commit b7e1e0fa0f157db3b5cc39e3f1dabf5381c892b3 Author: Deepak Narayanan Date: Thu Aug 17 11:43:17 2017 -0700 First commit diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..bdb7810 --- /dev/null +++ b/.gitignore @@ -0,0 +1,5 @@ +*.pyc +__pycache__/ +.eggs/ +*.egg-info/ +.cache diff --git a/pytorch/.gitignore b/pytorch/.gitignore new file mode 100644 index 0000000..f797795 --- /dev/null +++ b/pytorch/.gitignore @@ -0,0 +1,6 @@ +*.pyc +__pycache__/ +.eggs/ +*.egg-info/ +.cache +data/ diff --git a/pytorch/CIFAR10/benchmark/__init__.py b/pytorch/CIFAR10/benchmark/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/pytorch/CIFAR10/benchmark/infer.py b/pytorch/CIFAR10/benchmark/infer.py new file mode 100644 index 0000000..f52c194 --- /dev/null +++ b/pytorch/CIFAR10/benchmark/infer.py @@ -0,0 +1,140 @@ +import os +import timeit +from glob import glob +from collections import OrderedDict + +import click +import torch +import numpy as np +from torch.autograd import Variable +from torchvision import transforms +from torchvision import datasets + +from benchmark.train import load, MEAN, STD, save_result, MODELS + + +class PyTorchEngine: + def __init__(self, filename, use_cuda=False, name=None): + self.filename = filename + self.use_cuda = use_cuda + self.name = name + model, epoch, accuracy = load(self.filename) + + if self.use_cuda: + self.model = model.cuda() + else: + self.model = model.cpu() + self.epoch = epoch + self.accuracy = accuracy + + def pred(self, inputs): + inputs = Variable(inputs, requires_grad=False, volatile=True) + + if self.use_cuda: + inputs = inputs.cuda() + return self.model(inputs).data.cpu().numpy() + else: + return self.model(inputs).data.numpy() + + +def time_batch_size(dataset, batch_size, pred, use_cuda, repeat=100, bestof=3): + loader = torch.utils.data.DataLoader(dataset, batch_size=batch_size, + shuffle=False, pin_memory=use_cuda) + inputs, targets = loader.__iter__().next() + assert inputs.size(0) == batch_size + + times = timeit.repeat('pred(inputs)', globals=locals(), + repeat=repeat, number=1) + + return times + + +def infer_cifar10(dataset, engine, start=1, end=128, repeat=100, log2=True, + output=None): + if log2: + start = int(np.floor(np.log2(start))) + end = int(np.ceil(np.log2(end))) + assert start >= 0 + assert end >= start + batch_sizes = map(lambda x: 2**x, range(start, end + 1)) + else: + batch_sizes = range(start, end + 1) + results = [] + for batch_size in batch_sizes: + times = time_batch_size(dataset, batch_size, engine.pred, + engine.use_cuda, repeat=repeat) + + result = OrderedDict() + result['nodename'] = os.uname().nodename + result['model'] = engine.name + result['use_cuda'] = engine.use_cuda + result['batch_size'] = batch_size + result['mean'] = np.mean(times) + result['std'] = np.std(times) + result['throughput'] = batch_size / np.mean(times) + result['filename'] = engine.filename + if output is not None: + save_result(result, output) + + print('batch_size: {batch_size:4d}' + ' - mean: {mean:.4f}' + ' - std: {std:.4f}' + ' - throughput: {throughput:.4f}'.format(**result)) + results.append(result) + + return results + + +@click.command() +@click.option('--dataset-dir', default='./data/cifar10') +@click.option('--run-dir', default='./run/') +@click.option('--output-file', default='inference.csv') +@click.option('--start', '-s', default=1) +@click.option('--end', '-e', default=128) +@click.option('--repeat', '-r', default=100) +@click.option('--log2/--no-log2', default=True) +@click.option('--cpu/--no-cpu', default=True) +@click.option('--gpu/--no-gpu', default=True) +@click.option('--append', is_flag=True) +@click.option('--models', '-m', type=click.Choice(MODELS.keys()), + multiple=True) +def infer(dataset_dir, run_dir, output_file, start, end, repeat, log2, + cpu, gpu, append, models): + + transform_test = transforms.Compose([ + transforms.ToTensor(), + transforms.Normalize(MEAN, STD) + ]) + + testset = datasets.CIFAR10(root=dataset_dir, train=False, download=True, + transform=transform_test) + models = models or os.listdir(run_dir) + output_path = os.path.join(run_dir, output_file) + assert not os.path.exists(output_path) or append + for model in models: + model_dir = os.path.join(run_dir, model) + paths = glob(f"{model_dir}/*/checkpoint_best_model.t7") + assert len(paths) > 0 + path = os.path.abspath(paths[0]) + + print(f'Model: {model}') + print(f'Path: {path}') + + if cpu: + print('With CPU:') + engine = PyTorchEngine(path, use_cuda=False, name=model) + infer_cifar10(testset, engine, start=start, end=end, log2=log2, + repeat=repeat, output=output_path) + + if gpu and torch.cuda.is_available(): + print('With GPU:') + engine = PyTorchEngine(path, use_cuda=True, name=model) + # Warmup + time_batch_size(testset, 1, engine.pred, engine.use_cuda, repeat=1) + + infer_cifar10(testset, engine, start=start, end=end, log2=log2, + repeat=repeat, output=output_path) + + +if __name__ == '__main__': + infer() diff --git a/pytorch/CIFAR10/benchmark/models/densenet.py b/pytorch/CIFAR10/benchmark/models/densenet.py new file mode 100644 index 0000000..a1c59cd --- /dev/null +++ b/pytorch/CIFAR10/benchmark/models/densenet.py @@ -0,0 +1,108 @@ +'''DenseNet in PyTorch.''' +import math + +import torch +import torch.nn as nn +import torch.nn.functional as F + + +class Bottleneck(nn.Module): + def __init__(self, in_planes, growth_rate): + super(Bottleneck, self).__init__() + self.bn1 = nn.BatchNorm2d(in_planes) + self.conv1 = nn.Conv2d(in_planes, 4 * growth_rate, kernel_size=1, bias=False) + self.bn2 = nn.BatchNorm2d(4 * growth_rate) + self.conv2 = nn.Conv2d(4 * growth_rate, growth_rate, kernel_size=3, padding=1, bias=False) + + def forward(self, x): + out = self.conv1(F.relu(self.bn1(x))) + out = self.conv2(F.relu(self.bn2(out))) + out = torch.cat([out, x], 1) + return out + + +class Transition(nn.Module): + def __init__(self, in_planes, out_planes, last=False, pool_size=2): + super(Transition, self).__init__() + self.last = last + self.pool_size = pool_size + self.bn = nn.BatchNorm2d(in_planes) + if not self.last: + self.conv = nn.Conv2d(in_planes, out_planes, kernel_size=1, bias=False) + + def forward(self, x): + out = F.relu(self.bn(x)) + if not self.last: + out = self.conv(out) + out = F.avg_pool2d(out, self.pool_size) + return out + + +class DenseNet(nn.Module): + def __init__(self, block, nblocks, growth_rate=12, reduction=0.5, num_classes=10): + super(DenseNet, self).__init__() + # TODO: Add drop for CIFAR10 without data augmentation + self.growth_rate = growth_rate + + num_planes = 2 * growth_rate + self.conv1 = nn.Conv2d(3, num_planes, kernel_size=3, padding=1, bias=False) + + self.dense1 = self._make_dense_layers(block, num_planes, nblocks[0]) + num_planes += nblocks[0] * growth_rate + out_planes = int(math.floor(num_planes*reduction)) + self.trans1 = Transition(num_planes, out_planes) + num_planes = out_planes + + self.dense2 = self._make_dense_layers(block, num_planes, nblocks[1]) + num_planes += nblocks[1] * growth_rate + out_planes = int(math.floor(num_planes*reduction)) + self.trans2 = Transition(num_planes, out_planes) + num_planes = out_planes + + self.dense3 = self._make_dense_layers(block, num_planes, nblocks[2]) + num_planes += nblocks[2] * growth_rate + self.trans3 = Transition(num_planes, num_planes, last=True, pool_size=8) + + self.linear = nn.Linear(num_planes, num_classes) + + for m in self.modules(): + if isinstance(m, nn.Conv2d): + n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels + m.weight.data.normal_(0, math.sqrt(2. / n)) + elif isinstance(m, nn.BatchNorm2d): + m.weight.data.fill_(1) + m.bias.data.zero_() + + def _make_dense_layers(self, block, in_planes, nblock): + layers = [] + for i in range(nblock): + layers.append(block(in_planes, self.growth_rate)) + in_planes += self.growth_rate + return nn.Sequential(*layers) + + def forward(self, x): + out = self.conv1(x) + out = self.trans1(self.dense1(out)) + out = self.trans2(self.dense2(out)) + out = self.trans3(self.dense3(out)) + out = out.view(out.size(0), -1) + out = self.linear(out) + return out + + +def DenseNetBC(L, k): + assert (L - 4) % 6 == 0 + num_blocks = int((L - 4) / 6) + return DenseNet(Bottleneck, [num_blocks] * 3, growth_rate=k, reduction=0.5) + + +def DenseNetBC100(): + return DenseNetBC(100, 12) + + +def DenseNetBC250(): + return DenseNetBC(250, 24) + + +def DenseNetBC190(): + return DenseNetBC(190, 40) diff --git a/pytorch/CIFAR10/benchmark/models/resnet.py b/pytorch/CIFAR10/benchmark/models/resnet.py new file mode 100644 index 0000000..54d4d8e --- /dev/null +++ b/pytorch/CIFAR10/benchmark/models/resnet.py @@ -0,0 +1,372 @@ +import math +from functools import partial + +from torch import nn +from torch.nn import functional as F + + +class BasicBlock(nn.Module): + expansion = 1 + + def __init__(self, inplanes, planes, stride=1): + super().__init__() + self.conv1 = nn.Conv2d(inplanes, planes, 3, stride=stride, padding=1, + bias=False) + self.bn1 = nn.BatchNorm2d(planes) + + self.conv2 = nn.Conv2d(planes, planes, 3, padding=1, bias=False) + self.bn2 = nn.BatchNorm2d(planes) + + if stride != 1 or inplanes != (planes * self.expansion): + self.shortcut = nn.Sequential( + nn.Conv2d(inplanes, planes * self.expansion, 1, stride=stride, + bias=False), + nn.BatchNorm2d(planes * self.expansion) + ) + else: + self.shortcut = nn.Sequential() + + def forward(self, inputs): + H = self.conv1(inputs) + H = self.bn1(H) + H = F.relu(H) + + H = self.conv2(H) + H = self.bn2(H) + + H += self.shortcut(inputs) + outputs = F.relu(H) + + return outputs + + +class PreActBlock(nn.Module): + expansion = 1 + + def __init__(self, inplanes, planes, stride=1): + super().__init__() + self.bn1 = nn.BatchNorm2d(inplanes) + self.conv1 = nn.Conv2d(inplanes, planes, 3, stride=stride, padding=1, + bias=False) + + self.bn2 = nn.BatchNorm2d(planes) + self.conv2 = nn.Conv2d(planes, planes, 3, padding=1, bias=False) + + self.increasing = stride != 1 or inplanes != (planes * self.expansion) + if self.increasing: + self.shortcut = nn.Sequential( + nn.Conv2d(inplanes, planes * self.expansion, 1, stride=stride, + bias=False) + ) + else: + self.shortcut = nn.Sequential() + + def forward(self, inputs): + H = self.bn1(inputs) + H = F.relu(H) + if self.increasing: + inputs = H + H = self.conv1(H) + + H = self.bn2(H) + H = F.relu(H) + H = self.conv2(H) + + H += self.shortcut(inputs) + return H + + +class Bottleneck(nn.Module): + expansion = 4 + + def __init__(self, inplanes, planes, stride=1): + super().__init__() + self.conv1 = nn.Conv2d(inplanes, planes, 1, bias=False) + self.bn1 = nn.BatchNorm2d(planes) + + self.conv2 = nn.Conv2d(planes, planes, 3, stride=stride, + padding=1, bias=False) + self.bn2 = nn.BatchNorm2d(planes) + + self.conv3 = nn.Conv2d(planes, planes * 4, 1, bias=False) + self.bn3 = nn.BatchNorm2d(planes * 4) + + if stride != 1 or inplanes != (planes * self.expansion): + self.shortcut = nn.Sequential( + nn.Conv2d(inplanes, planes * self.expansion, 1, stride=stride, + bias=False), + nn.BatchNorm2d(planes * self.expansion) + ) + else: + self.shortcut = nn.Sequential() + + def forward(self, inputs): + H = self.conv1(inputs) + H = self.bn1(H) + H = F.relu(H) + + H = self.conv2(H) + H = self.bn2(H) + H = F.relu(H) + + H = self.conv3(H) + H = self.bn3(H) + + H += self.shortcut(inputs) + outputs = F.relu(H) + + return outputs + + +class ResNeXtBottleneck(nn.Module): + expansion = 4 + + def __init__(self, inplanes, planes, stride=1, cardinality=32, + base_width=4): + super().__init__() + + width = math.floor(planes * (base_width / 64.0)) + + self.conv1 = nn.Conv2d(inplanes, width * cardinality, 1, bias=False) + self.bn1 = nn.BatchNorm2d(width * cardinality) + + self.conv2 = nn.Conv2d(width * cardinality, width * cardinality, 3, + groups=cardinality, padding=1, stride=stride, + bias=False) + self.bn2 = nn.BatchNorm2d(width * cardinality) + + self.conv3 = nn.Conv2d(width * cardinality, planes * 4, 1, bias=False) + self.bn3 = nn.BatchNorm2d(planes * 4) + + if stride != 1 or inplanes != (planes * self.expansion): + self.shortcut = nn.Sequential( + nn.Conv2d(inplanes, planes * self.expansion, 1, stride=stride, + bias=False), + nn.BatchNorm2d(planes * self.expansion) + ) + else: + self.shortcut = nn.Sequential() + + def forward(self, inputs): + H = self.conv1(inputs) + H = self.bn1(H) + H = F.relu(H) + + H = self.conv2(H) + H = self.bn2(H) + H = F.relu(H) + + H = self.conv3(H) + H = self.bn3(H) + + H += self.shortcut(inputs) + outputs = F.relu(H) + + return outputs + + +class PreActBottleneck(nn.Module): + expansion = 4 + + def __init__(self, inplanes, planes, stride=1): + super().__init__() + self.bn1 = nn.BatchNorm2d(inplanes) + self.conv1 = nn.Conv2d(inplanes, planes, 1, bias=False) + + self.bn2 = nn.BatchNorm2d(planes) + self.conv2 = nn.Conv2d(planes, planes, 3, padding=1, stride=stride, + bias=False) + + self.bn3 = nn.BatchNorm2d(planes) + self.conv3 = nn.Conv2d(planes, planes * 4, 1, bias=False) + + self.increasing = stride != 1 or inplanes != (planes * self.expansion) + if self.increasing: + self.shortcut = nn.Sequential( + nn.Conv2d(inplanes, planes * self.expansion, 1, stride=stride, + bias=False) + ) + else: + self.shortcut = nn.Sequential() + + def forward(self, inputs): + H = self.bn1(inputs) + H = F.relu(H) + if self.increasing: + inputs = H + H = self.conv1(H) + + H = self.bn2(H) + H = F.relu(H) + H = self.conv2(H) + + H = self.bn3(H) + H = F.relu(H) + H = self.conv3(H) + + H += self.shortcut(inputs) + return H + + +class ResNet(nn.Module): + + def __init__(self, Block, layers, filters, num_classes=10, inplanes=None): + self.inplanes = inplanes or filters[0] + super().__init__() + + self.pre_act = 'Pre' in Block.__name__ + + self.conv1 = nn.Conv2d(3, self.inplanes, 3, padding=1, bias=False) + if not self.pre_act: + self.bn1 = nn.BatchNorm2d(self.inplanes) + + self.num_sections = len(layers) + for section_index, (size, planes) in enumerate(zip(layers, filters)): + section = [] + for layer_index in range(size): + if section_index != 0 and layer_index == 0: + stride = 2 + else: + stride = 1 + section.append(Block(self.inplanes, planes, stride=stride)) + self.inplanes = planes * Block.expansion + section = nn.Sequential(*section) + setattr(self, f'section_{section_index}', section) + + if self.pre_act: + self.bn1 = nn.BatchNorm2d(self.inplanes) + + self.fc = nn.Linear(filters[-1] * Block.expansion, num_classes) + + for m in self.modules(): + if isinstance(m, nn.Conv2d): + n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels + m.weight.data.normal_(0, math.sqrt(2. / n)) + elif isinstance(m, nn.BatchNorm2d): + m.weight.data.fill_(1) + m.bias.data.zero_() + + def forward(self, inputs): + H = self.conv1(inputs) + + if not self.pre_act: + H = self.bn1(H) + H = F.relu(H) + + for section_index in range(self.num_sections): + H = getattr(self, f'section_{section_index}')(H) + + if self.pre_act: + H = self.bn1(H) + H = F.relu(H) + + H = F.avg_pool2d(H, H.size()[2:]) + H = H.view(H.size(0), -1) + outputs = self.fc(H) + + return outputs + + +# From "Deep Residual Learning for Image Recognition" +def ResNet20(): + return ResNet(BasicBlock, layers=[3] * 3, filters=[16, 32, 64]) + + +def ResNet32(): + return ResNet(BasicBlock, layers=[5] * 3, filters=[16, 32, 64]) + + +def ResNet44(): + return ResNet(BasicBlock, layers=[7] * 3, filters=[16, 32, 64]) + + +def ResNet56(): + return ResNet(BasicBlock, layers=[9] * 3, filters=[16, 32, 64]) + + +def ResNet110(): + return ResNet(BasicBlock, layers=[18] * 3, filters=[16, 32, 64]) + + +def ResNet1202(): + return ResNet(BasicBlock, layers=[200] * 3, filters=[16, 32, 64]) + + +# Based on but not it "Identity Mappings in Deep Residual Networks" +def PreActResNet20(): + return ResNet(PreActBlock, layers=[3] * 3, filters=[16, 32, 64]) + + +def PreActResNet56(): + return ResNet(PreActBlock, layers=[9] * 3, filters=[16, 32, 64]) + + +def PreActResNet164Basic(): + return ResNet(PreActBlock, layers=[27] * 3, filters=[16, 32, 64]) + + +# From "Identity Mappings in Deep Residual Networks" +def PreActResNet110(): + return ResNet(PreActBlock, layers=[18] * 3, filters=[16, 32, 64]) + + +def PreActResNet164(): + return ResNet(PreActBottleneck, layers=[18] * 3, filters=[16, 32, 64]) + + +def PreActResNet1001(): + return ResNet(PreActBottleneck, layers=[111] * 3, filters=[16, 32, 64]) + + +# From "Wide Residual Networks" +def WRN(n, k): + assert (n - 4) % 6 == 0 + base_filters = [16, 32, 64] + filters = [num_filters * k for num_filters in base_filters] + d = (n - 4) / 2 # l = 2 + return ResNet(PreActBlock, layers=[int(d / 3)] * 3, filters=filters, + inplanes=16) + + +def WRN_40_4(): + return WRN(40, 4) + + +def WRN_16_8(): + return WRN(16, 8) + + +def WRN_28_10(): + return WRN(28, 10) + + +# From "Aggregated Residual Transformations for Deep Neural Networks" +def ResNeXt29(cardinality, base_width): + Block = partial(ResNeXtBottleneck, cardinality=cardinality, + base_width=base_width) + Block.__name__ = ResNeXtBottleneck.__name__ + Block.expansion = ResNeXtBottleneck.expansion + return ResNet(Block, layers=[3, 3, 3], filters=[64, 128, 256]) + + +# From kunagliu/pytorch +def ResNet18(): + return ResNet(BasicBlock, layers=[2, 2, 2, 2], filters=[64, 128, 256, 512]) + + +def ResNet34(): + return ResNet(BasicBlock, layers=[3, 4, 6, 3], filters=[64, 128, 256, 512]) + + +def ResNet50(): + return ResNet(Bottleneck, layers=[3, 4, 6, 3], filters=[64, 128, 256, 512]) + + +def ResNet101(): + return ResNet(Bottleneck, + layers=[3, 4, 23, 3], filters=[64, 128, 256, 512]) + + +def ResNet152(): + return ResNet(Bottleneck, + layers=[3, 8, 36, 3], filters=[64, 128, 256, 512]) diff --git a/pytorch/CIFAR10/benchmark/train.py b/pytorch/CIFAR10/benchmark/train.py new file mode 100644 index 0000000..78cdd8a --- /dev/null +++ b/pytorch/CIFAR10/benchmark/train.py @@ -0,0 +1,336 @@ +import os +import re +import json +from functools import reduce +from datetime import datetime +from collections import OrderedDict + +import click +import torch +import progressbar +from torch import nn, optim +from torch.autograd import Variable +from torchvision import transforms +from torchvision import datasets as dset + +from benchmark.models import resnet, densenet + +MEAN = (0.4914, 0.4822, 0.4465) +STD = (0.2023, 0.1994, 0.2010) + +MODELS = { + # "Deep Residual Learning for Image Recognition" + 'resnet20': resnet.ResNet20, + 'resnet32': resnet.ResNet32, + 'resnet44': resnet.ResNet44, + 'resnet56': resnet.ResNet56, + 'resnet110': resnet.ResNet110, + 'resnet1202': resnet.ResNet1202, + + # "Wide Residual Networks" + 'wrn-40-4': resnet.WRN_40_4, + 'wrn-16-8': resnet.WRN_16_8, + 'wrn-28-10': resnet.WRN_28_10, + + # Based on "Identity Mappings in Deep Residual Networks" + 'preact20': resnet.PreActResNet20, + 'preact56': resnet.PreActResNet56, + 'preact164-basic': resnet.PreActResNet164Basic, + + # "Identity Mappings in Deep Residual Networks" + 'preact110': resnet.PreActResNet110, + 'preact164': resnet.PreActResNet164, + 'preact1001': resnet.PreActResNet1001, + + # "Aggregated Residual Transformations for Deep Neural Networks" + 'resnext29-8-64': lambda _=None: resnet.ResNeXt29(8, 64), + 'resnext29-16-64': lambda _=None: resnet.ResNeXt29(16, 64), + + # "Densely Connected Convolutional Networks" + 'densenetbc100': densenet.DenseNetBC100, + 'densenetbc250': densenet.DenseNetBC250, + 'densenetbc190': densenet.DenseNetBC190, + + # Kuangliu/pytorch-cifar + 'resnet18': resnet.ResNet18, + 'resnet50': resnet.ResNet50, + 'resnet101': resnet.ResNet101, + 'resnet152': resnet.ResNet152, +} + + +def count_parameters(model): + c = map(lambda p: reduce(lambda x, y: x * y, p.size()), model.parameters()) + return sum(c) + + +def correct(outputs, targets, top=(1, )): + _, predictions = outputs.topk(max(top), dim=1, largest=True, sorted=True) + targets = targets.view(-1, 1).expand_as(predictions) + corrects = predictions.eq(targets).cpu().cumsum(1).sum(0) + tops = list(map(lambda k: corrects.data[0][k - 1], top)) + return tops + + +def save_result(result, path): + write_heading = not os.path.exists(path) + with open(path, mode='a') as out: + if write_heading: + out.write(",".join([str(k) for k, v in result.items()]) + '\n') + out.write(",".join([str(v) for k, v in result.items()]) + '\n') + + +def run(epoch, model, loader, criterion=None, optimizer=None, top=(1, 5), + use_cuda=False, tracking=None, max_value=None, train=True): + + assert criterion is not None or not train, 'Need criterion to train model' + assert optimizer is not None or not train, 'Need optimizer to train model' + max_value = max_value or progressbar.UnknownLength + bar = progressbar.ProgressBar(max_value=max_value) + total = 0 + correct_counts = {} + if train: + model.train() + else: + model.eval() + + start = datetime.now() + for batch_index, (inputs, targets) in enumerate(loader): + inputs = Variable(inputs, requires_grad=False, volatile=not train) + targets = Variable(targets, requires_grad=False, volatile=not train) + + if use_cuda: + inputs = inputs.cuda() + targets = targets.cuda() + + outputs = model(inputs) + + if train: + loss = criterion(outputs, targets) + optimizer.zero_grad() + loss.backward() + optimizer.step() + + _, predictions = torch.max(outputs.data, 1) + batch_size = targets.size(0) + top_correct = correct(outputs, targets, top=top) + total += batch_size + for k, count in zip(top, top_correct): + correct_counts[k] = correct_counts.get(k, 0) + count + + end = datetime.now() + if tracking is not None: + result = OrderedDict() + result['timestamp'] = datetime.now() + result['batch_duration'] = end - start + result['epoch'] = epoch + result['batch'] = batch_index + result['batch_size'] = batch_size + for i, k in enumerate(top): + result['top{}_correct'.format(k)] = top_correct[i] + if train: + result['loss'] = loss.data[0] + save_result(result, tracking) + + bar.update(batch_index + 1) + start = datetime.now() + + print() + if train: + message = 'Training accuracy of' + else: + message = 'Test accuracy of' + for k in top: + accuracy = correct_counts[k] / total + message += ' top-{}: {}'.format(k, accuracy) + print(message) + return (1. * correct_counts[top[0]]) / total, batch_index + 1 + + +def save(model, directory, epoch, accuracy, use_cuda=False, filename=None): + state = { + 'model': model.module if use_cuda else model, + 'epoch': epoch, + 'accuracy': accuracy + } + + filename = filename or 'checkpoint_{}.t7'.format(epoch) + torch.save(state, os.path.join(directory, filename)) + + +def save_config(config, run_dir): + path = os.path.join(run_dir, "config_{}.json".format(config['timestamp'])) + with open(path, 'w') as config_file: + json.dump(config, config_file) + config_file.write('\n') + + +def load(path): + assert os.path.exists(path) + state = torch.load(path) + model = state['model'] + epoch = state['epoch'] + accuracy = state['accuracy'] + return model, epoch, accuracy + + +def latest_file(model): + restore = f'./run/{model}' + timestamps = sorted(os.listdir(restore)) + assert len(timestamps) > 0 + run_dir = os.path.join(restore, timestamps[-1]) + files = os.listdir(run_dir) + max_checkpoint = -1 + for filename in files: + if re.search('checkpoint_\d+.t7', filename): + num = int(re.search('\d+', filename).group()) + + if num > max_checkpoint: + max_checkpoint = num + max_checkpoint_file = filename + + assert max_checkpoint != -1 + return os.path.join(run_dir, max_checkpoint_file) + + +@click.command() +@click.option('--dataset-dir', default='./data/cifar10') +@click.option('--checkpoint', '-c', type=click.Choice(['best', 'all', 'last']), + default='last') +@click.option('--restore', '-r') +@click.option('--tracking/--no-tracking', default=True) +@click.option('--cuda/--no-cuda', default=True) +@click.option('--epochs', '-e', default=200) +@click.option('--batch-size', '-b', default=32) +@click.option('--learning-rate', '-l', default=1e-3) +@click.option('--sgd', 'optimizer', flag_value='sgd') +@click.option('--adam', 'optimizer', flag_value='adam', default=True) +@click.option('--augmentation/--no-augmentation', default=True) +@click.option('--num-workers', type=int) +@click.option('--weight-decay', default=5e-4) +@click.option('--model', '-m', type=click.Choice(MODELS.keys()), + default='resnet20') +def main(dataset_dir, checkpoint, restore, tracking, cuda, epochs, + batch_size, learning_rate, optimizer, augmentation, num_workers, + weight_decay, model): + timestamp = "{:.0f}".format(datetime.utcnow().timestamp()) + config = {k: v for k, v in locals().items()} + + use_cuda = cuda and torch.cuda.is_available() + if use_cuda: + num_workers = num_workers or torch.cuda.device_count() + else: + num_workers = num_workers or 1 + + print(f"using {num_workers} workers for data loading") + + print("Preparing data:") + + if augmentation: + transform_train = [ + transforms.RandomCrop(32, padding=4), + transforms.RandomHorizontalFlip() + ] + else: + transform_train = [] + + transform_train = transforms.Compose(transform_train + [ + transforms.ToTensor(), + transforms.Normalize(MEAN, STD), + ]) + + trainset = dset.CIFAR10(root=dataset_dir, train=True, download=True, + transform=transform_train) + train_loader = torch.utils.data.DataLoader( + trainset, batch_size=batch_size, shuffle=True, num_workers=num_workers, + pin_memory=use_cuda) + + transform_test = transforms.Compose([ + transforms.ToTensor(), + transforms.Normalize(MEAN, STD), + ]) + + testset = dset.CIFAR10(root=dataset_dir, train=False, download=True, + transform=transform_test) + test_loader = torch.utils.data.DataLoader( + testset, batch_size=batch_size, shuffle=False, num_workers=num_workers, + pin_memory=use_cuda) + + if restore is not None: + if restore == 'latest': + restore = latest_file(model) + print(f'Restoring model from {restore}') + model, start_epoch, best_accuracy = load(restore) + start_epoch += 1 + print('Starting accuracy is {}'.format(best_accuracy)) + run_dir = os.path.split(restore)[0] + else: + print(f'Building {model} model') + best_accuracy = -1 + start_epoch = 1 + run_dir = f"./run/{model}/{timestamp}" + model = MODELS[model]() + + if not os.path.exists(run_dir): + os.makedirs(run_dir) + save_config(config, run_dir) + + print(model) + print("{} parameters".format(count_parameters(model))) + print(f"Run directory set to {run_dir}") + + # Save model text description + with open(os.path.join(run_dir, 'model.txt'), 'w') as file: + file.write(str(model)) + + if tracking: + train_results_file = os.path.join(run_dir, 'train_results.csv') + test_results_file = os.path.join(run_dir, 'test_results.csv') + else: + train_results_file = None + test_results_file = None + + if use_cuda: + print('Copying model to GPU') + model.cuda() + model = torch.nn.DataParallel( + model, device_ids=range(torch.cuda.device_count())) + criterion = nn.CrossEntropyLoss() + + # Other parameters? + if optimizer == 'adam': + optimizer = optim.Adam(model.parameters(), lr=learning_rate) + elif optimizer == 'sgd': + optimizer = optim.SGD(model.parameters(), lr=learning_rate, + momentum=0.9, + weight_decay=weight_decay) + else: + raise NotImplementedError("Unknown optimizer: {}".format(optimizer)) + + train_max_value = None + test_max_value = None + end_epoch = start_epoch + epochs + for epoch in range(start_epoch, end_epoch): + print('Epoch {} of {}'.format(epoch, end_epoch - 1)) + train_acc, train_max_value = run(epoch, model, train_loader, criterion, + optimizer, use_cuda=use_cuda, + tracking=train_results_file, + max_value=train_max_value, train=True) + + test_acc, test_max_value = run(epoch, model, test_loader, + use_cuda=use_cuda, + tracking=test_results_file, train=False) + + if test_acc > best_accuracy: + print('New best model!') + save(model, run_dir, epoch, test_acc, use_cuda=use_cuda, + filename='checkpoint_best_model.t7') + best_accuracy = test_acc + + last_epoch = epoch == (end_epoch - 1) + if checkpoint == 'all' or (checkpoint == 'last' and last_epoch): + save(model, run_dir, epoch, test_acc, use_cuda=use_cuda) + + +if __name__ == '__main__': + main() diff --git a/pytorch/CIFAR10/setup.py b/pytorch/CIFAR10/setup.py new file mode 100644 index 0000000..5f2a1ed --- /dev/null +++ b/pytorch/CIFAR10/setup.py @@ -0,0 +1,20 @@ +from setuptools import setup + +setup( + name='benchmark', + version='0.0.0', + url='http://www.codycoleman.com', + author='Cody Austun Coleman', + author_email='cody.coleman@cs.stanford.edu', + packages=['benchmark'], + entry_points={ + 'console_scripts': [ + 'bench = benchmark.train:main' + ] + }, + install_requires=[ + 'torchvision', + 'click', + 'progressbar2' + ] +) diff --git a/tensorflow/CIFAR10/README.md b/tensorflow/CIFAR10/README.md new file mode 100644 index 0000000..1e0fde6 --- /dev/null +++ b/tensorflow/CIFAR10/README.md @@ -0,0 +1,18 @@ +# ResNets on TensorFlow + +To train a ResNet, run, + +```bash +python3 resnet/resnet_main.py --train_data_path=cifar10/data_batch* --log_root=data/resnet20/log_root \ + --train_dir=data/resnet20/log_root/train --dataset='cifar10' --model=resnet20 \ + --num_gpus=1 --checkpoint_dir=data/resnet20/checkpoints --data_format=NCHW +``` + +To evaluate resulting checkpoints, run, + +```bash +python3 eval_checkpoints.py -i data/resnet20/checkpoints \ + -c "python3 resnet/resnet_main.py --mode=eval --eval_data_path=cifar10/test_batch.bin --eval_dir=data/resnet20/log_root/eval --dataset='cifar10' --model=resnet20 --num_gpus=1 --eval_batch_count=100 --eval_once=True --data_format=NCHW" +``` + +Make sure to first follow the instructions in `resnet/README.md` to get necessary data, etc. diff --git a/tensorflow/CIFAR10/eval_checkpoints.py b/tensorflow/CIFAR10/eval_checkpoints.py new file mode 100644 index 0000000..cd3f900 --- /dev/null +++ b/tensorflow/CIFAR10/eval_checkpoints.py @@ -0,0 +1,59 @@ +import argparse +import os +import subprocess +import sys + +def main(checkpoints_path, command, start_cnt): + cnt = start_cnt + + times = {} + cum_time = 0.0 + with open(os.path.join(checkpoints_path, "times.log"), 'r') as f: + output = f.read().strip() + output_lines = output.split('\n') + for output_line in output_lines: + [step, time] = output_line.split('\t') + step = int(step.split(': ')[1]) + time = float(time.split(': ')[1]) + cum_time += time + times[step] = cum_time + + print("Time (in secs)\tNumber of minibatches\tTop 1 accuracy\tTop 5 accuracy") + while True: + ckpt_path = ("%5d" % cnt).replace(' ', '0') + full_ckpt_path = os.path.join(checkpoints_path, ckpt_path) + if not os.path.exists(full_ckpt_path): + break + if len(os.listdir(full_ckpt_path)) <= 2: + cnt += 1 + continue + full_command = command + " --log_root=%s 2>/dev/null" % full_ckpt_path + output = subprocess.check_output(full_command, shell=True) + output = output.decode('utf8').strip() + for line in output.split('\n'): + if "Precision" in line and "Recall" in line: + tokens = line.split(", ") # TODO: Nasty hack, make more robust. + precision_at_1 = float(tokens[0].split()[-1]) + recall_at_5 = float(tokens[1].split()[-1]) + step = int(tokens[2].split()[3]) + stats = [times[step], step, precision_at_1, recall_at_5] + print("\t".join([str(stat) for stat in stats])) + sys.stdout.flush() + cnt += 1 + + +if __name__ == '__main__': + parser = argparse.ArgumentParser( + description=("Backup model checkpoints periodically") + ) + parser.add_argument('-i', "--checkpoints_path", type=str, required=True, + help="Path to dumped model checkpoints") + parser.add_argument('-c', "--command", type=str, required=True, + help="Command to evaluate each individual checkpoint") + parser.add_argument('-s', "--start_cnt", type=int, default=1, + help="Count to start evaluating checkpoints from") + + cmdline_args = parser.parse_args() + opt_dict = vars(cmdline_args) + + main(opt_dict["checkpoints_path"], opt_dict["command"], opt_dict["start_cnt"]) diff --git a/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet164_b_train.log b/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet164_b_train.log new file mode 100644 index 0000000..3d9bc77 --- /dev/null +++ b/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet164_b_train.log @@ -0,0 +1,2205 @@ +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 32, 32, 64) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 16, 16, 128) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +INFO:tensorflow:image after unit (128, 8, 8, 256) +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/1.69m params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (256x10, 2.56k/2.56k params) + logit/biases (10, 10/10 params) + unit_1_0/common_bn_relu/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (1x1x16x16, 256/256 params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub3/bn3/beta (16, 16/16 params) + unit_1_0/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_0/sub_add/project/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_1/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_1/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub3/bn3/beta (16, 16/16 params) + unit_1_1/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_10/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_10/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_10/sub2/bn2/beta (16, 16/16 params) + unit_1_10/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/sub3/bn3/beta (16, 16/16 params) + unit_1_10/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_11/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_11/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_11/sub2/bn2/beta (16, 16/16 params) + unit_1_11/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/sub3/bn3/beta (16, 16/16 params) + unit_1_11/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_12/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_12/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_12/sub2/bn2/beta (16, 16/16 params) + unit_1_12/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/sub3/bn3/beta (16, 16/16 params) + unit_1_12/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_13/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_13/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_13/sub2/bn2/beta (16, 16/16 params) + unit_1_13/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/sub3/bn3/beta (16, 16/16 params) + unit_1_13/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_14/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_14/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_14/sub2/bn2/beta (16, 16/16 params) + unit_1_14/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/sub3/bn3/beta (16, 16/16 params) + unit_1_14/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_15/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_15/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_15/sub2/bn2/beta (16, 16/16 params) + unit_1_15/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/sub3/bn3/beta (16, 16/16 params) + unit_1_15/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_16/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_16/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_16/sub2/bn2/beta (16, 16/16 params) + unit_1_16/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/sub3/bn3/beta (16, 16/16 params) + unit_1_16/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_17/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_17/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_17/sub2/bn2/beta (16, 16/16 params) + unit_1_17/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/sub3/bn3/beta (16, 16/16 params) + unit_1_17/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_2/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_2/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub3/bn3/beta (16, 16/16 params) + unit_1_2/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_3/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_3/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub3/bn3/beta (16, 16/16 params) + unit_1_3/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_4/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_4/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub3/bn3/beta (16, 16/16 params) + unit_1_4/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_5/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_5/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub3/bn3/beta (16, 16/16 params) + unit_1_5/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_6/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_6/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub3/bn3/beta (16, 16/16 params) + unit_1_6/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_7/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_7/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub3/bn3/beta (16, 16/16 params) + unit_1_7/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_8/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_8/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub3/bn3/beta (16, 16/16 params) + unit_1_8/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_9/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_9/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_9/sub2/bn2/beta (16, 16/16 params) + unit_1_9/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/sub3/bn3/beta (16, 16/16 params) + unit_1_9/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_2_0/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_2_0/sub1/conv1/DW (1x1x64x32, 2.05k/2.05k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_0/sub3/bn3/beta (32, 32/32 params) + unit_2_0/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_0/sub_add/project/DW (1x1x64x128, 8.19k/8.19k params) + unit_2_1/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_1/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub3/bn3/beta (32, 32/32 params) + unit_2_1/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_10/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_10/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_10/sub2/bn2/beta (32, 32/32 params) + unit_2_10/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/sub3/bn3/beta (32, 32/32 params) + unit_2_10/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_11/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_11/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_11/sub2/bn2/beta (32, 32/32 params) + unit_2_11/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/sub3/bn3/beta (32, 32/32 params) + unit_2_11/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_12/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_12/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_12/sub2/bn2/beta (32, 32/32 params) + unit_2_12/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/sub3/bn3/beta (32, 32/32 params) + unit_2_12/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_13/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_13/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_13/sub2/bn2/beta (32, 32/32 params) + unit_2_13/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/sub3/bn3/beta (32, 32/32 params) + unit_2_13/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_14/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_14/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_14/sub2/bn2/beta (32, 32/32 params) + unit_2_14/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/sub3/bn3/beta (32, 32/32 params) + unit_2_14/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_15/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_15/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_15/sub2/bn2/beta (32, 32/32 params) + unit_2_15/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/sub3/bn3/beta (32, 32/32 params) + unit_2_15/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_16/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_16/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_16/sub2/bn2/beta (32, 32/32 params) + unit_2_16/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/sub3/bn3/beta (32, 32/32 params) + unit_2_16/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_17/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_17/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_17/sub2/bn2/beta (32, 32/32 params) + unit_2_17/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/sub3/bn3/beta (32, 32/32 params) + unit_2_17/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_2/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_2/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub3/bn3/beta (32, 32/32 params) + unit_2_2/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_3/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_3/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub3/bn3/beta (32, 32/32 params) + unit_2_3/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_4/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_4/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub3/bn3/beta (32, 32/32 params) + unit_2_4/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_5/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_5/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub3/bn3/beta (32, 32/32 params) + unit_2_5/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_6/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_6/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub3/bn3/beta (32, 32/32 params) + unit_2_6/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_7/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_7/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub3/bn3/beta (32, 32/32 params) + unit_2_7/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_8/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_8/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub3/bn3/beta (32, 32/32 params) + unit_2_8/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_9/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_9/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_9/sub2/bn2/beta (32, 32/32 params) + unit_2_9/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/sub3/bn3/beta (32, 32/32 params) + unit_2_9/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_3_0/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_3_0/sub1/conv1/DW (1x1x128x64, 8.19k/8.19k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_0/sub3/bn3/beta (64, 64/64 params) + unit_3_0/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_0/sub_add/project/DW (1x1x128x256, 32.77k/32.77k params) + unit_3_1/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_1/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub3/bn3/beta (64, 64/64 params) + unit_3_1/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_10/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_10/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_10/sub2/bn2/beta (64, 64/64 params) + unit_3_10/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/sub3/bn3/beta (64, 64/64 params) + unit_3_10/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_11/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_11/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_11/sub2/bn2/beta (64, 64/64 params) + unit_3_11/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/sub3/bn3/beta (64, 64/64 params) + unit_3_11/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_12/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_12/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_12/sub2/bn2/beta (64, 64/64 params) + unit_3_12/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/sub3/bn3/beta (64, 64/64 params) + unit_3_12/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_13/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_13/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_13/sub2/bn2/beta (64, 64/64 params) + unit_3_13/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/sub3/bn3/beta (64, 64/64 params) + unit_3_13/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_14/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_14/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_14/sub2/bn2/beta (64, 64/64 params) + unit_3_14/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/sub3/bn3/beta (64, 64/64 params) + unit_3_14/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_15/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_15/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_15/sub2/bn2/beta (64, 64/64 params) + unit_3_15/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/sub3/bn3/beta (64, 64/64 params) + unit_3_15/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_16/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_16/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_16/sub2/bn2/beta (64, 64/64 params) + unit_3_16/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/sub3/bn3/beta (64, 64/64 params) + unit_3_16/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_17/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_17/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_17/sub2/bn2/beta (64, 64/64 params) + unit_3_17/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/sub3/bn3/beta (64, 64/64 params) + unit_3_17/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_2/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_2/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub3/bn3/beta (64, 64/64 params) + unit_3_2/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_3/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_3/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub3/bn3/beta (64, 64/64 params) + unit_3_3/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_4/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_4/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub3/bn3/beta (64, 64/64 params) + unit_3_4/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_5/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_5/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub3/bn3/beta (64, 64/64 params) + unit_3_5/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_6/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_6/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub3/bn3/beta (64, 64/64 params) + unit_3_6/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_7/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_7/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub3/bn3/beta (64, 64/64 params) + unit_3_7/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_8/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_8/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub3/bn3/beta (64, 64/64 params) + unit_3_8/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_9/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_9/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_9/sub2/bn2/beta (64, 64/64 params) + unit_3_9/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/sub3/bn3/beta (64, 64/64 params) + unit_3_9/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_last/final_bn/beta (256, 256/256 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/62.59b flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub_add/project/Conv2D (536.87m/536.87m flops) + unit_2_0/sub_add/project/Conv2D (536.87m/536.87m flops) + unit_3_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_0/sub_add/project/Conv2D (268.44m/268.44m flops) + unit_1_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_0/sub1/conv1/Conv2D (134.22m/134.22m flops) + unit_3_0/sub1/conv1/Conv2D (134.22m/134.22m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + unit_1_0/sub1/conv1/Conv2D (67.11m/67.11m flops) + logit/xw_plus_b (1.28k/656.64k flops) + logit/xw_plus_b/MatMul (655.36k/655.36k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (655.36k/655.36k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (655.36k/655.36k flops) + +======================End of Report========================== +2017-07-31 22:19:25.649555: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 16 visible devices +2017-07-31 22:19:25.654765: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x9bd5c20 executing computations on platform Host. Devices: +2017-07-31 22:19:25.654855: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +INFO:tensorflow:step = 1, loss = 8.39792, precision = 0.109375 +INFO:tensorflow:global_step/sec: 0.141326 +INFO:tensorflow:step = 101, loss = 7.80716, precision = 0.328125 (707.586 sec) +INFO:tensorflow:global_step/sec: 0.146364 +INFO:tensorflow:step = 201, loss = 7.57217, precision = 0.429688 (683.227 sec) +INFO:tensorflow:global_step/sec: 0.155155 +INFO:tensorflow:step = 301, loss = 7.39057, precision = 0.476562 (644.517 sec) +total_params: 1691146 +Saved checkpoint after 1 epoch(s) to data/resnet164/checkpoints/00001... +INFO:tensorflow:global_step/sec: 0.156548 +INFO:tensorflow:step = 401, loss = 7.8567, precision = 0.289062 (638.780 sec) +INFO:tensorflow:global_step/sec: 0.157992 +INFO:tensorflow:step = 501, loss = 6.90182, precision = 0.390625 (632.942 sec) +INFO:tensorflow:global_step/sec: 0.1486 +INFO:tensorflow:step = 601, loss = 5.96754, precision = 0.5625 (672.947 sec) +INFO:tensorflow:global_step/sec: 0.147971 +INFO:tensorflow:step = 701, loss = 5.58256, precision = 0.515625 (675.808 sec) +Saved checkpoint after 2 epoch(s) to data/resnet164/checkpoints/00002... +INFO:tensorflow:global_step/sec: 0.153219 +INFO:tensorflow:step = 801, loss = 4.70894, precision = 0.765625 (652.660 sec) +INFO:tensorflow:global_step/sec: 0.1483 +INFO:tensorflow:step = 901, loss = 4.46725, precision = 0.6875 (674.311 sec) +INFO:tensorflow:global_step/sec: 0.149315 +INFO:tensorflow:step = 1001, loss = 4.15475, precision = 0.648438 (669.727 sec) +INFO:tensorflow:global_step/sec: 0.149368 +INFO:tensorflow:step = 1101, loss = 3.72393, precision = 0.75 (669.487 sec) +Saved checkpoint after 3 epoch(s) to data/resnet164/checkpoints/00003... +INFO:tensorflow:global_step/sec: 0.153058 +INFO:tensorflow:step = 1201, loss = 3.62801, precision = 0.664062 (653.345 sec) +INFO:tensorflow:global_step/sec: 0.147054 +INFO:tensorflow:step = 1301, loss = 3.21184, precision = 0.671875 (680.024 sec) +INFO:tensorflow:global_step/sec: 0.132915 +INFO:tensorflow:step = 1401, loss = 3.11421, precision = 0.710938 (752.358 sec) +INFO:tensorflow:global_step/sec: 0.139848 +INFO:tensorflow:step = 1501, loss = 2.85305, precision = 0.71875 (715.062 sec) +Saved checkpoint after 4 epoch(s) to data/resnet164/checkpoints/00004... +INFO:tensorflow:global_step/sec: 0.146 +INFO:tensorflow:step = 1601, loss = 2.41076, precision = 0.78125 (684.934 sec) +INFO:tensorflow:global_step/sec: 0.156125 +INFO:tensorflow:step = 1701, loss = 2.42236, precision = 0.8125 (640.511 sec) +INFO:tensorflow:global_step/sec: 0.158342 +INFO:tensorflow:step = 1801, loss = 2.1539, precision = 0.789062 (631.543 sec) +INFO:tensorflow:global_step/sec: 0.156009 +INFO:tensorflow:step = 1901, loss = 2.08461, precision = 0.78125 (640.988 sec) +Saved checkpoint after 5 epoch(s) to data/resnet164/checkpoints/00005... +INFO:tensorflow:global_step/sec: 0.15496 +INFO:tensorflow:step = 2001, loss = 1.94207, precision = 0.8125 (645.326 sec) +INFO:tensorflow:global_step/sec: 0.147521 +INFO:tensorflow:step = 2101, loss = 1.96057, precision = 0.757812 (677.871 sec) +INFO:tensorflow:global_step/sec: 0.149878 +INFO:tensorflow:step = 2201, loss = 1.75131, precision = 0.773438 (667.210 sec) +INFO:tensorflow:global_step/sec: 0.148619 +INFO:tensorflow:step = 2301, loss = 1.70209, precision = 0.804688 (672.862 sec) +Saved checkpoint after 6 epoch(s) to data/resnet164/checkpoints/00006... +INFO:tensorflow:global_step/sec: 0.157437 +INFO:tensorflow:step = 2401, loss = 1.57474, precision = 0.8125 (635.174 sec) +INFO:tensorflow:global_step/sec: 0.159403 +INFO:tensorflow:step = 2501, loss = 1.50604, precision = 0.796875 (627.342 sec) +INFO:tensorflow:global_step/sec: 0.159047 +INFO:tensorflow:step = 2601, loss = 1.43421, precision = 0.78125 (628.746 sec) +INFO:tensorflow:global_step/sec: 0.131822 +INFO:tensorflow:step = 2701, loss = 1.27324, precision = 0.820312 (758.597 sec) +Saved checkpoint after 7 epoch(s) to data/resnet164/checkpoints/00007... +INFO:tensorflow:global_step/sec: 0.152358 +INFO:tensorflow:step = 2801, loss = 1.33855, precision = 0.804688 (656.350 sec) +INFO:tensorflow:global_step/sec: 0.155047 +INFO:tensorflow:step = 2901, loss = 1.42244, precision = 0.726562 (644.965 sec) +INFO:tensorflow:global_step/sec: 0.152586 +INFO:tensorflow:step = 3001, loss = 1.20759, precision = 0.773438 (655.367 sec) +INFO:tensorflow:global_step/sec: 0.154377 +INFO:tensorflow:step = 3101, loss = 1.26566, precision = 0.765625 (647.766 sec) +Saved checkpoint after 8 epoch(s) to data/resnet164/checkpoints/00008... +INFO:tensorflow:global_step/sec: 0.158963 +INFO:tensorflow:step = 3201, loss = 1.03803, precision = 0.84375 (629.077 sec) +INFO:tensorflow:global_step/sec: 0.160421 +INFO:tensorflow:step = 3301, loss = 1.06984, precision = 0.820312 (623.360 sec) +INFO:tensorflow:global_step/sec: 0.161753 +INFO:tensorflow:step = 3401, loss = 1.16329, precision = 0.804688 (618.227 sec) +INFO:tensorflow:global_step/sec: 0.162384 +INFO:tensorflow:step = 3501, loss = 1.00844, precision = 0.804688 (615.825 sec) +Saved checkpoint after 9 epoch(s) to data/resnet164/checkpoints/00009... +INFO:tensorflow:global_step/sec: 0.15796 +INFO:tensorflow:step = 3601, loss = 1.05155, precision = 0.835938 (633.071 sec) +INFO:tensorflow:global_step/sec: 0.155326 +INFO:tensorflow:step = 3701, loss = 0.880639, precision = 0.898438 (643.806 sec) +INFO:tensorflow:global_step/sec: 0.138792 +INFO:tensorflow:step = 3801, loss = 0.921566, precision = 0.875 (720.501 sec) +INFO:tensorflow:global_step/sec: 0.145138 +INFO:tensorflow:step = 3901, loss = 1.23889, precision = 0.765625 (689.000 sec) +Saved checkpoint after 10 epoch(s) to data/resnet164/checkpoints/00010... +INFO:tensorflow:global_step/sec: 0.148375 +INFO:tensorflow:step = 4001, loss = 0.900903, precision = 0.867188 (673.968 sec) +INFO:tensorflow:global_step/sec: 0.166377 +INFO:tensorflow:step = 4101, loss = 0.954465, precision = 0.789062 (601.043 sec) +INFO:tensorflow:global_step/sec: 0.16508 +INFO:tensorflow:step = 4201, loss = 0.900192, precision = 0.828125 (605.767 sec) +Saved checkpoint after 11 epoch(s) to data/resnet164/checkpoints/00011... +INFO:tensorflow:global_step/sec: 0.164835 +INFO:tensorflow:step = 4301, loss = 1.09259, precision = 0.78125 (606.667 sec) +INFO:tensorflow:global_step/sec: 0.164335 +INFO:tensorflow:step = 4401, loss = 0.920248, precision = 0.820312 (608.513 sec) +INFO:tensorflow:global_step/sec: 0.160938 +INFO:tensorflow:step = 4501, loss = 1.01962, precision = 0.804688 (621.358 sec) +INFO:tensorflow:global_step/sec: 0.157872 +INFO:tensorflow:step = 4601, loss = 0.794193, precision = 0.867188 (633.425 sec) +Saved checkpoint after 12 epoch(s) to data/resnet164/checkpoints/00012... +INFO:tensorflow:global_step/sec: 0.157548 +INFO:tensorflow:step = 4701, loss = 0.844666, precision = 0.828125 (634.729 sec) +INFO:tensorflow:global_step/sec: 0.160231 +INFO:tensorflow:step = 4801, loss = 0.650026, precision = 0.898438 (624.098 sec) +INFO:tensorflow:global_step/sec: 0.163578 +INFO:tensorflow:step = 4901, loss = 0.768703, precision = 0.867188 (611.330 sec) +INFO:tensorflow:global_step/sec: 0.169972 +INFO:tensorflow:step = 5001, loss = 0.83426, precision = 0.84375 (588.331 sec) +Saved checkpoint after 13 epoch(s) to data/resnet164/checkpoints/00013... +INFO:tensorflow:global_step/sec: 0.170379 +INFO:tensorflow:step = 5101, loss = 0.746989, precision = 0.898438 (586.926 sec) +INFO:tensorflow:global_step/sec: 0.171737 +INFO:tensorflow:step = 5201, loss = 0.881186, precision = 0.78125 (582.287 sec) +INFO:tensorflow:global_step/sec: 0.171976 +INFO:tensorflow:step = 5301, loss = 0.792536, precision = 0.835938 (581.478 sec) +INFO:tensorflow:global_step/sec: 0.168372 +INFO:tensorflow:step = 5401, loss = 0.903613, precision = 0.835938 (593.922 sec) +Saved checkpoint after 14 epoch(s) to data/resnet164/checkpoints/00014... +INFO:tensorflow:global_step/sec: 0.166072 +INFO:tensorflow:step = 5501, loss = 0.863237, precision = 0.820312 (602.149 sec) +INFO:tensorflow:global_step/sec: 0.158323 +INFO:tensorflow:step = 5601, loss = 0.732874, precision = 0.898438 (631.621 sec) +INFO:tensorflow:global_step/sec: 0.159066 +INFO:tensorflow:step = 5701, loss = 0.693553, precision = 0.890625 (628.671 sec) +INFO:tensorflow:global_step/sec: 0.166428 +INFO:tensorflow:step = 5801, loss = 0.777672, precision = 0.851562 (600.860 sec) +Saved checkpoint after 15 epoch(s) to data/resnet164/checkpoints/00015... +INFO:tensorflow:global_step/sec: 0.164942 +INFO:tensorflow:step = 5901, loss = 0.940717, precision = 0.804688 (606.274 sec) +INFO:tensorflow:global_step/sec: 0.16181 +INFO:tensorflow:step = 6001, loss = 0.736267, precision = 0.859375 (618.010 sec) +INFO:tensorflow:global_step/sec: 0.160645 +INFO:tensorflow:step = 6101, loss = 0.759816, precision = 0.84375 (622.491 sec) +INFO:tensorflow:global_step/sec: 0.161629 +INFO:tensorflow:step = 6201, loss = 0.800368, precision = 0.851562 (618.702 sec) +Saved checkpoint after 16 epoch(s) to data/resnet164/checkpoints/00016... +INFO:tensorflow:global_step/sec: 0.159894 +INFO:tensorflow:step = 6301, loss = 0.998926, precision = 0.796875 (625.416 sec) +INFO:tensorflow:global_step/sec: 0.164876 +INFO:tensorflow:step = 6401, loss = 0.767139, precision = 0.851562 (606.515 sec) +INFO:tensorflow:global_step/sec: 0.159457 +INFO:tensorflow:step = 6501, loss = 0.670936, precision = 0.882812 (627.127 sec) +INFO:tensorflow:global_step/sec: 0.152492 +INFO:tensorflow:step = 6601, loss = 0.784466, precision = 0.84375 (655.771 sec) +Saved checkpoint after 17 epoch(s) to data/resnet164/checkpoints/00017... +INFO:tensorflow:global_step/sec: 0.14903 +INFO:tensorflow:step = 6701, loss = 0.773602, precision = 0.859375 (671.008 sec) +INFO:tensorflow:global_step/sec: 0.156358 +INFO:tensorflow:step = 6801, loss = 0.774479, precision = 0.84375 (639.557 sec) +INFO:tensorflow:global_step/sec: 0.141686 +INFO:tensorflow:step = 6901, loss = 0.743528, precision = 0.851562 (705.788 sec) +INFO:tensorflow:global_step/sec: 0.14265 +INFO:tensorflow:step = 7001, loss = 0.705053, precision = 0.851562 (701.016 sec) +Saved checkpoint after 18 epoch(s) to data/resnet164/checkpoints/00018... +INFO:tensorflow:global_step/sec: 0.146515 +INFO:tensorflow:step = 7101, loss = 0.866764, precision = 0.8125 (682.524 sec) +INFO:tensorflow:global_step/sec: 0.150315 +INFO:tensorflow:step = 7201, loss = 0.812884, precision = 0.84375 (665.267 sec) +INFO:tensorflow:global_step/sec: 0.156689 +INFO:tensorflow:step = 7301, loss = 0.706213, precision = 0.851562 (638.206 sec) +INFO:tensorflow:global_step/sec: 0.142734 +INFO:tensorflow:step = 7401, loss = 0.688422, precision = 0.882812 (700.603 sec) +Saved checkpoint after 19 epoch(s) to data/resnet164/checkpoints/00019... +INFO:tensorflow:global_step/sec: 0.143477 +INFO:tensorflow:step = 7501, loss = 0.793061, precision = 0.828125 (696.978 sec) +INFO:tensorflow:global_step/sec: 0.152119 +INFO:tensorflow:step = 7601, loss = 0.648185, precision = 0.890625 (657.379 sec) +INFO:tensorflow:global_step/sec: 0.162587 +INFO:tensorflow:step = 7701, loss = 0.62901, precision = 0.890625 (615.057 sec) +INFO:tensorflow:global_step/sec: 0.167021 +INFO:tensorflow:step = 7801, loss = 0.640894, precision = 0.890625 (598.728 sec) +Saved checkpoint after 20 epoch(s) to data/resnet164/checkpoints/00020... +INFO:tensorflow:global_step/sec: 0.16676 +INFO:tensorflow:step = 7901, loss = 0.686377, precision = 0.851562 (599.664 sec) +INFO:tensorflow:global_step/sec: 0.146775 +INFO:tensorflow:step = 8001, loss = 0.769182, precision = 0.828125 (681.313 sec) +INFO:tensorflow:global_step/sec: 0.154138 +INFO:tensorflow:step = 8101, loss = 0.799919, precision = 0.804688 (648.769 sec) +INFO:tensorflow:global_step/sec: 0.156656 +INFO:tensorflow:step = 8201, loss = 0.70916, precision = 0.875 (638.341 sec) +Saved checkpoint after 21 epoch(s) to data/resnet164/checkpoints/00021... +INFO:tensorflow:global_step/sec: 0.162793 +INFO:tensorflow:step = 8301, loss = 0.703719, precision = 0.84375 (614.277 sec) +INFO:tensorflow:global_step/sec: 0.156607 +INFO:tensorflow:step = 8401, loss = 0.738867, precision = 0.875 (638.542 sec) +INFO:tensorflow:global_step/sec: 0.151513 +INFO:tensorflow:step = 8501, loss = 0.728686, precision = 0.859375 (660.009 sec) +INFO:tensorflow:global_step/sec: 0.151091 +INFO:tensorflow:step = 8601, loss = 0.695529, precision = 0.859375 (661.855 sec) +Saved checkpoint after 22 epoch(s) to data/resnet164/checkpoints/00022... +INFO:tensorflow:global_step/sec: 0.149399 +INFO:tensorflow:step = 8701, loss = 0.598267, precision = 0.90625 (669.350 sec) +INFO:tensorflow:global_step/sec: 0.152777 +INFO:tensorflow:step = 8801, loss = 0.729942, precision = 0.835938 (654.548 sec) +INFO:tensorflow:global_step/sec: 0.152511 +INFO:tensorflow:step = 8901, loss = 0.681867, precision = 0.890625 (655.690 sec) +Saved checkpoint after 23 epoch(s) to data/resnet164/checkpoints/00023... +INFO:tensorflow:global_step/sec: 0.14981 +INFO:tensorflow:step = 9001, loss = 0.688149, precision = 0.875 (667.513 sec) +INFO:tensorflow:global_step/sec: 0.149567 +INFO:tensorflow:step = 9101, loss = 0.673693, precision = 0.875 (668.597 sec) +INFO:tensorflow:global_step/sec: 0.154981 +INFO:tensorflow:step = 9201, loss = 0.631535, precision = 0.898438 (645.240 sec) +INFO:tensorflow:global_step/sec: 0.149799 +INFO:tensorflow:step = 9301, loss = 0.6215, precision = 0.882812 (667.559 sec) +Saved checkpoint after 24 epoch(s) to data/resnet164/checkpoints/00024... +INFO:tensorflow:global_step/sec: 0.150617 +INFO:tensorflow:step = 9401, loss = 0.685866, precision = 0.882812 (663.935 sec) +INFO:tensorflow:global_step/sec: 0.151154 +INFO:tensorflow:step = 9501, loss = 0.689527, precision = 0.890625 (661.576 sec) +INFO:tensorflow:global_step/sec: 0.147062 +INFO:tensorflow:step = 9601, loss = 0.70705, precision = 0.867188 (679.984 sec) +INFO:tensorflow:global_step/sec: 0.14425 +INFO:tensorflow:step = 9701, loss = 0.742276, precision = 0.851562 (693.242 sec) +Saved checkpoint after 25 epoch(s) to data/resnet164/checkpoints/00025... +INFO:tensorflow:global_step/sec: 0.144876 +INFO:tensorflow:step = 9801, loss = 0.916183, precision = 0.796875 (690.246 sec) +INFO:tensorflow:global_step/sec: 0.113616 +INFO:tensorflow:step = 9901, loss = 0.647125, precision = 0.890625 (880.157 sec) +INFO:tensorflow:global_step/sec: 0.117231 +INFO:tensorflow:step = 10001, loss = 0.736793, precision = 0.867188 (853.014 sec) +INFO:tensorflow:global_step/sec: 0.13448 +INFO:tensorflow:step = 10101, loss = 0.638379, precision = 0.859375 (743.606 sec) +Saved checkpoint after 26 epoch(s) to data/resnet164/checkpoints/00026... +INFO:tensorflow:global_step/sec: 0.13944 +INFO:tensorflow:step = 10201, loss = 0.746801, precision = 0.882812 (717.156 sec) +INFO:tensorflow:global_step/sec: 0.139451 +INFO:tensorflow:step = 10301, loss = 0.696116, precision = 0.851562 (717.097 sec) +INFO:tensorflow:global_step/sec: 0.148535 +INFO:tensorflow:step = 10401, loss = 0.539461, precision = 0.921875 (673.241 sec) +INFO:tensorflow:global_step/sec: 0.152032 +INFO:tensorflow:step = 10501, loss = 0.715655, precision = 0.859375 (657.758 sec) +Saved checkpoint after 27 epoch(s) to data/resnet164/checkpoints/00027... +INFO:tensorflow:global_step/sec: 0.145291 +INFO:tensorflow:step = 10601, loss = 0.629109, precision = 0.882812 (688.276 sec) +INFO:tensorflow:global_step/sec: 0.148361 +INFO:tensorflow:step = 10701, loss = 0.710225, precision = 0.867188 (674.032 sec) +INFO:tensorflow:global_step/sec: 0.143323 +INFO:tensorflow:step = 10801, loss = 0.775453, precision = 0.851562 (697.726 sec) +INFO:tensorflow:global_step/sec: 0.125356 +INFO:tensorflow:step = 10901, loss = 0.734063, precision = 0.859375 (797.729 sec) +Saved checkpoint after 28 epoch(s) to data/resnet164/checkpoints/00028... +INFO:tensorflow:global_step/sec: 0.143789 +INFO:tensorflow:step = 11001, loss = 0.824667, precision = 0.8125 (695.461 sec) +INFO:tensorflow:global_step/sec: 0.14376 +INFO:tensorflow:step = 11101, loss = 0.783941, precision = 0.84375 (695.605 sec) +INFO:tensorflow:global_step/sec: 0.150107 +INFO:tensorflow:step = 11201, loss = 0.67852, precision = 0.84375 (666.193 sec) +INFO:tensorflow:global_step/sec: 0.148117 +INFO:tensorflow:step = 11301, loss = 0.590707, precision = 0.898438 (675.141 sec) +Saved checkpoint after 29 epoch(s) to data/resnet164/checkpoints/00029... +INFO:tensorflow:global_step/sec: 0.150808 +INFO:tensorflow:step = 11401, loss = 0.648397, precision = 0.859375 (663.096 sec) +INFO:tensorflow:global_step/sec: 0.149868 +INFO:tensorflow:step = 11501, loss = 0.690367, precision = 0.875 (667.256 sec) +INFO:tensorflow:global_step/sec: 0.139239 +INFO:tensorflow:step = 11601, loss = 0.660261, precision = 0.882812 (718.189 sec) +INFO:tensorflow:global_step/sec: 0.12857 +INFO:tensorflow:step = 11701, loss = 0.692158, precision = 0.875 (777.789 sec) +Saved checkpoint after 30 epoch(s) to data/resnet164/checkpoints/00030... +INFO:tensorflow:global_step/sec: 0.143345 +INFO:tensorflow:step = 11801, loss = 0.71841, precision = 0.867188 (697.617 sec) +INFO:tensorflow:global_step/sec: 0.145497 +INFO:tensorflow:step = 11901, loss = 0.781778, precision = 0.851562 (687.301 sec) +INFO:tensorflow:global_step/sec: 0.144839 +INFO:tensorflow:step = 12001, loss = 0.757472, precision = 0.84375 (690.420 sec) +INFO:tensorflow:global_step/sec: 0.148287 +INFO:tensorflow:step = 12101, loss = 0.663395, precision = 0.882812 (674.370 sec) +Saved checkpoint after 31 epoch(s) to data/resnet164/checkpoints/00031... +INFO:tensorflow:global_step/sec: 0.145998 +INFO:tensorflow:step = 12201, loss = 0.804775, precision = 0.8125 (684.941 sec) +INFO:tensorflow:global_step/sec: 0.148809 +INFO:tensorflow:step = 12301, loss = 0.65259, precision = 0.929688 (672.004 sec) +INFO:tensorflow:global_step/sec: 0.145823 +INFO:tensorflow:step = 12401, loss = 0.671941, precision = 0.875 (685.765 sec) +INFO:tensorflow:global_step/sec: 0.147315 +INFO:tensorflow:step = 12501, loss = 0.649361, precision = 0.890625 (678.818 sec) +Saved checkpoint after 32 epoch(s) to data/resnet164/checkpoints/00032... +INFO:tensorflow:global_step/sec: 0.147072 +INFO:tensorflow:step = 12601, loss = 0.683354, precision = 0.882812 (679.938 sec) +INFO:tensorflow:global_step/sec: 0.14639 +INFO:tensorflow:step = 12701, loss = 0.724411, precision = 0.875 (683.105 sec) +INFO:tensorflow:global_step/sec: 0.145181 +INFO:tensorflow:step = 12801, loss = 0.897405, precision = 0.796875 (688.793 sec) +INFO:tensorflow:global_step/sec: 0.138773 +INFO:tensorflow:step = 12901, loss = 0.754058, precision = 0.867188 (720.599 sec) +Saved checkpoint after 33 epoch(s) to data/resnet164/checkpoints/00033... +INFO:tensorflow:global_step/sec: 0.139213 +INFO:tensorflow:step = 13001, loss = 0.534893, precision = 0.921875 (718.326 sec) +INFO:tensorflow:global_step/sec: 0.148151 +INFO:tensorflow:step = 13101, loss = 0.834407, precision = 0.820312 (674.985 sec) +INFO:tensorflow:global_step/sec: 0.149767 +INFO:tensorflow:step = 13201, loss = 0.728708, precision = 0.867188 (667.705 sec) +Saved checkpoint after 34 epoch(s) to data/resnet164/checkpoints/00034... +INFO:tensorflow:global_step/sec: 0.142933 +INFO:tensorflow:step = 13301, loss = 0.697871, precision = 0.851562 (699.627 sec) +INFO:tensorflow:global_step/sec: 0.143499 +INFO:tensorflow:step = 13401, loss = 0.614756, precision = 0.898438 (696.871 sec) +INFO:tensorflow:global_step/sec: 0.144316 +INFO:tensorflow:step = 13501, loss = 0.771262, precision = 0.820312 (692.923 sec) +INFO:tensorflow:global_step/sec: 0.143108 +INFO:tensorflow:step = 13601, loss = 0.884527, precision = 0.78125 (698.774 sec) +Saved checkpoint after 35 epoch(s) to data/resnet164/checkpoints/00035... +INFO:tensorflow:global_step/sec: 0.140836 +INFO:tensorflow:step = 13701, loss = 0.78354, precision = 0.859375 (710.048 sec) +INFO:tensorflow:global_step/sec: 0.148864 +INFO:tensorflow:step = 13801, loss = 0.602433, precision = 0.882812 (671.754 sec) +INFO:tensorflow:global_step/sec: 0.148479 +INFO:tensorflow:step = 13901, loss = 0.636735, precision = 0.875 (673.495 sec) +INFO:tensorflow:global_step/sec: 0.1456 +INFO:tensorflow:step = 14001, loss = 0.721753, precision = 0.859375 (686.813 sec) +Saved checkpoint after 36 epoch(s) to data/resnet164/checkpoints/00036... +INFO:tensorflow:global_step/sec: 0.144634 +INFO:tensorflow:step = 14101, loss = 0.687836, precision = 0.867188 (691.403 sec) +INFO:tensorflow:global_step/sec: 0.145598 +INFO:tensorflow:step = 14201, loss = 0.794823, precision = 0.828125 (686.821 sec) +INFO:tensorflow:global_step/sec: 0.144356 +INFO:tensorflow:step = 14301, loss = 0.632425, precision = 0.914062 (692.732 sec) +INFO:tensorflow:global_step/sec: 0.13628 +INFO:tensorflow:step = 14401, loss = 0.682689, precision = 0.875 (733.781 sec) +Saved checkpoint after 37 epoch(s) to data/resnet164/checkpoints/00037... +INFO:tensorflow:global_step/sec: 0.138856 +INFO:tensorflow:step = 14501, loss = 0.701991, precision = 0.851562 (720.171 sec) +INFO:tensorflow:global_step/sec: 0.13803 +INFO:tensorflow:step = 14601, loss = 0.768676, precision = 0.820312 (724.479 sec) +INFO:tensorflow:global_step/sec: 0.136869 +INFO:tensorflow:step = 14701, loss = 0.784695, precision = 0.828125 (730.625 sec) +INFO:tensorflow:global_step/sec: 0.140896 +INFO:tensorflow:step = 14801, loss = 0.652837, precision = 0.882812 (709.744 sec) +Saved checkpoint after 38 epoch(s) to data/resnet164/checkpoints/00038... +INFO:tensorflow:global_step/sec: 0.136942 +INFO:tensorflow:step = 14901, loss = 0.685812, precision = 0.875 (730.236 sec) +INFO:tensorflow:global_step/sec: 0.135022 +INFO:tensorflow:step = 15001, loss = 0.719847, precision = 0.890625 (740.619 sec) +INFO:tensorflow:global_step/sec: 0.134337 +INFO:tensorflow:step = 15101, loss = 0.609947, precision = 0.921875 (744.395 sec) +INFO:tensorflow:global_step/sec: 0.131742 +INFO:tensorflow:step = 15201, loss = 0.632524, precision = 0.921875 (759.057 sec) +Saved checkpoint after 39 epoch(s) to data/resnet164/checkpoints/00039... +INFO:tensorflow:global_step/sec: 0.136164 +INFO:tensorflow:step = 15301, loss = 0.676612, precision = 0.875 (734.409 sec) +INFO:tensorflow:global_step/sec: 0.137691 +INFO:tensorflow:step = 15401, loss = 0.731972, precision = 0.859375 (726.263 sec) +INFO:tensorflow:global_step/sec: 0.14126 +INFO:tensorflow:step = 15501, loss = 0.766407, precision = 0.84375 (707.914 sec) +INFO:tensorflow:global_step/sec: 0.144947 +INFO:tensorflow:step = 15601, loss = 0.736841, precision = 0.859375 (689.906 sec) +Saved checkpoint after 40 epoch(s) to data/resnet164/checkpoints/00040... +INFO:tensorflow:global_step/sec: 0.144888 +INFO:tensorflow:step = 15701, loss = 0.578763, precision = 0.90625 (690.188 sec) +INFO:tensorflow:global_step/sec: 0.146363 +INFO:tensorflow:step = 15801, loss = 0.713392, precision = 0.867188 (683.232 sec) +INFO:tensorflow:global_step/sec: 0.146103 +INFO:tensorflow:step = 15901, loss = 0.69793, precision = 0.84375 (684.448 sec) +INFO:tensorflow:global_step/sec: 0.145813 +INFO:tensorflow:step = 16001, loss = 0.727529, precision = 0.882812 (685.808 sec) +Saved checkpoint after 41 epoch(s) to data/resnet164/checkpoints/00041... +INFO:tensorflow:global_step/sec: 0.143597 +INFO:tensorflow:step = 16101, loss = 0.73857, precision = 0.867188 (696.395 sec) +INFO:tensorflow:global_step/sec: 0.141571 +INFO:tensorflow:step = 16201, loss = 0.532766, precision = 0.898438 (706.361 sec) +INFO:tensorflow:global_step/sec: 0.141905 +INFO:tensorflow:step = 16301, loss = 0.634989, precision = 0.890625 (704.697 sec) +INFO:tensorflow:global_step/sec: 0.123546 +INFO:tensorflow:step = 16401, loss = 0.682492, precision = 0.867188 (809.414 sec) +Saved checkpoint after 42 epoch(s) to data/resnet164/checkpoints/00042... +INFO:tensorflow:global_step/sec: 0.126857 +INFO:tensorflow:step = 16501, loss = 0.887843, precision = 0.78125 (788.288 sec) +INFO:tensorflow:global_step/sec: 0.143543 +INFO:tensorflow:step = 16601, loss = 0.792445, precision = 0.835938 (696.655 sec) +INFO:tensorflow:global_step/sec: 0.139219 +INFO:tensorflow:step = 16701, loss = 0.6321, precision = 0.914062 (718.295 sec) +INFO:tensorflow:global_step/sec: 0.141022 +INFO:tensorflow:step = 16801, loss = 0.611746, precision = 0.90625 (709.109 sec) +Saved checkpoint after 43 epoch(s) to data/resnet164/checkpoints/00043... +INFO:tensorflow:global_step/sec: 0.139375 +INFO:tensorflow:step = 16901, loss = 0.76093, precision = 0.859375 (717.491 sec) +INFO:tensorflow:global_step/sec: 0.142694 +INFO:tensorflow:step = 17001, loss = 0.781368, precision = 0.828125 (700.802 sec) +INFO:tensorflow:global_step/sec: 0.144379 +INFO:tensorflow:step = 17101, loss = 0.802106, precision = 0.882812 (692.623 sec) +INFO:tensorflow:global_step/sec: 0.143649 +INFO:tensorflow:step = 17201, loss = 0.76721, precision = 0.875 (696.140 sec) +Saved checkpoint after 44 epoch(s) to data/resnet164/checkpoints/00044... +INFO:tensorflow:global_step/sec: 0.147098 +INFO:tensorflow:step = 17301, loss = 0.707524, precision = 0.890625 (679.819 sec) +INFO:tensorflow:global_step/sec: 0.148139 +INFO:tensorflow:step = 17401, loss = 0.727504, precision = 0.859375 (675.043 sec) +INFO:tensorflow:global_step/sec: 0.148055 +INFO:tensorflow:step = 17501, loss = 0.717338, precision = 0.84375 (675.423 sec) +Saved checkpoint after 45 epoch(s) to data/resnet164/checkpoints/00045... +INFO:tensorflow:global_step/sec: 0.146807 +INFO:tensorflow:step = 17601, loss = 0.786196, precision = 0.851562 (681.168 sec) +INFO:tensorflow:global_step/sec: 0.138499 +INFO:tensorflow:step = 17701, loss = 0.71118, precision = 0.875 (722.028 sec) +INFO:tensorflow:global_step/sec: 0.135429 +INFO:tensorflow:step = 17801, loss = 0.572601, precision = 0.921875 (738.395 sec) +INFO:tensorflow:global_step/sec: 0.14428 +INFO:tensorflow:step = 17901, loss = 0.77641, precision = 0.835938 (693.095 sec) +Saved checkpoint after 46 epoch(s) to data/resnet164/checkpoints/00046... +INFO:tensorflow:global_step/sec: 0.152657 +INFO:tensorflow:step = 18001, loss = 0.570994, precision = 0.929688 (655.064 sec) +INFO:tensorflow:global_step/sec: 0.14926 +INFO:tensorflow:step = 18101, loss = 0.705175, precision = 0.84375 (669.974 sec) +INFO:tensorflow:global_step/sec: 0.118424 +INFO:tensorflow:step = 18201, loss = 0.541218, precision = 0.945312 (844.426 sec) +INFO:tensorflow:global_step/sec: 0.0910421 +INFO:tensorflow:step = 18301, loss = 0.822482, precision = 0.820312 (1098.393 sec) +Saved checkpoint after 47 epoch(s) to data/resnet164/checkpoints/00047... +INFO:tensorflow:global_step/sec: 0.096068 +INFO:tensorflow:step = 18401, loss = 0.632981, precision = 0.867188 (1040.930 sec) +INFO:tensorflow:global_step/sec: 0.15579 +INFO:tensorflow:step = 18501, loss = 0.719195, precision = 0.859375 (641.891 sec) +INFO:tensorflow:global_step/sec: 0.155643 +INFO:tensorflow:step = 18601, loss = 0.583066, precision = 0.929688 (642.494 sec) +INFO:tensorflow:global_step/sec: 0.154578 +INFO:tensorflow:step = 18701, loss = 0.751036, precision = 0.859375 (646.924 sec) +Saved checkpoint after 48 epoch(s) to data/resnet164/checkpoints/00048... +INFO:tensorflow:global_step/sec: 0.156502 +INFO:tensorflow:step = 18801, loss = 0.497276, precision = 0.9375 (638.970 sec) +INFO:tensorflow:global_step/sec: 0.159093 +INFO:tensorflow:step = 18901, loss = 0.635723, precision = 0.914062 (628.565 sec) +INFO:tensorflow:global_step/sec: 0.158834 +INFO:tensorflow:step = 19001, loss = 0.652796, precision = 0.890625 (629.590 sec) +INFO:tensorflow:global_step/sec: 0.153952 +INFO:tensorflow:step = 19101, loss = 0.795801, precision = 0.84375 (649.553 sec) +Saved checkpoint after 49 epoch(s) to data/resnet164/checkpoints/00049... +INFO:tensorflow:global_step/sec: 0.126971 +INFO:tensorflow:step = 19201, loss = 0.732796, precision = 0.859375 (787.581 sec) +INFO:tensorflow:global_step/sec: 0.102118 +INFO:tensorflow:step = 19301, loss = 0.660868, precision = 0.898438 (979.256 sec) +INFO:tensorflow:global_step/sec: 0.141378 +INFO:tensorflow:step = 19401, loss = 0.686222, precision = 0.867188 (707.321 sec) +INFO:tensorflow:global_step/sec: 0.135974 +INFO:tensorflow:step = 19501, loss = 0.638899, precision = 0.914062 (735.436 sec) +Saved checkpoint after 50 epoch(s) to data/resnet164/checkpoints/00050... +INFO:tensorflow:global_step/sec: 0.142478 +INFO:tensorflow:step = 19601, loss = 0.752302, precision = 0.867188 (701.865 sec) +INFO:tensorflow:global_step/sec: 0.154394 +INFO:tensorflow:step = 19701, loss = 0.586897, precision = 0.898438 (647.693 sec) +INFO:tensorflow:global_step/sec: 0.145157 +INFO:tensorflow:step = 19801, loss = 0.549641, precision = 0.921875 (688.911 sec) +INFO:tensorflow:global_step/sec: 0.145155 +INFO:tensorflow:step = 19901, loss = 0.778201, precision = 0.84375 (688.918 sec) +Saved checkpoint after 51 epoch(s) to data/resnet164/checkpoints/00051... +INFO:tensorflow:global_step/sec: 0.145188 +INFO:tensorflow:step = 20001, loss = 0.606326, precision = 0.890625 (688.762 sec) +INFO:tensorflow:global_step/sec: 0.149088 +INFO:tensorflow:step = 20101, loss = 0.605446, precision = 0.898438 (670.746 sec) +INFO:tensorflow:global_step/sec: 0.147524 +INFO:tensorflow:step = 20201, loss = 0.608187, precision = 0.875 (677.854 sec) +INFO:tensorflow:global_step/sec: 0.147542 +INFO:tensorflow:step = 20301, loss = 0.704985, precision = 0.867188 (677.773 sec) +Saved checkpoint after 52 epoch(s) to data/resnet164/checkpoints/00052... +INFO:tensorflow:global_step/sec: 0.147566 +INFO:tensorflow:step = 20401, loss = 0.687108, precision = 0.875 (677.665 sec) +INFO:tensorflow:global_step/sec: 0.140353 +INFO:tensorflow:step = 20501, loss = 0.671274, precision = 0.890625 (712.489 sec) +INFO:tensorflow:global_step/sec: 0.136458 +INFO:tensorflow:step = 20601, loss = 0.607266, precision = 0.898438 (732.828 sec) +INFO:tensorflow:global_step/sec: 0.140568 +INFO:tensorflow:step = 20701, loss = 0.611009, precision = 0.90625 (711.399 sec) +Saved checkpoint after 53 epoch(s) to data/resnet164/checkpoints/00053... +INFO:tensorflow:global_step/sec: 0.141228 +INFO:tensorflow:step = 20801, loss = 0.561581, precision = 0.945312 (708.075 sec) +INFO:tensorflow:global_step/sec: 0.149342 +INFO:tensorflow:step = 20901, loss = 0.612865, precision = 0.90625 (669.604 sec) +INFO:tensorflow:global_step/sec: 0.151445 +INFO:tensorflow:step = 21001, loss = 0.620529, precision = 0.898438 (660.307 sec) +INFO:tensorflow:global_step/sec: 0.149323 +INFO:tensorflow:step = 21101, loss = 0.651292, precision = 0.921875 (669.690 sec) +Saved checkpoint after 54 epoch(s) to data/resnet164/checkpoints/00054... +INFO:tensorflow:global_step/sec: 0.148208 +INFO:tensorflow:step = 21201, loss = 0.856856, precision = 0.8125 (674.727 sec) +INFO:tensorflow:global_step/sec: 0.138778 +INFO:tensorflow:step = 21301, loss = 0.659288, precision = 0.890625 (720.576 sec) +INFO:tensorflow:global_step/sec: 0.130736 +INFO:tensorflow:step = 21401, loss = 0.702874, precision = 0.851562 (764.902 sec) +INFO:tensorflow:global_step/sec: 0.133969 +INFO:tensorflow:step = 21501, loss = 0.749378, precision = 0.875 (746.444 sec) +Saved checkpoint after 55 epoch(s) to data/resnet164/checkpoints/00055... +INFO:tensorflow:global_step/sec: 0.134931 +INFO:tensorflow:step = 21601, loss = 0.58047, precision = 0.929688 (741.119 sec) +INFO:tensorflow:global_step/sec: 0.134325 +INFO:tensorflow:step = 21701, loss = 0.648098, precision = 0.898438 (744.465 sec) +INFO:tensorflow:global_step/sec: 0.133599 +INFO:tensorflow:step = 21801, loss = 0.641283, precision = 0.875 (748.511 sec) +Saved checkpoint after 56 epoch(s) to data/resnet164/checkpoints/00056... +INFO:tensorflow:global_step/sec: 0.133988 +INFO:tensorflow:step = 21901, loss = 0.735364, precision = 0.882812 (746.335 sec) +INFO:tensorflow:global_step/sec: 0.13473 +INFO:tensorflow:step = 22001, loss = 0.637471, precision = 0.914062 (742.224 sec) +INFO:tensorflow:global_step/sec: 0.132079 +INFO:tensorflow:step = 22101, loss = 0.694523, precision = 0.882812 (757.124 sec) +INFO:tensorflow:global_step/sec: 0.133439 +INFO:tensorflow:step = 22201, loss = 0.804554, precision = 0.828125 (749.407 sec) +Saved checkpoint after 57 epoch(s) to data/resnet164/checkpoints/00057... +INFO:tensorflow:global_step/sec: 0.129775 +INFO:tensorflow:step = 22301, loss = 0.787446, precision = 0.8125 (770.567 sec) +INFO:tensorflow:global_step/sec: 0.128857 +INFO:tensorflow:step = 22401, loss = 0.602665, precision = 0.929688 (776.055 sec) +INFO:tensorflow:global_step/sec: 0.130309 +INFO:tensorflow:step = 22501, loss = 0.644666, precision = 0.890625 (767.405 sec) +INFO:tensorflow:global_step/sec: 0.131189 +INFO:tensorflow:step = 22601, loss = 0.656975, precision = 0.875 (762.256 sec) +Saved checkpoint after 58 epoch(s) to data/resnet164/checkpoints/00058... +INFO:tensorflow:global_step/sec: 0.130276 +INFO:tensorflow:step = 22701, loss = 0.653692, precision = 0.90625 (767.599 sec) +INFO:tensorflow:global_step/sec: 0.130517 +INFO:tensorflow:step = 22801, loss = 0.632458, precision = 0.898438 (766.186 sec) +INFO:tensorflow:global_step/sec: 0.130927 +INFO:tensorflow:step = 22901, loss = 0.616373, precision = 0.898438 (763.785 sec) +INFO:tensorflow:global_step/sec: 0.133516 +INFO:tensorflow:step = 23001, loss = 0.702975, precision = 0.882812 (748.972 sec) +Saved checkpoint after 59 epoch(s) to data/resnet164/checkpoints/00059... +INFO:tensorflow:global_step/sec: 0.133806 +INFO:tensorflow:step = 23101, loss = 0.660854, precision = 0.890625 (747.351 sec) +INFO:tensorflow:global_step/sec: 0.133774 +INFO:tensorflow:step = 23201, loss = 0.636034, precision = 0.890625 (747.530 sec) +INFO:tensorflow:global_step/sec: 0.134103 +INFO:tensorflow:step = 23301, loss = 0.724265, precision = 0.851562 (745.697 sec) +INFO:tensorflow:global_step/sec: 0.13289 +INFO:tensorflow:step = 23401, loss = 0.657623, precision = 0.882812 (752.504 sec) +Saved checkpoint after 60 epoch(s) to data/resnet164/checkpoints/00060... +INFO:tensorflow:global_step/sec: 0.13291 +INFO:tensorflow:step = 23501, loss = 0.708539, precision = 0.882812 (752.390 sec) +INFO:tensorflow:global_step/sec: 0.132612 +INFO:tensorflow:step = 23601, loss = 0.720198, precision = 0.851562 (754.077 sec) +INFO:tensorflow:global_step/sec: 0.132367 +INFO:tensorflow:step = 23701, loss = 0.731016, precision = 0.875 (755.473 sec) +INFO:tensorflow:global_step/sec: 0.134151 +INFO:tensorflow:step = 23801, loss = 0.776437, precision = 0.828125 (745.427 sec) +Saved checkpoint after 61 epoch(s) to data/resnet164/checkpoints/00061... +INFO:tensorflow:global_step/sec: 0.135625 +INFO:tensorflow:step = 23901, loss = 0.710227, precision = 0.882812 (737.326 sec) +INFO:tensorflow:global_step/sec: 0.134424 +INFO:tensorflow:step = 24001, loss = 0.697301, precision = 0.890625 (743.916 sec) +INFO:tensorflow:global_step/sec: 0.133496 +INFO:tensorflow:step = 24101, loss = 0.706882, precision = 0.898438 (749.088 sec) +INFO:tensorflow:global_step/sec: 0.132918 +INFO:tensorflow:step = 24201, loss = 0.6589, precision = 0.875 (752.345 sec) +Saved checkpoint after 62 epoch(s) to data/resnet164/checkpoints/00062... +INFO:tensorflow:global_step/sec: 0.131881 +INFO:tensorflow:step = 24301, loss = 0.889101, precision = 0.796875 (758.258 sec) +INFO:tensorflow:global_step/sec: 0.134452 +INFO:tensorflow:step = 24401, loss = 0.726608, precision = 0.851562 (743.762 sec) +INFO:tensorflow:global_step/sec: 0.132936 +INFO:tensorflow:step = 24501, loss = 0.596243, precision = 0.914062 (752.242 sec) +INFO:tensorflow:global_step/sec: 0.131915 +INFO:tensorflow:step = 24601, loss = 0.634795, precision = 0.882812 (758.064 sec) +Saved checkpoint after 63 epoch(s) to data/resnet164/checkpoints/00063... +INFO:tensorflow:global_step/sec: 0.131433 +INFO:tensorflow:step = 24701, loss = 0.678893, precision = 0.882812 (760.842 sec) +INFO:tensorflow:global_step/sec: 0.129169 +INFO:tensorflow:step = 24801, loss = 0.638255, precision = 0.867188 (774.178 sec) +INFO:tensorflow:global_step/sec: 0.130902 +INFO:tensorflow:step = 24901, loss = 0.881363, precision = 0.835938 (763.929 sec) +INFO:tensorflow:global_step/sec: 0.129858 +INFO:tensorflow:step = 25001, loss = 0.660767, precision = 0.890625 (770.071 sec) +Saved checkpoint after 64 epoch(s) to data/resnet164/checkpoints/00064... +INFO:tensorflow:global_step/sec: 0.130412 +INFO:tensorflow:step = 25101, loss = 0.580645, precision = 0.9375 (766.800 sec) +INFO:tensorflow:global_step/sec: 0.131933 +INFO:tensorflow:step = 25201, loss = 0.642647, precision = 0.90625 (757.959 sec) +INFO:tensorflow:global_step/sec: 0.131507 +INFO:tensorflow:step = 25301, loss = 0.697281, precision = 0.84375 (760.419 sec) +INFO:tensorflow:global_step/sec: 0.133102 +INFO:tensorflow:step = 25401, loss = 0.60622, precision = 0.90625 (751.303 sec) +Saved checkpoint after 65 epoch(s) to data/resnet164/checkpoints/00065... +INFO:tensorflow:global_step/sec: 0.133981 +INFO:tensorflow:step = 25501, loss = 0.709263, precision = 0.882812 (746.377 sec) +INFO:tensorflow:global_step/sec: 0.13267 +INFO:tensorflow:step = 25601, loss = 0.769578, precision = 0.828125 (753.753 sec) +INFO:tensorflow:global_step/sec: 0.134711 +INFO:tensorflow:step = 25701, loss = 0.695019, precision = 0.898438 (742.329 sec) +INFO:tensorflow:global_step/sec: 0.137389 +INFO:tensorflow:step = 25801, loss = 0.555473, precision = 0.914062 (727.860 sec) +Saved checkpoint after 66 epoch(s) to data/resnet164/checkpoints/00066... +INFO:tensorflow:global_step/sec: 0.136737 +INFO:tensorflow:step = 25901, loss = 0.734909, precision = 0.867188 (731.329 sec) +INFO:tensorflow:global_step/sec: 0.133528 +INFO:tensorflow:step = 26001, loss = 0.563629, precision = 0.898438 (748.908 sec) +INFO:tensorflow:global_step/sec: 0.134169 +INFO:tensorflow:step = 26101, loss = 0.678412, precision = 0.875 (745.331 sec) +Saved checkpoint after 67 epoch(s) to data/resnet164/checkpoints/00067... +INFO:tensorflow:global_step/sec: 0.131341 +INFO:tensorflow:step = 26201, loss = 0.576831, precision = 0.9375 (761.376 sec) +INFO:tensorflow:global_step/sec: 0.13537 +INFO:tensorflow:step = 26301, loss = 0.619746, precision = 0.882812 (738.714 sec) +INFO:tensorflow:global_step/sec: 0.132464 +INFO:tensorflow:step = 26401, loss = 0.675096, precision = 0.882812 (754.922 sec) +INFO:tensorflow:global_step/sec: 0.13024 +INFO:tensorflow:step = 26501, loss = 0.816114, precision = 0.859375 (767.816 sec) +Saved checkpoint after 68 epoch(s) to data/resnet164/checkpoints/00068... +INFO:tensorflow:global_step/sec: 0.130043 +INFO:tensorflow:step = 26601, loss = 0.575397, precision = 0.929688 (768.975 sec) +INFO:tensorflow:global_step/sec: 0.134797 +INFO:tensorflow:step = 26701, loss = 0.629717, precision = 0.90625 (741.857 sec) +INFO:tensorflow:global_step/sec: 0.136486 +INFO:tensorflow:step = 26801, loss = 0.583626, precision = 0.882812 (732.677 sec) +INFO:tensorflow:global_step/sec: 0.130736 +INFO:tensorflow:step = 26901, loss = 0.695886, precision = 0.875 (764.901 sec) +Saved checkpoint after 69 epoch(s) to data/resnet164/checkpoints/00069... +INFO:tensorflow:global_step/sec: 0.13088 +INFO:tensorflow:step = 27001, loss = 0.556448, precision = 0.929688 (764.057 sec) +INFO:tensorflow:global_step/sec: 0.13076 +INFO:tensorflow:step = 27101, loss = 0.720241, precision = 0.882812 (764.758 sec) +INFO:tensorflow:global_step/sec: 0.132232 +INFO:tensorflow:step = 27201, loss = 0.677291, precision = 0.867188 (756.248 sec) +INFO:tensorflow:global_step/sec: 0.133149 +INFO:tensorflow:step = 27301, loss = 0.717456, precision = 0.851562 (751.037 sec) +Saved checkpoint after 70 epoch(s) to data/resnet164/checkpoints/00070... +INFO:tensorflow:global_step/sec: 0.141768 +INFO:tensorflow:step = 27401, loss = 0.620971, precision = 0.90625 (705.378 sec) +INFO:tensorflow:global_step/sec: 0.148196 +INFO:tensorflow:step = 27501, loss = 0.709948, precision = 0.859375 (674.780 sec) +INFO:tensorflow:global_step/sec: 0.151861 +INFO:tensorflow:step = 27601, loss = 0.686694, precision = 0.867188 (658.497 sec) +INFO:tensorflow:global_step/sec: 0.151055 +INFO:tensorflow:step = 27701, loss = 0.687338, precision = 0.898438 (662.010 sec) +Saved checkpoint after 71 epoch(s) to data/resnet164/checkpoints/00071... +INFO:tensorflow:global_step/sec: 0.149056 +INFO:tensorflow:step = 27801, loss = 0.727965, precision = 0.8125 (670.889 sec) +INFO:tensorflow:global_step/sec: 0.144143 +INFO:tensorflow:step = 27901, loss = 0.575043, precision = 0.9375 (693.755 sec) +INFO:tensorflow:global_step/sec: 0.137142 +INFO:tensorflow:step = 28001, loss = 0.632586, precision = 0.898438 (729.172 sec) +INFO:tensorflow:global_step/sec: 0.138825 +INFO:tensorflow:step = 28101, loss = 0.682613, precision = 0.890625 (720.332 sec) +Saved checkpoint after 72 epoch(s) to data/resnet164/checkpoints/00072... +INFO:tensorflow:global_step/sec: 0.121898 +INFO:tensorflow:step = 28201, loss = 0.649817, precision = 0.890625 (820.360 sec) +INFO:tensorflow:global_step/sec: 0.121919 +INFO:tensorflow:step = 28301, loss = 0.586312, precision = 0.929688 (820.216 sec) +INFO:tensorflow:global_step/sec: 0.147797 +INFO:tensorflow:step = 28401, loss = 0.634121, precision = 0.882812 (676.602 sec) +INFO:tensorflow:global_step/sec: 0.14744 +INFO:tensorflow:step = 28501, loss = 0.663141, precision = 0.882812 (678.242 sec) +Saved checkpoint after 73 epoch(s) to data/resnet164/checkpoints/00073... +INFO:tensorflow:global_step/sec: 0.146062 +INFO:tensorflow:step = 28601, loss = 0.664759, precision = 0.890625 (684.642 sec) +INFO:tensorflow:global_step/sec: 0.143405 +INFO:tensorflow:step = 28701, loss = 0.578263, precision = 0.9375 (697.327 sec) +INFO:tensorflow:global_step/sec: 0.1204 +INFO:tensorflow:step = 28801, loss = 0.643198, precision = 0.90625 (830.561 sec) +INFO:tensorflow:global_step/sec: 0.148461 +INFO:tensorflow:step = 28901, loss = 0.576857, precision = 0.953125 (673.579 sec) +Saved checkpoint after 74 epoch(s) to data/resnet164/checkpoints/00074... +INFO:tensorflow:global_step/sec: 0.151377 +INFO:tensorflow:step = 29001, loss = 0.592912, precision = 0.921875 (660.600 sec) +INFO:tensorflow:global_step/sec: 0.149237 +INFO:tensorflow:step = 29101, loss = 0.769734, precision = 0.835938 (670.076 sec) +INFO:tensorflow:global_step/sec: 0.131792 +INFO:tensorflow:step = 29201, loss = 0.769668, precision = 0.820312 (758.773 sec) +INFO:tensorflow:global_step/sec: 0.11856 +INFO:tensorflow:step = 29301, loss = 0.627649, precision = 0.882812 (843.452 sec) +Saved checkpoint after 75 epoch(s) to data/resnet164/checkpoints/00075... +INFO:tensorflow:global_step/sec: 0.128535 +INFO:tensorflow:step = 29401, loss = 0.627231, precision = 0.914062 (777.998 sec) +INFO:tensorflow:global_step/sec: 0.148282 +INFO:tensorflow:step = 29501, loss = 0.707525, precision = 0.859375 (674.391 sec) +INFO:tensorflow:global_step/sec: 0.145073 +INFO:tensorflow:step = 29601, loss = 0.567499, precision = 0.929688 (689.310 sec) +INFO:tensorflow:global_step/sec: 0.147486 +INFO:tensorflow:step = 29701, loss = 0.620218, precision = 0.859375 (678.029 sec) +Saved checkpoint after 76 epoch(s) to data/resnet164/checkpoints/00076... +INFO:tensorflow:global_step/sec: 0.150782 +INFO:tensorflow:step = 29801, loss = 0.764623, precision = 0.867188 (663.209 sec) +INFO:tensorflow:global_step/sec: 0.152117 +INFO:tensorflow:step = 29901, loss = 0.640176, precision = 0.898438 (657.388 sec) +INFO:tensorflow:global_step/sec: 0.144865 +INFO:tensorflow:step = 30001, loss = 0.692015, precision = 0.882812 (690.298 sec) +INFO:tensorflow:global_step/sec: 0.143284 +INFO:tensorflow:step = 30101, loss = 0.666982, precision = 0.890625 (697.914 sec) +Saved checkpoint after 77 epoch(s) to data/resnet164/checkpoints/00077... +INFO:tensorflow:global_step/sec: 0.144943 +INFO:tensorflow:step = 30201, loss = 0.626937, precision = 0.9375 (689.928 sec) +INFO:tensorflow:global_step/sec: 0.143512 +INFO:tensorflow:step = 30301, loss = 0.776291, precision = 0.851562 (696.807 sec) +INFO:tensorflow:global_step/sec: 0.147518 +INFO:tensorflow:step = 30401, loss = 0.729203, precision = 0.84375 (677.883 sec) +Saved checkpoint after 78 epoch(s) to data/resnet164/checkpoints/00078... +INFO:tensorflow:global_step/sec: 0.145685 +INFO:tensorflow:step = 30501, loss = 0.587819, precision = 0.890625 (686.411 sec) +INFO:tensorflow:global_step/sec: 0.147806 +INFO:tensorflow:step = 30601, loss = 0.667186, precision = 0.921875 (676.564 sec) +INFO:tensorflow:global_step/sec: 0.1473 +INFO:tensorflow:step = 30701, loss = 0.674999, precision = 0.867188 (678.885 sec) +INFO:tensorflow:global_step/sec: 0.145405 +INFO:tensorflow:step = 30801, loss = 0.723376, precision = 0.859375 (687.732 sec) +Saved checkpoint after 79 epoch(s) to data/resnet164/checkpoints/00079... +INFO:tensorflow:global_step/sec: 0.141209 +INFO:tensorflow:step = 30901, loss = 0.687747, precision = 0.867188 (708.168 sec) +INFO:tensorflow:global_step/sec: 0.150642 +INFO:tensorflow:step = 31001, loss = 0.701329, precision = 0.875 (663.827 sec) +INFO:tensorflow:global_step/sec: 0.150052 +INFO:tensorflow:step = 31101, loss = 0.701779, precision = 0.875 (666.434 sec) +INFO:tensorflow:global_step/sec: 0.143307 +INFO:tensorflow:step = 31201, loss = 0.605538, precision = 0.914062 (697.801 sec) +Saved checkpoint after 80 epoch(s) to data/resnet164/checkpoints/00080... +INFO:tensorflow:global_step/sec: 0.141389 +INFO:tensorflow:step = 31301, loss = 0.630363, precision = 0.90625 (707.266 sec) +INFO:tensorflow:global_step/sec: 0.145063 +INFO:tensorflow:step = 31401, loss = 0.656484, precision = 0.875 (689.353 sec) +INFO:tensorflow:global_step/sec: 0.13254 +INFO:tensorflow:step = 31501, loss = 0.739198, precision = 0.867188 (754.491 sec) +INFO:tensorflow:global_step/sec: 0.150469 +INFO:tensorflow:step = 31601, loss = 0.670097, precision = 0.890625 (664.587 sec) +Saved checkpoint after 81 epoch(s) to data/resnet164/checkpoints/00081... +INFO:tensorflow:global_step/sec: 0.130054 +INFO:tensorflow:step = 31701, loss = 0.631923, precision = 0.890625 (768.911 sec) +INFO:tensorflow:global_step/sec: 0.152692 +INFO:tensorflow:step = 31801, loss = 0.608848, precision = 0.929688 (654.914 sec) +INFO:tensorflow:global_step/sec: 0.147098 +INFO:tensorflow:step = 31901, loss = 0.662505, precision = 0.890625 (679.821 sec) +INFO:tensorflow:global_step/sec: 0.146728 +INFO:tensorflow:step = 32001, loss = 0.634212, precision = 0.90625 (681.535 sec) +Saved checkpoint after 82 epoch(s) to data/resnet164/checkpoints/00082... +INFO:tensorflow:global_step/sec: 0.146844 +INFO:tensorflow:step = 32101, loss = 0.69414, precision = 0.890625 (680.997 sec) +INFO:tensorflow:global_step/sec: 0.149511 +INFO:tensorflow:step = 32201, loss = 0.673262, precision = 0.882812 (668.847 sec) +INFO:tensorflow:global_step/sec: 0.142698 +INFO:tensorflow:step = 32301, loss = 0.58684, precision = 0.898438 (700.781 sec) +INFO:tensorflow:global_step/sec: 0.140716 +INFO:tensorflow:step = 32401, loss = 0.646469, precision = 0.882812 (710.652 sec) +Saved checkpoint after 83 epoch(s) to data/resnet164/checkpoints/00083... +INFO:tensorflow:global_step/sec: 0.148644 +INFO:tensorflow:step = 32501, loss = 0.684458, precision = 0.84375 (672.749 sec) +INFO:tensorflow:global_step/sec: 0.147171 +INFO:tensorflow:step = 32601, loss = 0.68817, precision = 0.90625 (679.480 sec) +INFO:tensorflow:global_step/sec: 0.147622 +INFO:tensorflow:step = 32701, loss = 0.65056, precision = 0.898438 (677.406 sec) +INFO:tensorflow:global_step/sec: 0.154824 +INFO:tensorflow:step = 32801, loss = 0.814233, precision = 0.859375 (645.896 sec) +Saved checkpoint after 84 epoch(s) to data/resnet164/checkpoints/00084... +INFO:tensorflow:global_step/sec: 0.145739 +INFO:tensorflow:step = 32901, loss = 0.605846, precision = 0.921875 (686.159 sec) +INFO:tensorflow:global_step/sec: 0.142888 +INFO:tensorflow:step = 33001, loss = 0.677476, precision = 0.867188 (699.847 sec) +INFO:tensorflow:global_step/sec: 0.145765 +INFO:tensorflow:step = 33101, loss = 0.737437, precision = 0.867188 (686.036 sec) +INFO:tensorflow:global_step/sec: 0.142957 +INFO:tensorflow:step = 33201, loss = 0.655267, precision = 0.914062 (699.512 sec) +Saved checkpoint after 85 epoch(s) to data/resnet164/checkpoints/00085... +INFO:tensorflow:global_step/sec: 0.150271 +INFO:tensorflow:step = 33301, loss = 0.640084, precision = 0.898438 (665.466 sec) +INFO:tensorflow:global_step/sec: 0.144345 +INFO:tensorflow:step = 33401, loss = 0.676603, precision = 0.867188 (692.783 sec) +INFO:tensorflow:global_step/sec: 0.149159 +INFO:tensorflow:step = 33501, loss = 0.63091, precision = 0.90625 (670.424 sec) +INFO:tensorflow:global_step/sec: 0.145139 +INFO:tensorflow:step = 33601, loss = 0.664038, precision = 0.851562 (688.996 sec) +Saved checkpoint after 86 epoch(s) to data/resnet164/checkpoints/00086... +INFO:tensorflow:global_step/sec: 0.140157 +INFO:tensorflow:step = 33701, loss = 0.596853, precision = 0.9375 (713.484 sec) +INFO:tensorflow:global_step/sec: 0.135461 +INFO:tensorflow:step = 33801, loss = 0.62646, precision = 0.90625 (738.220 sec) +INFO:tensorflow:global_step/sec: 0.141122 +INFO:tensorflow:step = 33901, loss = 0.95676, precision = 0.789062 (708.608 sec) +INFO:tensorflow:global_step/sec: 0.14148 +INFO:tensorflow:step = 34001, loss = 0.727888, precision = 0.84375 (706.813 sec) +Saved checkpoint after 87 epoch(s) to data/resnet164/checkpoints/00087... +INFO:tensorflow:global_step/sec: 0.144128 +INFO:tensorflow:step = 34101, loss = 0.892649, precision = 0.859375 (693.829 sec) +INFO:tensorflow:global_step/sec: 0.144173 +INFO:tensorflow:step = 34201, loss = 0.649179, precision = 0.882812 (693.612 sec) +INFO:tensorflow:global_step/sec: 0.144489 +INFO:tensorflow:step = 34301, loss = 0.678938, precision = 0.90625 (692.094 sec) +INFO:tensorflow:global_step/sec: 0.145478 +INFO:tensorflow:step = 34401, loss = 0.551747, precision = 0.921875 (687.390 sec) +Saved checkpoint after 88 epoch(s) to data/resnet164/checkpoints/00088... +INFO:tensorflow:global_step/sec: 0.141688 +INFO:tensorflow:step = 34501, loss = 0.637089, precision = 0.898438 (705.774 sec) +INFO:tensorflow:global_step/sec: 0.145926 +INFO:tensorflow:step = 34601, loss = 0.670813, precision = 0.890625 (685.281 sec) +INFO:tensorflow:global_step/sec: 0.144173 +INFO:tensorflow:step = 34701, loss = 0.707091, precision = 0.882812 (693.609 sec) +Saved checkpoint after 89 epoch(s) to data/resnet164/checkpoints/00089... +INFO:tensorflow:global_step/sec: 0.143928 +INFO:tensorflow:step = 34801, loss = 0.657548, precision = 0.851562 (694.792 sec) +INFO:tensorflow:global_step/sec: 0.138412 +INFO:tensorflow:step = 34901, loss = 0.660363, precision = 0.890625 (722.483 sec) +INFO:tensorflow:global_step/sec: 0.141574 +INFO:tensorflow:step = 35001, loss = 0.582893, precision = 0.929688 (706.345 sec) +INFO:tensorflow:global_step/sec: 0.148466 +INFO:tensorflow:step = 35101, loss = 0.729197, precision = 0.867188 (673.554 sec) +Saved checkpoint after 90 epoch(s) to data/resnet164/checkpoints/00090... +INFO:tensorflow:global_step/sec: 0.145851 +INFO:tensorflow:step = 35201, loss = 0.627183, precision = 0.914062 (685.633 sec) +INFO:tensorflow:global_step/sec: 0.144716 +INFO:tensorflow:step = 35301, loss = 0.598298, precision = 0.914062 (691.009 sec) +INFO:tensorflow:global_step/sec: 0.143274 +INFO:tensorflow:step = 35401, loss = 0.67834, precision = 0.867188 (697.965 sec) +INFO:tensorflow:global_step/sec: 0.144054 +INFO:tensorflow:step = 35501, loss = 0.674444, precision = 0.867188 (694.185 sec) +Saved checkpoint after 91 epoch(s) to data/resnet164/checkpoints/00091... +INFO:tensorflow:global_step/sec: 0.142595 +INFO:tensorflow:step = 35601, loss = 0.505515, precision = 0.929688 (701.285 sec) +INFO:tensorflow:global_step/sec: 0.143601 +INFO:tensorflow:step = 35701, loss = 0.537581, precision = 0.929688 (696.372 sec) +INFO:tensorflow:global_step/sec: 0.137299 +INFO:tensorflow:step = 35801, loss = 0.496464, precision = 0.945312 (728.336 sec) +INFO:tensorflow:global_step/sec: 0.148991 +INFO:tensorflow:step = 35901, loss = 0.422166, precision = 0.976562 (671.182 sec) +Saved checkpoint after 92 epoch(s) to data/resnet164/checkpoints/00092... +INFO:tensorflow:global_step/sec: 0.140622 +INFO:tensorflow:step = 36001, loss = 0.432214, precision = 0.960938 (711.128 sec) +INFO:tensorflow:global_step/sec: 0.143395 +INFO:tensorflow:step = 36101, loss = 0.480946, precision = 0.9375 (697.375 sec) +INFO:tensorflow:global_step/sec: 0.138091 +INFO:tensorflow:step = 36201, loss = 0.414451, precision = 0.976562 (724.162 sec) +INFO:tensorflow:global_step/sec: 0.140919 +INFO:tensorflow:step = 36301, loss = 0.470744, precision = 0.953125 (709.627 sec) +Saved checkpoint after 93 epoch(s) to data/resnet164/checkpoints/00093... +INFO:tensorflow:global_step/sec: 0.146695 +INFO:tensorflow:step = 36401, loss = 0.430229, precision = 0.960938 (681.686 sec) +INFO:tensorflow:global_step/sec: 0.134561 +INFO:tensorflow:step = 36501, loss = 0.459159, precision = 0.945312 (743.156 sec) +INFO:tensorflow:global_step/sec: 0.142229 +INFO:tensorflow:step = 36601, loss = 0.422979, precision = 0.960938 (703.089 sec) +INFO:tensorflow:global_step/sec: 0.150019 +INFO:tensorflow:step = 36701, loss = 0.517502, precision = 0.914062 (666.581 sec) +Saved checkpoint after 94 epoch(s) to data/resnet164/checkpoints/00094... +INFO:tensorflow:global_step/sec: 0.143245 +INFO:tensorflow:step = 36801, loss = 0.401117, precision = 0.960938 (698.107 sec) +INFO:tensorflow:global_step/sec: 0.132302 +INFO:tensorflow:step = 36901, loss = 0.408784, precision = 0.96875 (755.847 sec) +INFO:tensorflow:global_step/sec: 0.154901 +INFO:tensorflow:step = 37001, loss = 0.392673, precision = 0.96875 (645.576 sec) +INFO:tensorflow:global_step/sec: 0.155767 +INFO:tensorflow:step = 37101, loss = 0.373222, precision = 0.960938 (641.985 sec) +Saved checkpoint after 95 epoch(s) to data/resnet164/checkpoints/00095... +INFO:tensorflow:global_step/sec: 0.157321 +INFO:tensorflow:step = 37201, loss = 0.404901, precision = 0.96875 (635.643 sec) +INFO:tensorflow:global_step/sec: 0.155863 +INFO:tensorflow:step = 37301, loss = 0.357266, precision = 0.960938 (641.590 sec) +INFO:tensorflow:global_step/sec: 0.15964 +INFO:tensorflow:step = 37401, loss = 0.408385, precision = 0.960938 (626.408 sec) +INFO:tensorflow:global_step/sec: 0.160972 +INFO:tensorflow:step = 37501, loss = 0.376795, precision = 0.96875 (621.226 sec) +Saved checkpoint after 96 epoch(s) to data/resnet164/checkpoints/00096... +INFO:tensorflow:global_step/sec: 0.157306 +INFO:tensorflow:step = 37601, loss = 0.375515, precision = 0.96875 (635.706 sec) +INFO:tensorflow:global_step/sec: 0.155944 +INFO:tensorflow:step = 37701, loss = 0.378208, precision = 0.945312 (641.255 sec) +INFO:tensorflow:global_step/sec: 0.157768 +INFO:tensorflow:step = 37801, loss = 0.346149, precision = 0.984375 (633.843 sec) +INFO:tensorflow:global_step/sec: 0.157367 +INFO:tensorflow:step = 37901, loss = 0.364548, precision = 0.976562 (635.459 sec) +Saved checkpoint after 97 epoch(s) to data/resnet164/checkpoints/00097... +INFO:tensorflow:global_step/sec: 0.155929 +INFO:tensorflow:step = 38001, loss = 0.393942, precision = 0.953125 (641.317 sec) +INFO:tensorflow:global_step/sec: 0.150581 +INFO:tensorflow:step = 38101, loss = 0.327269, precision = 0.976562 (664.096 sec) +INFO:tensorflow:global_step/sec: 0.152237 +INFO:tensorflow:step = 38201, loss = 0.37508, precision = 0.953125 (656.869 sec) +INFO:tensorflow:global_step/sec: 0.160916 +INFO:tensorflow:step = 38301, loss = 0.333509, precision = 0.976562 (621.443 sec) +Saved checkpoint after 98 epoch(s) to data/resnet164/checkpoints/00098... +INFO:tensorflow:global_step/sec: 0.158194 +INFO:tensorflow:step = 38401, loss = 0.39377, precision = 0.960938 (632.135 sec) +INFO:tensorflow:global_step/sec: 0.154953 +INFO:tensorflow:step = 38501, loss = 0.335084, precision = 0.960938 (645.357 sec) +INFO:tensorflow:global_step/sec: 0.156459 +INFO:tensorflow:step = 38601, loss = 0.31044, precision = 0.992188 (639.145 sec) +INFO:tensorflow:global_step/sec: 0.157642 +INFO:tensorflow:step = 38701, loss = 0.360965, precision = 0.984375 (634.350 sec) +Saved checkpoint after 99 epoch(s) to data/resnet164/checkpoints/00099... +INFO:tensorflow:global_step/sec: 0.157629 +INFO:tensorflow:step = 38801, loss = 0.355025, precision = 0.976562 (634.402 sec) +INFO:tensorflow:global_step/sec: 0.160774 +INFO:tensorflow:step = 38901, loss = 0.324815, precision = 0.976562 (621.992 sec) +INFO:tensorflow:global_step/sec: 0.159275 +INFO:tensorflow:step = 39001, loss = 0.399329, precision = 0.945312 (627.845 sec) +Saved checkpoint after 100 epoch(s) to data/resnet164/checkpoints/00100... +INFO:tensorflow:global_step/sec: 0.157322 +INFO:tensorflow:step = 39101, loss = 0.342492, precision = 0.960938 (635.641 sec) +INFO:tensorflow:global_step/sec: 0.149086 +INFO:tensorflow:step = 39201, loss = 0.324027, precision = 0.96875 (670.756 sec) +INFO:tensorflow:global_step/sec: 0.155071 +INFO:tensorflow:step = 39301, loss = 0.2857, precision = 1.0 (644.867 sec) +INFO:tensorflow:global_step/sec: 0.15366 +INFO:tensorflow:step = 39401, loss = 0.386251, precision = 0.953125 (650.785 sec) +Saved checkpoint after 101 epoch(s) to data/resnet164/checkpoints/00101... +INFO:tensorflow:global_step/sec: 0.14611 +INFO:tensorflow:step = 39501, loss = 0.287093, precision = 0.992188 (684.418 sec) +INFO:tensorflow:global_step/sec: 0.155256 +INFO:tensorflow:step = 39601, loss = 0.290623, precision = 0.984375 (644.098 sec) +INFO:tensorflow:global_step/sec: 0.157681 +INFO:tensorflow:step = 39701, loss = 0.31866, precision = 0.960938 (634.193 sec) +INFO:tensorflow:global_step/sec: 0.157094 +INFO:tensorflow:step = 39801, loss = 0.385159, precision = 0.953125 (636.563 sec) +Saved checkpoint after 102 epoch(s) to data/resnet164/checkpoints/00102... +INFO:tensorflow:global_step/sec: 0.158831 +INFO:tensorflow:step = 39901, loss = 0.282834, precision = 0.984375 (629.602 sec) +INFO:tensorflow:global_step/sec: 0.161727 +INFO:tensorflow:step = 40001, loss = 0.359623, precision = 0.953125 (618.324 sec) +INFO:tensorflow:global_step/sec: 0.162763 +INFO:tensorflow:step = 40101, loss = 0.303239, precision = 0.96875 (614.391 sec) +INFO:tensorflow:global_step/sec: 0.163627 +INFO:tensorflow:step = 40201, loss = 0.304932, precision = 0.976562 (611.146 sec) +Saved checkpoint after 103 epoch(s) to data/resnet164/checkpoints/00103... +INFO:tensorflow:global_step/sec: 0.158174 +INFO:tensorflow:step = 40301, loss = 0.278588, precision = 0.984375 (632.214 sec) +INFO:tensorflow:global_step/sec: 0.161473 +INFO:tensorflow:step = 40401, loss = 0.279425, precision = 0.976562 (619.297 sec) +INFO:tensorflow:global_step/sec: 0.16069 +INFO:tensorflow:step = 40501, loss = 0.334199, precision = 0.960938 (622.317 sec) +INFO:tensorflow:global_step/sec: 0.16118 +INFO:tensorflow:step = 40601, loss = 0.312336, precision = 0.976562 (620.423 sec) +Saved checkpoint after 104 epoch(s) to data/resnet164/checkpoints/00104... +INFO:tensorflow:global_step/sec: 0.162931 +INFO:tensorflow:step = 40701, loss = 0.363591, precision = 0.953125 (613.758 sec) +INFO:tensorflow:global_step/sec: 0.164294 +INFO:tensorflow:step = 40801, loss = 0.324772, precision = 0.953125 (608.666 sec) +INFO:tensorflow:global_step/sec: 0.163596 +INFO:tensorflow:step = 40901, loss = 0.263847, precision = 0.984375 (611.260 sec) +INFO:tensorflow:global_step/sec: 0.164072 +INFO:tensorflow:step = 41001, loss = 0.33258, precision = 0.960938 (609.488 sec) +Saved checkpoint after 105 epoch(s) to data/resnet164/checkpoints/00105... +INFO:tensorflow:global_step/sec: 0.164102 +INFO:tensorflow:step = 41101, loss = 0.246443, precision = 0.992188 (609.377 sec) +INFO:tensorflow:global_step/sec: 0.164653 +INFO:tensorflow:step = 41201, loss = 0.248174, precision = 1.0 (607.338 sec) +INFO:tensorflow:global_step/sec: 0.164906 +INFO:tensorflow:step = 41301, loss = 0.280519, precision = 0.984375 (606.407 sec) +INFO:tensorflow:global_step/sec: 0.165702 +INFO:tensorflow:step = 41401, loss = 0.271288, precision = 0.976562 (603.493 sec) +Saved checkpoint after 106 epoch(s) to data/resnet164/checkpoints/00106... +INFO:tensorflow:global_step/sec: 0.164278 +INFO:tensorflow:step = 41501, loss = 0.302357, precision = 0.96875 (608.723 sec) +INFO:tensorflow:global_step/sec: 0.162905 +INFO:tensorflow:step = 41601, loss = 0.249591, precision = 0.992188 (613.853 sec) +INFO:tensorflow:global_step/sec: 0.16022 +INFO:tensorflow:step = 41701, loss = 0.231301, precision = 1.0 (624.141 sec) +INFO:tensorflow:global_step/sec: 0.161671 +INFO:tensorflow:step = 41801, loss = 0.249345, precision = 0.976562 (618.539 sec) +Saved checkpoint after 107 epoch(s) to data/resnet164/checkpoints/00107... +INFO:tensorflow:global_step/sec: 0.15949 +INFO:tensorflow:step = 41901, loss = 0.282217, precision = 0.96875 (626.998 sec) +INFO:tensorflow:global_step/sec: 0.165675 +INFO:tensorflow:step = 42001, loss = 0.283811, precision = 0.953125 (603.591 sec) +INFO:tensorflow:global_step/sec: 0.166124 +INFO:tensorflow:step = 42101, loss = 0.236907, precision = 0.992188 (601.961 sec) +INFO:tensorflow:global_step/sec: 0.168398 +INFO:tensorflow:step = 42201, loss = 0.235062, precision = 0.992188 (593.832 sec) +Saved checkpoint after 108 epoch(s) to data/resnet164/checkpoints/00108... +INFO:tensorflow:global_step/sec: 0.159642 +INFO:tensorflow:step = 42301, loss = 0.261954, precision = 0.976562 (626.400 sec) +INFO:tensorflow:global_step/sec: 0.154156 +INFO:tensorflow:step = 42401, loss = 0.25411, precision = 0.984375 (648.692 sec) +INFO:tensorflow:global_step/sec: 0.162232 +INFO:tensorflow:step = 42501, loss = 0.227869, precision = 0.984375 (616.401 sec) +INFO:tensorflow:global_step/sec: 0.163424 +INFO:tensorflow:step = 42601, loss = 0.23997, precision = 0.976562 (611.907 sec) +Saved checkpoint after 109 epoch(s) to data/resnet164/checkpoints/00109... +INFO:tensorflow:global_step/sec: 0.16158 +INFO:tensorflow:step = 42701, loss = 0.251731, precision = 0.976562 (618.887 sec) +INFO:tensorflow:global_step/sec: 0.161796 +INFO:tensorflow:step = 42801, loss = 0.21331, precision = 1.0 (618.062 sec) +INFO:tensorflow:global_step/sec: 0.156984 +INFO:tensorflow:step = 42901, loss = 0.24355, precision = 0.984375 (637.006 sec) +INFO:tensorflow:global_step/sec: 0.159453 +INFO:tensorflow:step = 43001, loss = 0.262525, precision = 0.984375 (627.145 sec) +Saved checkpoint after 110 epoch(s) to data/resnet164/checkpoints/00110... +INFO:tensorflow:global_step/sec: 0.159061 +INFO:tensorflow:step = 43101, loss = 0.263977, precision = 0.960938 (628.691 sec) +INFO:tensorflow:global_step/sec: 0.162511 +INFO:tensorflow:step = 43201, loss = 0.2892, precision = 0.96875 (615.344 sec) +INFO:tensorflow:global_step/sec: 0.163587 +INFO:tensorflow:step = 43301, loss = 0.237139, precision = 0.984375 (611.295 sec) +Saved checkpoint after 111 epoch(s) to data/resnet164/checkpoints/00111... +INFO:tensorflow:global_step/sec: 0.163058 +INFO:tensorflow:step = 43401, loss = 0.262986, precision = 0.96875 (613.278 sec) +INFO:tensorflow:global_step/sec: 0.164242 +INFO:tensorflow:step = 43501, loss = 0.23993, precision = 0.976562 (608.856 sec) +INFO:tensorflow:global_step/sec: 0.163597 +INFO:tensorflow:step = 43601, loss = 0.232155, precision = 0.984375 (611.260 sec) +INFO:tensorflow:global_step/sec: 0.164132 +INFO:tensorflow:step = 43701, loss = 0.236077, precision = 0.976562 (609.265 sec) +Saved checkpoint after 112 epoch(s) to data/resnet164/checkpoints/00112... +INFO:tensorflow:global_step/sec: 0.162955 +INFO:tensorflow:step = 43801, loss = 0.207027, precision = 1.0 (613.666 sec) +INFO:tensorflow:global_step/sec: 0.162585 +INFO:tensorflow:step = 43901, loss = 0.245584, precision = 0.96875 (615.061 sec) +INFO:tensorflow:global_step/sec: 0.163982 +INFO:tensorflow:step = 44001, loss = 0.289529, precision = 0.953125 (609.824 sec) +INFO:tensorflow:global_step/sec: 0.161034 +INFO:tensorflow:step = 44101, loss = 0.228284, precision = 0.992188 (620.988 sec) +Saved checkpoint after 113 epoch(s) to data/resnet164/checkpoints/00113... +INFO:tensorflow:global_step/sec: 0.152748 +INFO:tensorflow:step = 44201, loss = 0.220843, precision = 0.984375 (654.675 sec) +INFO:tensorflow:global_step/sec: 0.153706 +INFO:tensorflow:step = 44301, loss = 0.274975, precision = 0.96875 (650.593 sec) +INFO:tensorflow:global_step/sec: 0.157108 +INFO:tensorflow:step = 44401, loss = 0.208387, precision = 0.984375 (636.505 sec) +INFO:tensorflow:global_step/sec: 0.157905 +INFO:tensorflow:step = 44501, loss = 0.222054, precision = 1.0 (633.294 sec) +Saved checkpoint after 114 epoch(s) to data/resnet164/checkpoints/00114... +INFO:tensorflow:global_step/sec: 0.156614 +INFO:tensorflow:step = 44601, loss = 0.201227, precision = 1.0 (638.512 sec) +INFO:tensorflow:global_step/sec: 0.153519 +INFO:tensorflow:step = 44701, loss = 0.236281, precision = 0.984375 (651.383 sec) +INFO:tensorflow:global_step/sec: 0.150723 +INFO:tensorflow:step = 44801, loss = 0.281754, precision = 0.960938 (663.470 sec) +INFO:tensorflow:global_step/sec: 0.154932 +INFO:tensorflow:step = 44901, loss = 0.245821, precision = 0.976562 (645.446 sec) +Saved checkpoint after 115 epoch(s) to data/resnet164/checkpoints/00115... +INFO:tensorflow:global_step/sec: 0.158777 +INFO:tensorflow:step = 45001, loss = 0.204514, precision = 1.0 (629.816 sec) +INFO:tensorflow:global_step/sec: 0.158663 +INFO:tensorflow:step = 45101, loss = 0.214542, precision = 0.984375 (630.268 sec) +INFO:tensorflow:global_step/sec: 0.157996 +INFO:tensorflow:step = 45201, loss = 0.226698, precision = 0.984375 (632.927 sec) +INFO:tensorflow:global_step/sec: 0.156754 +INFO:tensorflow:step = 45301, loss = 0.213178, precision = 0.984375 (637.942 sec) +Saved checkpoint after 116 epoch(s) to data/resnet164/checkpoints/00116... +INFO:tensorflow:global_step/sec: 0.155281 +INFO:tensorflow:step = 45401, loss = 0.213095, precision = 0.984375 (643.995 sec) +INFO:tensorflow:global_step/sec: 0.156643 +INFO:tensorflow:step = 45501, loss = 0.238302, precision = 0.976562 (638.393 sec) +INFO:tensorflow:global_step/sec: 0.156452 +INFO:tensorflow:step = 45601, loss = 0.235222, precision = 0.976562 (639.174 sec) +INFO:tensorflow:global_step/sec: 0.156027 +INFO:tensorflow:step = 45701, loss = 0.19894, precision = 0.984375 (640.914 sec) +Saved checkpoint after 117 epoch(s) to data/resnet164/checkpoints/00117... +INFO:tensorflow:global_step/sec: 0.157113 +INFO:tensorflow:step = 45801, loss = 0.234374, precision = 0.984375 (636.486 sec) +INFO:tensorflow:global_step/sec: 0.133976 +INFO:tensorflow:step = 45901, loss = 0.208881, precision = 0.984375 (746.402 sec) +INFO:tensorflow:global_step/sec: 0.139957 +INFO:tensorflow:step = 46001, loss = 0.210964, precision = 0.984375 (714.504 sec) +INFO:tensorflow:global_step/sec: 0.140628 +INFO:tensorflow:step = 46101, loss = 0.185669, precision = 1.0 (711.095 sec) +Saved checkpoint after 118 epoch(s) to data/resnet164/checkpoints/00118... +INFO:tensorflow:global_step/sec: 0.140674 +INFO:tensorflow:step = 46201, loss = 0.22615, precision = 0.96875 (710.866 sec) +INFO:tensorflow:global_step/sec: 0.148623 +INFO:tensorflow:step = 46301, loss = 0.223429, precision = 0.984375 (672.843 sec) +INFO:tensorflow:global_step/sec: 0.151552 +INFO:tensorflow:step = 46401, loss = 0.207503, precision = 0.984375 (659.841 sec) +INFO:tensorflow:global_step/sec: 0.151195 +INFO:tensorflow:step = 46501, loss = 0.198142, precision = 0.984375 (661.400 sec) +Saved checkpoint after 119 epoch(s) to data/resnet164/checkpoints/00119... +INFO:tensorflow:global_step/sec: 0.153927 +INFO:tensorflow:step = 46601, loss = 0.221271, precision = 0.976562 (649.660 sec) +INFO:tensorflow:global_step/sec: 0.153943 +INFO:tensorflow:step = 46701, loss = 0.264067, precision = 0.96875 (649.593 sec) +INFO:tensorflow:global_step/sec: 0.153506 +INFO:tensorflow:step = 46801, loss = 0.226684, precision = 0.96875 (651.442 sec) +INFO:tensorflow:global_step/sec: 0.153173 +INFO:tensorflow:step = 46901, loss = 0.216887, precision = 0.976562 (652.856 sec) +Saved checkpoint after 120 epoch(s) to data/resnet164/checkpoints/00120... +INFO:tensorflow:global_step/sec: 0.155417 +INFO:tensorflow:step = 47001, loss = 0.209015, precision = 0.984375 (643.432 sec) +INFO:tensorflow:global_step/sec: 0.156064 +INFO:tensorflow:step = 47101, loss = 0.208346, precision = 0.984375 (640.761 sec) +INFO:tensorflow:global_step/sec: 0.151575 +INFO:tensorflow:step = 47201, loss = 0.209839, precision = 0.984375 (659.739 sec) +INFO:tensorflow:global_step/sec: 0.152256 +INFO:tensorflow:step = 47301, loss = 0.224304, precision = 0.96875 (656.787 sec) +Saved checkpoint after 121 epoch(s) to data/resnet164/checkpoints/00121... +INFO:tensorflow:global_step/sec: 0.146133 +INFO:tensorflow:step = 47401, loss = 0.322207, precision = 0.929688 (684.307 sec) +INFO:tensorflow:global_step/sec: 0.144949 +INFO:tensorflow:step = 47501, loss = 0.246491, precision = 0.96875 (689.900 sec) +INFO:tensorflow:global_step/sec: 0.145185 +INFO:tensorflow:step = 47601, loss = 0.20538, precision = 0.976562 (688.778 sec) +INFO:tensorflow:global_step/sec: 0.146629 +INFO:tensorflow:step = 47701, loss = 0.190359, precision = 0.992188 (681.995 sec) +Saved checkpoint after 122 epoch(s) to data/resnet164/checkpoints/00122... +INFO:tensorflow:global_step/sec: 0.148266 +INFO:tensorflow:step = 47801, loss = 0.250156, precision = 0.96875 (674.465 sec) +INFO:tensorflow:global_step/sec: 0.147413 +INFO:tensorflow:step = 47901, loss = 0.186555, precision = 0.992188 (678.368 sec) +INFO:tensorflow:global_step/sec: 0.145982 +INFO:tensorflow:step = 48001, loss = 0.255286, precision = 0.960938 (685.016 sec) +Saved checkpoint after 123 epoch(s) to data/resnet164/checkpoints/00123... +INFO:tensorflow:global_step/sec: 0.151724 +INFO:tensorflow:step = 48101, loss = 0.278672, precision = 0.96875 (659.090 sec) +INFO:tensorflow:global_step/sec: 0.153408 +INFO:tensorflow:step = 48201, loss = 0.228693, precision = 0.976562 (651.856 sec) +INFO:tensorflow:global_step/sec: 0.153651 +INFO:tensorflow:step = 48301, loss = 0.247914, precision = 0.96875 (650.825 sec) +INFO:tensorflow:global_step/sec: 0.155566 +INFO:tensorflow:step = 48401, loss = 0.289431, precision = 0.9375 (642.812 sec) +Saved checkpoint after 124 epoch(s) to data/resnet164/checkpoints/00124... +INFO:tensorflow:global_step/sec: 0.154315 +INFO:tensorflow:step = 48501, loss = 0.191362, precision = 0.992188 (648.027 sec) +INFO:tensorflow:global_step/sec: 0.154226 +INFO:tensorflow:step = 48601, loss = 0.258529, precision = 0.976562 (648.401 sec) +INFO:tensorflow:global_step/sec: 0.154527 +INFO:tensorflow:step = 48701, loss = 0.227872, precision = 0.960938 (647.135 sec) +INFO:tensorflow:global_step/sec: 0.154604 +INFO:tensorflow:step = 48801, loss = 0.231088, precision = 0.96875 (646.814 sec) +Saved checkpoint after 125 epoch(s) to data/resnet164/checkpoints/00125... +INFO:tensorflow:global_step/sec: 0.153479 +INFO:tensorflow:step = 48901, loss = 0.210748, precision = 0.976562 (651.556 sec) +INFO:tensorflow:global_step/sec: 0.15197 +INFO:tensorflow:step = 49001, loss = 0.215948, precision = 0.984375 (658.023 sec) +INFO:tensorflow:global_step/sec: 0.15314 +INFO:tensorflow:step = 49101, loss = 0.2017, precision = 0.992188 (652.997 sec) +INFO:tensorflow:global_step/sec: 0.149785 +INFO:tensorflow:step = 49201, loss = 0.198605, precision = 0.984375 (667.623 sec) +Saved checkpoint after 126 epoch(s) to data/resnet164/checkpoints/00126... +INFO:tensorflow:global_step/sec: 0.147838 +INFO:tensorflow:step = 49301, loss = 0.213742, precision = 0.976562 (676.415 sec) +INFO:tensorflow:global_step/sec: 0.149435 +INFO:tensorflow:step = 49401, loss = 0.181294, precision = 0.992188 (669.186 sec) +INFO:tensorflow:global_step/sec: 0.150599 +INFO:tensorflow:step = 49501, loss = 0.188793, precision = 0.984375 (664.014 sec) +INFO:tensorflow:global_step/sec: 0.153373 +INFO:tensorflow:step = 49601, loss = 0.180606, precision = 0.992188 (652.004 sec) +Saved checkpoint after 127 epoch(s) to data/resnet164/checkpoints/00127... +INFO:tensorflow:global_step/sec: 0.153545 +INFO:tensorflow:step = 49701, loss = 0.221729, precision = 0.992188 (651.274 sec) +INFO:tensorflow:global_step/sec: 0.155161 +INFO:tensorflow:step = 49801, loss = 0.191599, precision = 0.976562 (644.492 sec) +INFO:tensorflow:global_step/sec: 0.156451 +INFO:tensorflow:step = 49901, loss = 0.249142, precision = 0.953125 (639.179 sec) +INFO:tensorflow:global_step/sec: 0.156902 +INFO:tensorflow:step = 50001, loss = 0.224984, precision = 0.953125 (637.340 sec) +Saved checkpoint after 128 epoch(s) to data/resnet164/checkpoints/00128... +INFO:tensorflow:global_step/sec: 0.155237 +INFO:tensorflow:step = 50101, loss = 0.176548, precision = 0.992188 (644.176 sec) +INFO:tensorflow:global_step/sec: 0.15596 +INFO:tensorflow:step = 50201, loss = 0.215119, precision = 0.992188 (641.190 sec) +INFO:tensorflow:global_step/sec: 0.156941 +INFO:tensorflow:step = 50301, loss = 0.266945, precision = 0.96875 (637.181 sec) +INFO:tensorflow:global_step/sec: 0.158699 +INFO:tensorflow:step = 50401, loss = 0.174299, precision = 1.0 (630.123 sec) +Saved checkpoint after 129 epoch(s) to data/resnet164/checkpoints/00129... +INFO:tensorflow:global_step/sec: 0.159097 +INFO:tensorflow:step = 50501, loss = 0.210652, precision = 0.976562 (628.549 sec) +INFO:tensorflow:global_step/sec: 0.149896 +INFO:tensorflow:step = 50601, loss = 0.201244, precision = 0.976562 (667.128 sec) +INFO:tensorflow:global_step/sec: 0.140317 +INFO:tensorflow:step = 50701, loss = 0.250014, precision = 0.960938 (712.671 sec) +INFO:tensorflow:global_step/sec: 0.150909 +INFO:tensorflow:step = 50801, loss = 0.200557, precision = 0.976562 (662.652 sec) +Saved checkpoint after 130 epoch(s) to data/resnet164/checkpoints/00130... +INFO:tensorflow:global_step/sec: 0.149045 +INFO:tensorflow:step = 50901, loss = 0.17888, precision = 0.992188 (670.938 sec) +INFO:tensorflow:global_step/sec: 0.15703 +INFO:tensorflow:step = 51001, loss = 0.213061, precision = 0.976562 (636.820 sec) +INFO:tensorflow:global_step/sec: 0.156998 +INFO:tensorflow:step = 51101, loss = 0.235842, precision = 0.953125 (636.951 sec) +INFO:tensorflow:global_step/sec: 0.155798 +INFO:tensorflow:step = 51201, loss = 0.183121, precision = 0.984375 (641.858 sec) +Saved checkpoint after 131 epoch(s) to data/resnet164/checkpoints/00131... +INFO:tensorflow:global_step/sec: 0.155146 +INFO:tensorflow:step = 51301, loss = 0.177609, precision = 0.992188 (644.552 sec) +INFO:tensorflow:global_step/sec: 0.154003 +INFO:tensorflow:step = 51401, loss = 0.219825, precision = 0.960938 (649.339 sec) +INFO:tensorflow:global_step/sec: 0.155198 +INFO:tensorflow:step = 51501, loss = 0.195547, precision = 0.984375 (644.339 sec) +INFO:tensorflow:global_step/sec: 0.157021 +INFO:tensorflow:step = 51601, loss = 0.186723, precision = 0.984375 (636.858 sec) +Saved checkpoint after 132 epoch(s) to data/resnet164/checkpoints/00132... +INFO:tensorflow:global_step/sec: 0.155128 +INFO:tensorflow:step = 51701, loss = 0.166947, precision = 0.992188 (644.627 sec) +INFO:tensorflow:global_step/sec: 0.151105 +INFO:tensorflow:step = 51801, loss = 0.268701, precision = 0.96875 (661.793 sec) +INFO:tensorflow:global_step/sec: 0.149727 +INFO:tensorflow:step = 51901, loss = 0.182158, precision = 1.0 (667.881 sec) +INFO:tensorflow:global_step/sec: 0.151594 +INFO:tensorflow:step = 52001, loss = 0.192351, precision = 0.992188 (659.658 sec) +Saved checkpoint after 133 epoch(s) to data/resnet164/checkpoints/00133... +INFO:tensorflow:global_step/sec: 0.13926 +INFO:tensorflow:step = 52101, loss = 0.214991, precision = 0.984375 (718.079 sec) +INFO:tensorflow:global_step/sec: 0.149347 +INFO:tensorflow:step = 52201, loss = 0.172769, precision = 0.984375 (669.582 sec) +INFO:tensorflow:global_step/sec: 0.147611 +INFO:tensorflow:step = 52301, loss = 0.202033, precision = 0.984375 (677.458 sec) +Saved checkpoint after 134 epoch(s) to data/resnet164/checkpoints/00134... +INFO:tensorflow:global_step/sec: 0.14957 +INFO:tensorflow:step = 52401, loss = 0.180629, precision = 0.992188 (668.584 sec) +INFO:tensorflow:global_step/sec: 0.144967 +INFO:tensorflow:step = 52501, loss = 0.190121, precision = 0.984375 (689.813 sec) +INFO:tensorflow:global_step/sec: 0.14676 +INFO:tensorflow:step = 52601, loss = 0.187825, precision = 0.984375 (681.386 sec) +INFO:tensorflow:global_step/sec: 0.148036 +INFO:tensorflow:step = 52701, loss = 0.188305, precision = 0.984375 (675.510 sec) +Saved checkpoint after 135 epoch(s) to data/resnet164/checkpoints/00135... +INFO:tensorflow:global_step/sec: 0.128163 +INFO:tensorflow:step = 52801, loss = 0.202063, precision = 0.984375 (780.256 sec) +INFO:tensorflow:global_step/sec: 0.120829 +INFO:tensorflow:step = 52901, loss = 0.239575, precision = 0.960938 (827.612 sec) +INFO:tensorflow:global_step/sec: 0.136841 +INFO:tensorflow:step = 53001, loss = 0.186953, precision = 0.984375 (730.773 sec) +INFO:tensorflow:global_step/sec: 0.149521 +INFO:tensorflow:step = 53101, loss = 0.229837, precision = 0.960938 (668.803 sec) +Saved checkpoint after 136 epoch(s) to data/resnet164/checkpoints/00136... +INFO:tensorflow:global_step/sec: 0.147575 +INFO:tensorflow:step = 53201, loss = 0.161166, precision = 0.992188 (677.622 sec) +INFO:tensorflow:global_step/sec: 0.150968 +INFO:tensorflow:step = 53301, loss = 0.156133, precision = 0.992188 (662.391 sec) +INFO:tensorflow:global_step/sec: 0.150661 +INFO:tensorflow:step = 53401, loss = 0.148818, precision = 1.0 (663.744 sec) +INFO:tensorflow:global_step/sec: 0.15084 +INFO:tensorflow:step = 53501, loss = 0.154856, precision = 1.0 (662.956 sec) +Saved checkpoint after 137 epoch(s) to data/resnet164/checkpoints/00137... +INFO:tensorflow:global_step/sec: 0.150425 +INFO:tensorflow:step = 53601, loss = 0.162522, precision = 0.992188 (664.784 sec) +INFO:tensorflow:global_step/sec: 0.150682 +INFO:tensorflow:step = 53701, loss = 0.1693, precision = 0.992188 (663.650 sec) +INFO:tensorflow:global_step/sec: 0.151246 +INFO:tensorflow:step = 53801, loss = 0.163769, precision = 1.0 (661.174 sec) +INFO:tensorflow:global_step/sec: 0.15152 +INFO:tensorflow:step = 53901, loss = 0.159226, precision = 0.992188 (659.979 sec) +Saved checkpoint after 138 epoch(s) to data/resnet164/checkpoints/00138... +INFO:tensorflow:global_step/sec: 0.149815 +INFO:tensorflow:step = 54001, loss = 0.174077, precision = 0.984375 (667.488 sec) +INFO:tensorflow:global_step/sec: 0.15684 +INFO:tensorflow:step = 54101, loss = 0.152228, precision = 1.0 (637.592 sec) +INFO:tensorflow:global_step/sec: 0.155697 +INFO:tensorflow:step = 54201, loss = 0.164592, precision = 0.992188 (642.273 sec) +INFO:tensorflow:global_step/sec: 0.155335 +INFO:tensorflow:step = 54301, loss = 0.147065, precision = 1.0 (643.771 sec) +Saved checkpoint after 139 epoch(s) to data/resnet164/checkpoints/00139... +INFO:tensorflow:global_step/sec: 0.153612 +INFO:tensorflow:step = 54401, loss = 0.157168, precision = 1.0 (650.989 sec) +INFO:tensorflow:global_step/sec: 0.155698 +INFO:tensorflow:step = 54501, loss = 0.168107, precision = 0.992188 (642.268 sec) +INFO:tensorflow:global_step/sec: 0.157195 +INFO:tensorflow:step = 54601, loss = 0.152529, precision = 1.0 (636.151 sec) +INFO:tensorflow:global_step/sec: 0.157063 +INFO:tensorflow:step = 54701, loss = 0.150633, precision = 1.0 (636.686 sec) +Saved checkpoint after 140 epoch(s) to data/resnet164/checkpoints/00140... +INFO:tensorflow:global_step/sec: 0.156007 +INFO:tensorflow:step = 54801, loss = 0.167502, precision = 0.992188 (640.997 sec) +INFO:tensorflow:global_step/sec: 0.157424 +INFO:tensorflow:step = 54901, loss = 0.149682, precision = 1.0 (635.227 sec) +INFO:tensorflow:global_step/sec: 0.156976 +INFO:tensorflow:step = 55001, loss = 0.147029, precision = 1.0 (637.040 sec) +INFO:tensorflow:global_step/sec: 0.155959 +INFO:tensorflow:step = 55101, loss = 0.145601, precision = 1.0 (641.196 sec) +Saved checkpoint after 141 epoch(s) to data/resnet164/checkpoints/00141... +INFO:tensorflow:global_step/sec: 0.157284 +INFO:tensorflow:step = 55201, loss = 0.143317, precision = 1.0 (635.793 sec) +INFO:tensorflow:global_step/sec: 0.159325 +INFO:tensorflow:step = 55301, loss = 0.150904, precision = 1.0 (627.649 sec) +INFO:tensorflow:global_step/sec: 0.158604 +INFO:tensorflow:step = 55401, loss = 0.14376, precision = 1.0 (630.502 sec) +INFO:tensorflow:global_step/sec: 0.159082 +INFO:tensorflow:step = 55501, loss = 0.144975, precision = 1.0 (628.607 sec) +Saved checkpoint after 142 epoch(s) to data/resnet164/checkpoints/00142... +INFO:tensorflow:global_step/sec: 0.162292 +INFO:tensorflow:step = 55601, loss = 0.154095, precision = 1.0 (616.174 sec) +INFO:tensorflow:global_step/sec: 0.164373 +INFO:tensorflow:step = 55701, loss = 0.156877, precision = 1.0 (608.373 sec) +INFO:tensorflow:global_step/sec: 0.163577 +INFO:tensorflow:step = 55801, loss = 0.145624, precision = 1.0 (611.332 sec) +INFO:tensorflow:global_step/sec: 0.16423 +INFO:tensorflow:step = 55901, loss = 0.147407, precision = 1.0 (608.902 sec) +Saved checkpoint after 143 epoch(s) to data/resnet164/checkpoints/00143... +INFO:tensorflow:global_step/sec: 0.163525 +INFO:tensorflow:step = 56001, loss = 0.149945, precision = 1.0 (611.528 sec) +INFO:tensorflow:global_step/sec: 0.164349 +INFO:tensorflow:step = 56101, loss = 0.145465, precision = 1.0 (608.462 sec) +INFO:tensorflow:global_step/sec: 0.163801 +INFO:tensorflow:step = 56201, loss = 0.143024, precision = 1.0 (610.497 sec) +INFO:tensorflow:global_step/sec: 0.164621 +INFO:tensorflow:step = 56301, loss = 0.142224, precision = 1.0 (607.456 sec) +Saved checkpoint after 144 epoch(s) to data/resnet164/checkpoints/00144... +INFO:tensorflow:global_step/sec: 0.164491 +INFO:tensorflow:step = 56401, loss = 0.160175, precision = 0.992188 (607.935 sec) +INFO:tensorflow:global_step/sec: 0.165064 +INFO:tensorflow:step = 56501, loss = 0.143079, precision = 1.0 (605.824 sec) +INFO:tensorflow:global_step/sec: 0.162933 +INFO:tensorflow:step = 56601, loss = 0.14511, precision = 1.0 (613.751 sec) +Saved checkpoint after 145 epoch(s) to data/resnet164/checkpoints/00145... +INFO:tensorflow:global_step/sec: 0.163036 +INFO:tensorflow:step = 56701, loss = 0.142664, precision = 1.0 (613.361 sec) +INFO:tensorflow:global_step/sec: 0.158011 +INFO:tensorflow:step = 56801, loss = 0.142947, precision = 1.0 (632.869 sec) +INFO:tensorflow:global_step/sec: 0.157043 +INFO:tensorflow:step = 56901, loss = 0.144279, precision = 1.0 (636.769 sec) +INFO:tensorflow:global_step/sec: 0.156923 +INFO:tensorflow:step = 57001, loss = 0.151851, precision = 1.0 (637.256 sec) +Saved checkpoint after 146 epoch(s) to data/resnet164/checkpoints/00146... +INFO:tensorflow:global_step/sec: 0.156584 +INFO:tensorflow:step = 57101, loss = 0.146061, precision = 1.0 (638.635 sec) +INFO:tensorflow:global_step/sec: 0.158142 +INFO:tensorflow:step = 57201, loss = 0.143484, precision = 1.0 (632.343 sec) +INFO:tensorflow:global_step/sec: 0.160966 +INFO:tensorflow:step = 57301, loss = 0.140748, precision = 1.0 (621.250 sec) +INFO:tensorflow:global_step/sec: 0.16282 +INFO:tensorflow:step = 57401, loss = 0.150171, precision = 0.992188 (614.174 sec) +Saved checkpoint after 147 epoch(s) to data/resnet164/checkpoints/00147... +INFO:tensorflow:global_step/sec: 0.162166 +INFO:tensorflow:step = 57501, loss = 0.139738, precision = 1.0 (616.651 sec) +INFO:tensorflow:global_step/sec: 0.163696 +INFO:tensorflow:step = 57601, loss = 0.15241, precision = 1.0 (610.888 sec) +INFO:tensorflow:global_step/sec: 0.160576 +INFO:tensorflow:step = 57701, loss = 0.142578, precision = 1.0 (622.758 sec) +INFO:tensorflow:global_step/sec: 0.159457 +INFO:tensorflow:step = 57801, loss = 0.142678, precision = 1.0 (627.127 sec) +Saved checkpoint after 148 epoch(s) to data/resnet164/checkpoints/00148... +INFO:tensorflow:global_step/sec: 0.160044 +INFO:tensorflow:step = 57901, loss = 0.156103, precision = 0.992188 (624.827 sec) +INFO:tensorflow:global_step/sec: 0.158537 +INFO:tensorflow:step = 58001, loss = 0.143221, precision = 1.0 (630.768 sec) +INFO:tensorflow:global_step/sec: 0.159084 +INFO:tensorflow:step = 58101, loss = 0.140446, precision = 1.0 (628.599 sec) +INFO:tensorflow:global_step/sec: 0.158207 +INFO:tensorflow:step = 58201, loss = 0.139296, precision = 1.0 (632.083 sec) +Saved checkpoint after 149 epoch(s) to data/resnet164/checkpoints/00149... +INFO:tensorflow:global_step/sec: 0.159485 +INFO:tensorflow:step = 58301, loss = 0.143854, precision = 1.0 (627.018 sec) +INFO:tensorflow:global_step/sec: 0.159515 +INFO:tensorflow:step = 58401, loss = 0.159904, precision = 0.984375 (626.900 sec) +INFO:tensorflow:global_step/sec: 0.155894 +INFO:tensorflow:step = 58501, loss = 0.140656, precision = 1.0 (641.462 sec) +INFO:tensorflow:global_step/sec: 0.129232 +INFO:tensorflow:step = 58601, loss = 0.147512, precision = 1.0 (773.802 sec) +Saved checkpoint after 150 epoch(s) to data/resnet164/checkpoints/00150... +INFO:tensorflow:global_step/sec: 0.127654 +INFO:tensorflow:step = 58701, loss = 0.138182, precision = 1.0 (783.368 sec) +INFO:tensorflow:global_step/sec: 0.127397 +INFO:tensorflow:step = 58801, loss = 0.142917, precision = 1.0 (784.948 sec) +INFO:tensorflow:global_step/sec: 0.126909 +INFO:tensorflow:step = 58901, loss = 0.138863, precision = 1.0 (787.969 sec) +INFO:tensorflow:global_step/sec: 0.127138 +INFO:tensorflow:step = 59001, loss = 0.13777, precision = 1.0 (786.544 sec) +Saved checkpoint after 151 epoch(s) to data/resnet164/checkpoints/00151... +INFO:tensorflow:global_step/sec: 0.126748 +INFO:tensorflow:step = 59101, loss = 0.140502, precision = 1.0 (788.968 sec) +INFO:tensorflow:global_step/sec: 0.127089 +INFO:tensorflow:step = 59201, loss = 0.139027, precision = 1.0 (786.848 sec) +INFO:tensorflow:global_step/sec: 0.128891 +INFO:tensorflow:step = 59301, loss = 0.13925, precision = 1.0 (775.850 sec) +INFO:tensorflow:global_step/sec: 0.128053 +INFO:tensorflow:step = 59401, loss = 0.141141, precision = 1.0 (780.925 sec) +Saved checkpoint after 152 epoch(s) to data/resnet164/checkpoints/00152... +INFO:tensorflow:global_step/sec: 0.127472 +INFO:tensorflow:step = 59501, loss = 0.144652, precision = 0.992188 (784.483 sec) +INFO:tensorflow:global_step/sec: 0.125977 +INFO:tensorflow:step = 59601, loss = 0.141126, precision = 1.0 (793.798 sec) +INFO:tensorflow:global_step/sec: 0.124801 +INFO:tensorflow:step = 59701, loss = 0.144262, precision = 0.992188 (801.275 sec) +INFO:tensorflow:global_step/sec: 0.124114 +INFO:tensorflow:step = 59801, loss = 0.142346, precision = 1.0 (805.712 sec) +Saved checkpoint after 153 epoch(s) to data/resnet164/checkpoints/00153... +INFO:tensorflow:global_step/sec: 0.124326 +INFO:tensorflow:step = 59901, loss = 0.139989, precision = 1.0 (804.340 sec) +INFO:tensorflow:global_step/sec: 0.125165 +INFO:tensorflow:step = 60001, loss = 0.156196, precision = 1.0 (798.948 sec) +INFO:tensorflow:global_step/sec: 0.124301 +INFO:tensorflow:step = 60101, loss = 0.136946, precision = 1.0 (804.502 sec) +INFO:tensorflow:global_step/sec: 0.124603 +INFO:tensorflow:step = 60201, loss = 0.138113, precision = 1.0 (802.551 sec) +Saved checkpoint after 154 epoch(s) to data/resnet164/checkpoints/00154... +INFO:tensorflow:global_step/sec: 0.121597 +INFO:tensorflow:step = 60301, loss = 0.137366, precision = 1.0 (822.388 sec) +INFO:tensorflow:global_step/sec: 0.124607 +INFO:tensorflow:step = 60401, loss = 0.139609, precision = 1.0 (802.526 sec) +INFO:tensorflow:global_step/sec: 0.125059 +INFO:tensorflow:step = 60501, loss = 0.135533, precision = 1.0 (799.621 sec) +INFO:tensorflow:global_step/sec: 0.125645 +INFO:tensorflow:step = 60601, loss = 0.137067, precision = 1.0 (795.891 sec) +Saved checkpoint after 155 epoch(s) to data/resnet164/checkpoints/00155... +INFO:tensorflow:global_step/sec: 0.125063 +INFO:tensorflow:step = 60701, loss = 0.135484, precision = 1.0 (799.595 sec) +INFO:tensorflow:global_step/sec: 0.122433 +INFO:tensorflow:step = 60801, loss = 0.137893, precision = 1.0 (816.776 sec) +INFO:tensorflow:global_step/sec: 0.124998 +INFO:tensorflow:step = 60901, loss = 0.137006, precision = 1.0 (800.011 sec) +Saved checkpoint after 156 epoch(s) to data/resnet164/checkpoints/00156... +INFO:tensorflow:global_step/sec: 0.125052 +INFO:tensorflow:step = 61001, loss = 0.134746, precision = 1.0 (799.668 sec) +INFO:tensorflow:global_step/sec: 0.12532 +INFO:tensorflow:step = 61101, loss = 0.135804, precision = 1.0 (797.956 sec) +INFO:tensorflow:global_step/sec: 0.124656 +INFO:tensorflow:step = 61201, loss = 0.137898, precision = 1.0 (802.206 sec) +INFO:tensorflow:global_step/sec: 0.124336 +INFO:tensorflow:step = 61301, loss = 0.135031, precision = 1.0 (804.271 sec) +Saved checkpoint after 157 epoch(s) to data/resnet164/checkpoints/00157... +INFO:tensorflow:global_step/sec: 0.125567 +INFO:tensorflow:step = 61401, loss = 0.135305, precision = 1.0 (796.385 sec) +INFO:tensorflow:global_step/sec: 0.126038 +INFO:tensorflow:step = 61501, loss = 0.136032, precision = 1.0 (793.414 sec) +INFO:tensorflow:global_step/sec: 0.125881 +INFO:tensorflow:step = 61601, loss = 0.136179, precision = 1.0 (794.401 sec) +INFO:tensorflow:global_step/sec: 0.125509 +INFO:tensorflow:step = 61701, loss = 0.132972, precision = 1.0 (796.757 sec) +Saved checkpoint after 158 epoch(s) to data/resnet164/checkpoints/00158... +INFO:tensorflow:global_step/sec: 0.124254 +INFO:tensorflow:step = 61801, loss = 0.134003, precision = 1.0 (804.805 sec) +INFO:tensorflow:global_step/sec: 0.123809 +INFO:tensorflow:step = 61901, loss = 0.133809, precision = 1.0 (807.693 sec) +INFO:tensorflow:global_step/sec: 0.127028 +INFO:tensorflow:step = 62001, loss = 0.134944, precision = 1.0 (787.229 sec) +INFO:tensorflow:global_step/sec: 0.127084 +INFO:tensorflow:step = 62101, loss = 0.136915, precision = 1.0 (786.880 sec) +Saved checkpoint after 159 epoch(s) to data/resnet164/checkpoints/00159... +INFO:tensorflow:global_step/sec: 0.122145 +INFO:tensorflow:step = 62201, loss = 0.135118, precision = 1.0 (818.696 sec) +INFO:tensorflow:global_step/sec: 0.12819 +INFO:tensorflow:step = 62301, loss = 0.132503, precision = 1.0 (780.091 sec) +INFO:tensorflow:global_step/sec: 0.126398 +INFO:tensorflow:step = 62401, loss = 0.134004, precision = 1.0 (791.154 sec) +INFO:tensorflow:global_step/sec: 0.132267 +INFO:tensorflow:step = 62501, loss = 0.133425, precision = 1.0 (756.044 sec) +Saved checkpoint after 160 epoch(s) to data/resnet164/checkpoints/00160... +INFO:tensorflow:global_step/sec: 0.131781 +INFO:tensorflow:step = 62601, loss = 0.132279, precision = 1.0 (758.832 sec) +INFO:tensorflow:global_step/sec: 0.132892 +INFO:tensorflow:step = 62701, loss = 0.132417, precision = 1.0 (752.490 sec) +INFO:tensorflow:global_step/sec: 0.134147 +INFO:tensorflow:step = 62801, loss = 0.133283, precision = 1.0 (745.450 sec) +INFO:tensorflow:global_step/sec: 0.133161 +INFO:tensorflow:step = 62901, loss = 0.132385, precision = 1.0 (750.973 sec) +Saved checkpoint after 161 epoch(s) to data/resnet164/checkpoints/00161... +INFO:tensorflow:global_step/sec: 0.133093 +INFO:tensorflow:step = 63001, loss = 0.134781, precision = 1.0 (751.354 sec) +INFO:tensorflow:global_step/sec: 0.132545 +INFO:tensorflow:step = 63101, loss = 0.134631, precision = 1.0 (754.459 sec) +INFO:tensorflow:global_step/sec: 0.132216 +INFO:tensorflow:step = 63201, loss = 0.13236, precision = 1.0 (756.336 sec) +INFO:tensorflow:global_step/sec: 0.133445 +INFO:tensorflow:step = 63301, loss = 0.134416, precision = 1.0 (749.373 sec) +Saved checkpoint after 162 epoch(s) to data/resnet164/checkpoints/00162... +INFO:tensorflow:global_step/sec: 0.133263 +INFO:tensorflow:step = 63401, loss = 0.131395, precision = 1.0 (750.394 sec) +INFO:tensorflow:global_step/sec: 0.132844 +INFO:tensorflow:step = 63501, loss = 0.132259, precision = 1.0 (752.761 sec) +INFO:tensorflow:global_step/sec: 0.129905 +INFO:tensorflow:step = 63601, loss = 0.135634, precision = 1.0 (769.794 sec) +INFO:tensorflow:global_step/sec: 0.13227 +INFO:tensorflow:step = 63701, loss = 0.133926, precision = 1.0 (756.031 sec) +Saved checkpoint after 163 epoch(s) to data/resnet164/checkpoints/00163... +INFO:tensorflow:global_step/sec: 0.131417 +INFO:tensorflow:step = 63801, loss = 0.137141, precision = 0.992188 (760.937 sec) +INFO:tensorflow:global_step/sec: 0.1321 +INFO:tensorflow:step = 63901, loss = 0.130936, precision = 1.0 (757.000 sec) +INFO:tensorflow:global_step/sec: 0.132273 +INFO:tensorflow:step = 64001, loss = 0.143096, precision = 0.992188 (756.014 sec) +INFO:tensorflow:global_step/sec: 0.131555 +INFO:tensorflow:step = 64101, loss = 0.136508, precision = 1.0 (760.137 sec) +Saved checkpoint after 164 epoch(s) to data/resnet164/checkpoints/00164... +INFO:tensorflow:global_step/sec: 0.133683 +INFO:tensorflow:step = 64201, loss = 0.139176, precision = 1.0 (748.040 sec) +INFO:tensorflow:global_step/sec: 0.133927 +INFO:tensorflow:step = 64301, loss = 0.132092, precision = 1.0 (746.677 sec) +INFO:tensorflow:global_step/sec: 0.133393 +INFO:tensorflow:step = 64401, loss = 0.130172, precision = 1.0 (749.666 sec) +INFO:tensorflow:global_step/sec: 0.133126 +INFO:tensorflow:step = 64501, loss = 0.131595, precision = 1.0 (751.171 sec) +Saved checkpoint after 165 epoch(s) to data/resnet164/checkpoints/00165... +INFO:tensorflow:global_step/sec: 0.132632 +INFO:tensorflow:step = 64601, loss = 0.131378, precision = 1.0 (753.967 sec) +INFO:tensorflow:global_step/sec: 0.133838 +INFO:tensorflow:step = 64701, loss = 0.13196, precision = 1.0 (747.175 sec) +INFO:tensorflow:global_step/sec: 0.132161 +INFO:tensorflow:step = 64801, loss = 0.130589, precision = 1.0 (756.650 sec) +INFO:tensorflow:global_step/sec: 0.133215 +INFO:tensorflow:step = 64901, loss = 0.129769, precision = 1.0 (750.668 sec) +Saved checkpoint after 166 epoch(s) to data/resnet164/checkpoints/00166... +INFO:tensorflow:global_step/sec: 0.132643 +INFO:tensorflow:step = 65001, loss = 0.132681, precision = 1.0 (753.901 sec) +INFO:tensorflow:global_step/sec: 0.135416 +INFO:tensorflow:step = 65101, loss = 0.130529, precision = 1.0 (738.463 sec) +INFO:tensorflow:global_step/sec: 0.135323 +INFO:tensorflow:step = 65201, loss = 0.129484, precision = 1.0 (738.971 sec) +Saved checkpoint after 167 epoch(s) to data/resnet164/checkpoints/00167... +INFO:tensorflow:global_step/sec: 0.136151 +INFO:tensorflow:step = 65301, loss = 0.129228, precision = 1.0 (734.479 sec) +INFO:tensorflow:global_step/sec: 0.133884 +INFO:tensorflow:step = 65401, loss = 0.131522, precision = 1.0 (746.915 sec) +INFO:tensorflow:global_step/sec: 0.130975 +INFO:tensorflow:step = 65501, loss = 0.128409, precision = 1.0 (763.503 sec) +INFO:tensorflow:global_step/sec: 0.130877 +INFO:tensorflow:step = 65601, loss = 0.129377, precision = 1.0 (764.075 sec) +Saved checkpoint after 168 epoch(s) to data/resnet164/checkpoints/00168... +INFO:tensorflow:global_step/sec: 0.132804 +INFO:tensorflow:step = 65701, loss = 0.130825, precision = 1.0 (752.990 sec) +INFO:tensorflow:global_step/sec: 0.129463 +INFO:tensorflow:step = 65801, loss = 0.129182, precision = 1.0 (772.424 sec) +INFO:tensorflow:global_step/sec: 0.131603 +INFO:tensorflow:step = 65901, loss = 0.12983, precision = 1.0 (759.863 sec) +INFO:tensorflow:global_step/sec: 0.134728 +INFO:tensorflow:step = 66001, loss = 0.127916, precision = 1.0 (742.234 sec) +Saved checkpoint after 169 epoch(s) to data/resnet164/checkpoints/00169... +INFO:tensorflow:global_step/sec: 0.134546 +INFO:tensorflow:step = 66101, loss = 0.129637, precision = 1.0 (743.238 sec) +INFO:tensorflow:global_step/sec: 0.133447 +INFO:tensorflow:step = 66201, loss = 0.133232, precision = 1.0 (749.363 sec) +INFO:tensorflow:global_step/sec: 0.133822 +INFO:tensorflow:step = 66301, loss = 0.128482, precision = 1.0 (747.263 sec) +INFO:tensorflow:global_step/sec: 0.13079 +INFO:tensorflow:step = 66401, loss = 0.134107, precision = 1.0 (764.583 sec) +Saved checkpoint after 170 epoch(s) to data/resnet164/checkpoints/00170... +INFO:tensorflow:global_step/sec: 0.134211 +INFO:tensorflow:step = 66501, loss = 0.132301, precision = 1.0 (745.094 sec) +INFO:tensorflow:global_step/sec: 0.135311 +INFO:tensorflow:step = 66601, loss = 0.12761, precision = 1.0 (739.038 sec) +INFO:tensorflow:global_step/sec: 0.133971 +INFO:tensorflow:step = 66701, loss = 0.13292, precision = 1.0 (746.432 sec) +INFO:tensorflow:global_step/sec: 0.134037 +INFO:tensorflow:step = 66801, loss = 0.127109, precision = 1.0 (746.064 sec) +Saved checkpoint after 171 epoch(s) to data/resnet164/checkpoints/00171... +INFO:tensorflow:global_step/sec: 0.135549 +INFO:tensorflow:step = 66901, loss = 0.127778, precision = 1.0 (737.739 sec) +INFO:tensorflow:global_step/sec: 0.137227 +INFO:tensorflow:step = 67001, loss = 0.129602, precision = 1.0 (728.720 sec) +INFO:tensorflow:global_step/sec: 0.136656 +INFO:tensorflow:step = 67101, loss = 0.127247, precision = 1.0 (731.765 sec) +INFO:tensorflow:global_step/sec: 0.137459 +INFO:tensorflow:step = 67201, loss = 0.132811, precision = 1.0 (727.488 sec) +Saved checkpoint after 172 epoch(s) to data/resnet164/checkpoints/00172... +INFO:tensorflow:global_step/sec: 0.136273 +INFO:tensorflow:step = 67301, loss = 0.130717, precision = 1.0 (733.821 sec) +INFO:tensorflow:global_step/sec: 0.135495 +INFO:tensorflow:step = 67401, loss = 0.126418, precision = 1.0 (738.036 sec) +INFO:tensorflow:global_step/sec: 0.131696 +INFO:tensorflow:step = 67501, loss = 0.129598, precision = 1.0 (759.322 sec) +INFO:tensorflow:global_step/sec: 0.131706 +INFO:tensorflow:step = 67601, loss = 0.136538, precision = 0.992188 (759.265 sec) +Saved checkpoint after 173 epoch(s) to data/resnet164/checkpoints/00173... +INFO:tensorflow:global_step/sec: 0.131497 +INFO:tensorflow:step = 67701, loss = 0.126117, precision = 1.0 (760.474 sec) +INFO:tensorflow:global_step/sec: 0.131332 +INFO:tensorflow:step = 67801, loss = 0.127598, precision = 1.0 (761.431 sec) +INFO:tensorflow:global_step/sec: 0.128833 +INFO:tensorflow:step = 67901, loss = 0.126059, precision = 1.0 (776.196 sec) +INFO:tensorflow:global_step/sec: 0.131278 +INFO:tensorflow:step = 68001, loss = 0.126483, precision = 1.0 (761.740 sec) +Saved checkpoint after 174 epoch(s) to data/resnet164/checkpoints/00174... +INFO:tensorflow:global_step/sec: 0.132758 +INFO:tensorflow:step = 68101, loss = 0.126238, precision = 1.0 (753.250 sec) +INFO:tensorflow:global_step/sec: 0.136884 +INFO:tensorflow:step = 68201, loss = 0.125024, precision = 1.0 (730.546 sec) +INFO:tensorflow:global_step/sec: 0.137299 +INFO:tensorflow:step = 68301, loss = 0.125815, precision = 1.0 (728.338 sec) +INFO:tensorflow:global_step/sec: 0.133614 +INFO:tensorflow:step = 68401, loss = 0.124781, precision = 1.0 (748.424 sec) +Saved checkpoint after 175 epoch(s) to data/resnet164/checkpoints/00175... +INFO:tensorflow:global_step/sec: 0.13266 +INFO:tensorflow:step = 68501, loss = 0.125413, precision = 1.0 (753.805 sec) +INFO:tensorflow:global_step/sec: 0.128511 +INFO:tensorflow:step = 68601, loss = 0.125197, precision = 1.0 (778.143 sec) +INFO:tensorflow:global_step/sec: 0.132719 +INFO:tensorflow:step = 68701, loss = 0.124569, precision = 1.0 (753.472 sec) +INFO:tensorflow:global_step/sec: 0.137367 +INFO:tensorflow:step = 68801, loss = 0.124541, precision = 1.0 (727.977 sec) +Saved checkpoint after 176 epoch(s) to data/resnet164/checkpoints/00176... +INFO:tensorflow:global_step/sec: 0.130128 +INFO:tensorflow:step = 68901, loss = 0.132813, precision = 0.992188 (768.475 sec) +INFO:tensorflow:global_step/sec: 0.132726 +INFO:tensorflow:step = 69001, loss = 0.127385, precision = 1.0 (753.429 sec) +INFO:tensorflow:global_step/sec: 0.134453 +INFO:tensorflow:step = 69101, loss = 0.127095, precision = 1.0 (743.754 sec) +INFO:tensorflow:global_step/sec: 0.136065 +INFO:tensorflow:step = 69201, loss = 0.126107, precision = 1.0 (734.941 sec) +Saved checkpoint after 177 epoch(s) to data/resnet164/checkpoints/00177... +INFO:tensorflow:global_step/sec: 0.137257 +INFO:tensorflow:step = 69301, loss = 0.125633, precision = 1.0 (728.562 sec) +INFO:tensorflow:global_step/sec: 0.136084 +INFO:tensorflow:step = 69401, loss = 0.123969, precision = 1.0 (734.840 sec) +INFO:tensorflow:global_step/sec: 0.131025 +INFO:tensorflow:step = 69501, loss = 0.124902, precision = 1.0 (763.215 sec) +Saved checkpoint after 178 epoch(s) to data/resnet164/checkpoints/00178... +INFO:tensorflow:global_step/sec: 0.131994 +INFO:tensorflow:step = 69601, loss = 0.125795, precision = 1.0 (757.609 sec) +INFO:tensorflow:global_step/sec: 0.137432 +INFO:tensorflow:step = 69701, loss = 0.123321, precision = 1.0 (727.630 sec) +INFO:tensorflow:global_step/sec: 0.136239 +INFO:tensorflow:step = 69801, loss = 0.123202, precision = 1.0 (734.006 sec) +INFO:tensorflow:global_step/sec: 0.135798 +INFO:tensorflow:step = 69901, loss = 0.123582, precision = 1.0 (736.387 sec) +Saved checkpoint after 179 epoch(s) to data/resnet164/checkpoints/00179... +INFO:tensorflow:global_step/sec: 0.136465 +INFO:tensorflow:step = 70001, loss = 0.124344, precision = 1.0 (732.790 sec) +INFO:tensorflow:global_step/sec: 0.133937 +INFO:tensorflow:step = 70101, loss = 0.123373, precision = 1.0 (746.618 sec) +INFO:tensorflow:global_step/sec: 0.130178 +INFO:tensorflow:step = 70201, loss = 0.122346, precision = 1.0 (768.179 sec) +INFO:tensorflow:global_step/sec: 0.133952 +INFO:tensorflow:step = 70301, loss = 0.122109, precision = 1.0 (746.536 sec) +Saved checkpoint after 180 epoch(s) to data/resnet164/checkpoints/00180... +INFO:tensorflow:global_step/sec: 0.134832 +INFO:tensorflow:step = 70401, loss = 0.122864, precision = 1.0 (741.664 sec) +INFO:tensorflow:global_step/sec: 0.134165 +INFO:tensorflow:step = 70501, loss = 0.125385, precision = 1.0 (745.353 sec) +INFO:tensorflow:global_step/sec: 0.133657 +INFO:tensorflow:step = 70601, loss = 0.123316, precision = 1.0 (748.181 sec) +INFO:tensorflow:global_step/sec: 0.133193 +INFO:tensorflow:step = 70701, loss = 0.123494, precision = 1.0 (750.788 sec) +Saved checkpoint after 181 epoch(s) to data/resnet164/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet164_nb_train.log b/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet164_nb_train.log new file mode 100644 index 0000000..cc135c8 --- /dev/null +++ b/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet164_nb_train.log @@ -0,0 +1,2145 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/2.60m params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_10/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/sub2/bn2/beta (16, 16/16 params) + unit_1_10/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_11/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/sub2/bn2/beta (16, 16/16 params) + unit_1_11/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_12/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/sub2/bn2/beta (16, 16/16 params) + unit_1_12/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_13/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/sub2/bn2/beta (16, 16/16 params) + unit_1_13/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_14/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/sub2/bn2/beta (16, 16/16 params) + unit_1_14/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_15/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/sub2/bn2/beta (16, 16/16 params) + unit_1_15/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_16/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/sub2/bn2/beta (16, 16/16 params) + unit_1_16/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_17/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/sub2/bn2/beta (16, 16/16 params) + unit_1_17/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_18/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_18/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_18/sub2/bn2/beta (16, 16/16 params) + unit_1_18/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_19/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_19/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_19/sub2/bn2/beta (16, 16/16 params) + unit_1_19/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_20/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_20/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_20/sub2/bn2/beta (16, 16/16 params) + unit_1_20/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_21/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_21/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_21/sub2/bn2/beta (16, 16/16 params) + unit_1_21/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_22/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_22/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_22/sub2/bn2/beta (16, 16/16 params) + unit_1_22/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_23/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_23/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_23/sub2/bn2/beta (16, 16/16 params) + unit_1_23/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_24/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_24/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_24/sub2/bn2/beta (16, 16/16 params) + unit_1_24/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_25/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_25/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_25/sub2/bn2/beta (16, 16/16 params) + unit_1_25/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_26/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_26/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_26/sub2/bn2/beta (16, 16/16 params) + unit_1_26/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_3/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_4/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_5/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_6/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_7/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_8/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_9/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/sub2/bn2/beta (16, 16/16 params) + unit_1_9/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_10/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/sub2/bn2/beta (32, 32/32 params) + unit_2_10/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_11/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/sub2/bn2/beta (32, 32/32 params) + unit_2_11/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_12/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/sub2/bn2/beta (32, 32/32 params) + unit_2_12/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_13/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/sub2/bn2/beta (32, 32/32 params) + unit_2_13/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_14/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/sub2/bn2/beta (32, 32/32 params) + unit_2_14/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_15/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/sub2/bn2/beta (32, 32/32 params) + unit_2_15/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_16/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/sub2/bn2/beta (32, 32/32 params) + unit_2_16/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_17/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/sub2/bn2/beta (32, 32/32 params) + unit_2_17/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_18/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_18/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_18/sub2/bn2/beta (32, 32/32 params) + unit_2_18/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_19/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_19/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_19/sub2/bn2/beta (32, 32/32 params) + unit_2_19/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_20/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_20/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_20/sub2/bn2/beta (32, 32/32 params) + unit_2_20/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_21/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_21/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_21/sub2/bn2/beta (32, 32/32 params) + unit_2_21/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_22/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_22/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_22/sub2/bn2/beta (32, 32/32 params) + unit_2_22/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_23/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_23/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_23/sub2/bn2/beta (32, 32/32 params) + unit_2_23/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_24/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_24/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_24/sub2/bn2/beta (32, 32/32 params) + unit_2_24/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_25/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_25/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_25/sub2/bn2/beta (32, 32/32 params) + unit_2_25/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_26/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_26/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_26/sub2/bn2/beta (32, 32/32 params) + unit_2_26/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_3/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_4/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_5/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_6/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_7/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_8/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_9/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/sub2/bn2/beta (32, 32/32 params) + unit_2_9/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_10/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/sub2/bn2/beta (64, 64/64 params) + unit_3_10/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_11/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/sub2/bn2/beta (64, 64/64 params) + unit_3_11/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_12/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/sub2/bn2/beta (64, 64/64 params) + unit_3_12/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_13/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/sub2/bn2/beta (64, 64/64 params) + unit_3_13/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_14/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/sub2/bn2/beta (64, 64/64 params) + unit_3_14/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_15/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/sub2/bn2/beta (64, 64/64 params) + unit_3_15/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_16/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/sub2/bn2/beta (64, 64/64 params) + unit_3_16/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_17/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/sub2/bn2/beta (64, 64/64 params) + unit_3_17/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_18/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_18/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_18/sub2/bn2/beta (64, 64/64 params) + unit_3_18/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_19/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_19/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_19/sub2/bn2/beta (64, 64/64 params) + unit_3_19/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_20/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_20/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_20/sub2/bn2/beta (64, 64/64 params) + unit_3_20/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_21/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_21/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_21/sub2/bn2/beta (64, 64/64 params) + unit_3_21/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_22/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_22/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_22/sub2/bn2/beta (64, 64/64 params) + unit_3_22/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_23/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_23/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_23/sub2/bn2/beta (64, 64/64 params) + unit_3_23/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_24/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_24/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_24/sub2/bn2/beta (64, 64/64 params) + unit_3_24/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_25/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_25/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_25/sub2/bn2/beta (64, 64/64 params) + unit_3_25/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_26/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_26/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_26/sub2/bn2/beta (64, 64/64 params) + unit_3_26/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_3/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_4/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_5/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_6/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_7/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_8/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_9/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/sub2/bn2/beta (64, 64/64 params) + unit_3_9/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/97.35b flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + +======================End of Report========================== +2017-08-01 06:02:14.159202: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 16 visible devices +2017-08-01 06:02:14.164502: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x75ad130 executing computations on platform Host. Devices: +2017-08-01 06:02:14.164590: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +INFO:tensorflow:step = 1, loss = 5.35604, precision = 0.125 +INFO:tensorflow:global_step/sec: 0.17848 +INFO:tensorflow:step = 101, loss = 4.67859, precision = 0.351562 (560.288 sec) +INFO:tensorflow:global_step/sec: 0.183571 +INFO:tensorflow:step = 201, loss = 4.53404, precision = 0.421875 (544.749 sec) +INFO:tensorflow:global_step/sec: 0.179337 +INFO:tensorflow:step = 301, loss = 4.28766, precision = 0.515625 (557.610 sec) +total_params: 2596842 +Saved checkpoint after 1 epoch(s) to data/resnet164/checkpoints/00001... +INFO:tensorflow:global_step/sec: 0.180509 +INFO:tensorflow:step = 401, loss = 4.6556, precision = 0.335938 (553.988 sec) +INFO:tensorflow:global_step/sec: 0.181872 +INFO:tensorflow:step = 501, loss = 3.94621, precision = 0.539062 (549.838 sec) +INFO:tensorflow:global_step/sec: 0.188116 +INFO:tensorflow:step = 601, loss = 3.62083, precision = 0.546875 (531.587 sec) +INFO:tensorflow:global_step/sec: 0.191816 +INFO:tensorflow:step = 701, loss = 3.17487, precision = 0.632812 (521.332 sec) +Saved checkpoint after 2 epoch(s) to data/resnet164/checkpoints/00002... +INFO:tensorflow:global_step/sec: 0.191288 +INFO:tensorflow:step = 801, loss = 2.97458, precision = 0.648438 (522.772 sec) +INFO:tensorflow:global_step/sec: 0.191805 +INFO:tensorflow:step = 901, loss = 2.8601, precision = 0.648438 (521.364 sec) +INFO:tensorflow:global_step/sec: 0.190528 +INFO:tensorflow:step = 1001, loss = 2.50224, precision = 0.726562 (524.854 sec) +INFO:tensorflow:global_step/sec: 0.192674 +INFO:tensorflow:step = 1101, loss = 2.36066, precision = 0.6875 (519.012 sec) +Saved checkpoint after 3 epoch(s) to data/resnet164/checkpoints/00003... +INFO:tensorflow:global_step/sec: 0.198541 +INFO:tensorflow:step = 1201, loss = 2.23766, precision = 0.773438 (503.673 sec) +INFO:tensorflow:global_step/sec: 0.200702 +INFO:tensorflow:step = 1301, loss = 1.90419, precision = 0.8125 (498.251 sec) +INFO:tensorflow:global_step/sec: 0.199762 +INFO:tensorflow:step = 1401, loss = 1.84054, precision = 0.78125 (500.597 sec) +INFO:tensorflow:global_step/sec: 0.19916 +INFO:tensorflow:step = 1501, loss = 1.77711, precision = 0.78125 (502.109 sec) +Saved checkpoint after 4 epoch(s) to data/resnet164/checkpoints/00004... +INFO:tensorflow:global_step/sec: 0.19733 +INFO:tensorflow:step = 1601, loss = 1.78155, precision = 0.773438 (506.764 sec) +INFO:tensorflow:global_step/sec: 0.201359 +INFO:tensorflow:step = 1701, loss = 1.71432, precision = 0.742188 (496.626 sec) +INFO:tensorflow:global_step/sec: 0.198715 +INFO:tensorflow:step = 1801, loss = 1.43287, precision = 0.851562 (503.233 sec) +INFO:tensorflow:global_step/sec: 0.200201 +INFO:tensorflow:step = 1901, loss = 1.57339, precision = 0.71875 (499.498 sec) +Saved checkpoint after 5 epoch(s) to data/resnet164/checkpoints/00005... +INFO:tensorflow:global_step/sec: 0.197513 +INFO:tensorflow:step = 2001, loss = 1.36344, precision = 0.804688 (506.295 sec) +INFO:tensorflow:global_step/sec: 0.200802 +INFO:tensorflow:step = 2101, loss = 1.376, precision = 0.789062 (498.002 sec) +INFO:tensorflow:global_step/sec: 0.200892 +INFO:tensorflow:step = 2201, loss = 1.23549, precision = 0.820312 (497.781 sec) +INFO:tensorflow:global_step/sec: 0.194727 +INFO:tensorflow:step = 2301, loss = 1.36974, precision = 0.757812 (513.540 sec) +Saved checkpoint after 6 epoch(s) to data/resnet164/checkpoints/00006... +INFO:tensorflow:global_step/sec: 0.194755 +INFO:tensorflow:step = 2401, loss = 1.14844, precision = 0.796875 (513.467 sec) +INFO:tensorflow:global_step/sec: 0.193709 +INFO:tensorflow:step = 2501, loss = 1.2832, precision = 0.796875 (516.239 sec) +INFO:tensorflow:global_step/sec: 0.197396 +INFO:tensorflow:step = 2601, loss = 1.03119, precision = 0.859375 (506.595 sec) +INFO:tensorflow:global_step/sec: 0.193272 +INFO:tensorflow:step = 2701, loss = 1.02643, precision = 0.835938 (517.406 sec) +Saved checkpoint after 7 epoch(s) to data/resnet164/checkpoints/00007... +INFO:tensorflow:global_step/sec: 0.192971 +INFO:tensorflow:step = 2801, loss = 0.959358, precision = 0.828125 (518.213 sec) +INFO:tensorflow:global_step/sec: 0.193276 +INFO:tensorflow:step = 2901, loss = 1.01918, precision = 0.820312 (517.394 sec) +INFO:tensorflow:global_step/sec: 0.193532 +INFO:tensorflow:step = 3001, loss = 0.976689, precision = 0.8125 (516.711 sec) +INFO:tensorflow:global_step/sec: 0.19098 +INFO:tensorflow:step = 3101, loss = 1.02437, precision = 0.804688 (523.614 sec) +Saved checkpoint after 8 epoch(s) to data/resnet164/checkpoints/00008... +INFO:tensorflow:global_step/sec: 0.191273 +INFO:tensorflow:step = 3201, loss = 1.08514, precision = 0.757812 (522.812 sec) +INFO:tensorflow:global_step/sec: 0.192541 +INFO:tensorflow:step = 3301, loss = 1.07214, precision = 0.773438 (519.369 sec) +INFO:tensorflow:global_step/sec: 0.192909 +INFO:tensorflow:step = 3401, loss = 1.02026, precision = 0.804688 (518.380 sec) +INFO:tensorflow:global_step/sec: 0.1949 +INFO:tensorflow:step = 3501, loss = 0.823547, precision = 0.835938 (513.083 sec) +Saved checkpoint after 9 epoch(s) to data/resnet164/checkpoints/00009... +INFO:tensorflow:global_step/sec: 0.19486 +INFO:tensorflow:step = 3601, loss = 0.786214, precision = 0.875 (513.189 sec) +INFO:tensorflow:global_step/sec: 0.192811 +INFO:tensorflow:step = 3701, loss = 0.947634, precision = 0.78125 (518.643 sec) +INFO:tensorflow:global_step/sec: 0.195476 +INFO:tensorflow:step = 3801, loss = 0.877234, precision = 0.828125 (511.571 sec) +INFO:tensorflow:global_step/sec: 0.191029 +INFO:tensorflow:step = 3901, loss = 0.825129, precision = 0.84375 (523.481 sec) +Saved checkpoint after 10 epoch(s) to data/resnet164/checkpoints/00010... +INFO:tensorflow:global_step/sec: 0.189916 +INFO:tensorflow:step = 4001, loss = 0.810758, precision = 0.828125 (526.548 sec) +INFO:tensorflow:global_step/sec: 0.196811 +INFO:tensorflow:step = 4101, loss = 0.823833, precision = 0.820312 (508.102 sec) +INFO:tensorflow:global_step/sec: 0.198452 +INFO:tensorflow:step = 4201, loss = 0.878682, precision = 0.8125 (503.901 sec) +Saved checkpoint after 11 epoch(s) to data/resnet164/checkpoints/00011... +INFO:tensorflow:global_step/sec: 0.196923 +INFO:tensorflow:step = 4301, loss = 0.818976, precision = 0.851562 (507.812 sec) +INFO:tensorflow:global_step/sec: 0.196984 +INFO:tensorflow:step = 4401, loss = 0.897797, precision = 0.796875 (507.655 sec) +INFO:tensorflow:global_step/sec: 0.197709 +INFO:tensorflow:step = 4501, loss = 0.831542, precision = 0.820312 (505.793 sec) +INFO:tensorflow:global_step/sec: 0.19848 +INFO:tensorflow:step = 4601, loss = 0.825011, precision = 0.867188 (503.829 sec) +Saved checkpoint after 12 epoch(s) to data/resnet164/checkpoints/00012... +INFO:tensorflow:global_step/sec: 0.195553 +INFO:tensorflow:step = 4701, loss = 0.894565, precision = 0.8125 (511.369 sec) +INFO:tensorflow:global_step/sec: 0.19522 +INFO:tensorflow:step = 4801, loss = 0.713798, precision = 0.867188 (512.241 sec) +INFO:tensorflow:global_step/sec: 0.195872 +INFO:tensorflow:step = 4901, loss = 0.755067, precision = 0.828125 (510.538 sec) +INFO:tensorflow:global_step/sec: 0.198309 +INFO:tensorflow:step = 5001, loss = 0.725294, precision = 0.867188 (504.264 sec) +Saved checkpoint after 13 epoch(s) to data/resnet164/checkpoints/00013... +INFO:tensorflow:global_step/sec: 0.197006 +INFO:tensorflow:step = 5101, loss = 0.819527, precision = 0.8125 (507.600 sec) +INFO:tensorflow:global_step/sec: 0.197193 +INFO:tensorflow:step = 5201, loss = 0.670075, precision = 0.875 (507.117 sec) +INFO:tensorflow:global_step/sec: 0.195003 +INFO:tensorflow:step = 5301, loss = 0.827831, precision = 0.8125 (512.814 sec) +INFO:tensorflow:global_step/sec: 0.193787 +INFO:tensorflow:step = 5401, loss = 0.707037, precision = 0.867188 (516.031 sec) +Saved checkpoint after 14 epoch(s) to data/resnet164/checkpoints/00014... +INFO:tensorflow:global_step/sec: 0.195134 +INFO:tensorflow:step = 5501, loss = 0.788947, precision = 0.828125 (512.467 sec) +INFO:tensorflow:global_step/sec: 0.198053 +INFO:tensorflow:step = 5601, loss = 0.84989, precision = 0.796875 (504.916 sec) +INFO:tensorflow:global_step/sec: 0.198459 +INFO:tensorflow:step = 5701, loss = 0.670318, precision = 0.875 (503.881 sec) +INFO:tensorflow:global_step/sec: 0.196822 +INFO:tensorflow:step = 5801, loss = 0.818447, precision = 0.820312 (508.073 sec) +Saved checkpoint after 15 epoch(s) to data/resnet164/checkpoints/00015... +INFO:tensorflow:global_step/sec: 0.195075 +INFO:tensorflow:step = 5901, loss = 0.637272, precision = 0.882812 (512.623 sec) +INFO:tensorflow:global_step/sec: 0.194413 +INFO:tensorflow:step = 6001, loss = 0.678607, precision = 0.875 (514.369 sec) +INFO:tensorflow:global_step/sec: 0.195068 +INFO:tensorflow:step = 6101, loss = 0.573749, precision = 0.890625 (512.643 sec) +INFO:tensorflow:global_step/sec: 0.197279 +INFO:tensorflow:step = 6201, loss = 0.737918, precision = 0.84375 (506.896 sec) +Saved checkpoint after 16 epoch(s) to data/resnet164/checkpoints/00016... +INFO:tensorflow:global_step/sec: 0.194209 +INFO:tensorflow:step = 6301, loss = 0.897067, precision = 0.796875 (514.911 sec) +INFO:tensorflow:global_step/sec: 0.192368 +INFO:tensorflow:step = 6401, loss = 0.706234, precision = 0.875 (519.836 sec) +INFO:tensorflow:global_step/sec: 0.190531 +INFO:tensorflow:step = 6501, loss = 0.660721, precision = 0.875 (524.848 sec) +INFO:tensorflow:global_step/sec: 0.189381 +INFO:tensorflow:step = 6601, loss = 0.829516, precision = 0.820312 (528.036 sec) +Saved checkpoint after 17 epoch(s) to data/resnet164/checkpoints/00017... +INFO:tensorflow:global_step/sec: 0.199063 +INFO:tensorflow:step = 6701, loss = 0.74692, precision = 0.820312 (502.354 sec) +INFO:tensorflow:global_step/sec: 0.197593 +INFO:tensorflow:step = 6801, loss = 0.579496, precision = 0.882812 (506.091 sec) +INFO:tensorflow:global_step/sec: 0.191066 +INFO:tensorflow:step = 6901, loss = 0.809897, precision = 0.8125 (523.379 sec) +INFO:tensorflow:global_step/sec: 0.197094 +INFO:tensorflow:step = 7001, loss = 0.677325, precision = 0.867188 (507.371 sec) +Saved checkpoint after 18 epoch(s) to data/resnet164/checkpoints/00018... +INFO:tensorflow:global_step/sec: 0.18173 +INFO:tensorflow:step = 7101, loss = 0.747478, precision = 0.84375 (550.267 sec) +INFO:tensorflow:global_step/sec: 0.162262 +INFO:tensorflow:step = 7201, loss = 0.689672, precision = 0.875 (616.286 sec) +INFO:tensorflow:global_step/sec: 0.173363 +INFO:tensorflow:step = 7301, loss = 0.664129, precision = 0.867188 (576.824 sec) +INFO:tensorflow:global_step/sec: 0.184817 +INFO:tensorflow:step = 7401, loss = 0.608913, precision = 0.90625 (541.077 sec) +Saved checkpoint after 19 epoch(s) to data/resnet164/checkpoints/00019... +INFO:tensorflow:global_step/sec: 0.170692 +INFO:tensorflow:step = 7501, loss = 0.661742, precision = 0.828125 (585.850 sec) +INFO:tensorflow:global_step/sec: 0.180439 +INFO:tensorflow:step = 7601, loss = 0.684137, precision = 0.84375 (554.206 sec) +INFO:tensorflow:global_step/sec: 0.186824 +INFO:tensorflow:step = 7701, loss = 0.693892, precision = 0.859375 (535.264 sec) +INFO:tensorflow:global_step/sec: 0.191737 +INFO:tensorflow:step = 7801, loss = 0.647871, precision = 0.898438 (521.547 sec) +Saved checkpoint after 20 epoch(s) to data/resnet164/checkpoints/00020... +INFO:tensorflow:global_step/sec: 0.177098 +INFO:tensorflow:step = 7901, loss = 0.685492, precision = 0.867188 (564.658 sec) +INFO:tensorflow:global_step/sec: 0.177698 +INFO:tensorflow:step = 8001, loss = 0.668309, precision = 0.859375 (562.754 sec) +INFO:tensorflow:global_step/sec: 0.171701 +INFO:tensorflow:step = 8101, loss = 0.862865, precision = 0.78125 (582.408 sec) +INFO:tensorflow:global_step/sec: 0.17466 +INFO:tensorflow:step = 8201, loss = 0.62115, precision = 0.882812 (572.540 sec) +Saved checkpoint after 21 epoch(s) to data/resnet164/checkpoints/00021... +INFO:tensorflow:global_step/sec: 0.175775 +INFO:tensorflow:step = 8301, loss = 0.592151, precision = 0.898438 (568.910 sec) +INFO:tensorflow:global_step/sec: 0.176398 +INFO:tensorflow:step = 8401, loss = 0.635541, precision = 0.875 (566.898 sec) +INFO:tensorflow:global_step/sec: 0.176699 +INFO:tensorflow:step = 8501, loss = 0.848582, precision = 0.796875 (565.936 sec) +INFO:tensorflow:global_step/sec: 0.175337 +INFO:tensorflow:step = 8601, loss = 0.997624, precision = 0.78125 (570.332 sec) +Saved checkpoint after 22 epoch(s) to data/resnet164/checkpoints/00022... +INFO:tensorflow:global_step/sec: 0.176865 +INFO:tensorflow:step = 8701, loss = 0.771111, precision = 0.820312 (565.405 sec) +INFO:tensorflow:global_step/sec: 0.171548 +INFO:tensorflow:step = 8801, loss = 0.705819, precision = 0.867188 (582.927 sec) +INFO:tensorflow:global_step/sec: 0.175299 +INFO:tensorflow:step = 8901, loss = 0.648098, precision = 0.90625 (570.454 sec) +Saved checkpoint after 23 epoch(s) to data/resnet164/checkpoints/00023... +INFO:tensorflow:global_step/sec: 0.176771 +INFO:tensorflow:step = 9001, loss = 0.700524, precision = 0.84375 (565.704 sec) +INFO:tensorflow:global_step/sec: 0.176232 +INFO:tensorflow:step = 9101, loss = 0.704139, precision = 0.84375 (567.433 sec) +INFO:tensorflow:global_step/sec: 0.174531 +INFO:tensorflow:step = 9201, loss = 0.607209, precision = 0.898438 (572.965 sec) +INFO:tensorflow:global_step/sec: 0.170763 +INFO:tensorflow:step = 9301, loss = 0.65637, precision = 0.859375 (585.606 sec) +Saved checkpoint after 24 epoch(s) to data/resnet164/checkpoints/00024... +INFO:tensorflow:global_step/sec: 0.169988 +INFO:tensorflow:step = 9401, loss = 0.568754, precision = 0.90625 (588.275 sec) +INFO:tensorflow:global_step/sec: 0.176825 +INFO:tensorflow:step = 9501, loss = 0.743587, precision = 0.851562 (565.531 sec) +INFO:tensorflow:global_step/sec: 0.177914 +INFO:tensorflow:step = 9601, loss = 0.496695, precision = 0.945312 (562.068 sec) +INFO:tensorflow:global_step/sec: 0.185921 +INFO:tensorflow:step = 9701, loss = 0.851891, precision = 0.789062 (537.861 sec) +Saved checkpoint after 25 epoch(s) to data/resnet164/checkpoints/00025... +INFO:tensorflow:global_step/sec: 0.179773 +INFO:tensorflow:step = 9801, loss = 0.783082, precision = 0.828125 (556.256 sec) +INFO:tensorflow:global_step/sec: 0.175458 +INFO:tensorflow:step = 9901, loss = 0.661031, precision = 0.882812 (569.937 sec) +INFO:tensorflow:global_step/sec: 0.1777 +INFO:tensorflow:step = 10001, loss = 0.644406, precision = 0.875 (562.747 sec) +INFO:tensorflow:global_step/sec: 0.184049 +INFO:tensorflow:step = 10101, loss = 0.679022, precision = 0.875 (543.335 sec) +Saved checkpoint after 26 epoch(s) to data/resnet164/checkpoints/00026... +INFO:tensorflow:global_step/sec: 0.171152 +INFO:tensorflow:step = 10201, loss = 0.725568, precision = 0.859375 (584.276 sec) +INFO:tensorflow:global_step/sec: 0.17249 +INFO:tensorflow:step = 10301, loss = 0.75512, precision = 0.804688 (579.743 sec) +INFO:tensorflow:global_step/sec: 0.173839 +INFO:tensorflow:step = 10401, loss = 0.630701, precision = 0.875 (575.245 sec) +INFO:tensorflow:global_step/sec: 0.170317 +INFO:tensorflow:step = 10501, loss = 0.606312, precision = 0.90625 (587.140 sec) +Saved checkpoint after 27 epoch(s) to data/resnet164/checkpoints/00027... +INFO:tensorflow:global_step/sec: 0.174024 +INFO:tensorflow:step = 10601, loss = 0.695077, precision = 0.851562 (574.633 sec) +INFO:tensorflow:global_step/sec: 0.176685 +INFO:tensorflow:step = 10701, loss = 0.771429, precision = 0.835938 (565.980 sec) +INFO:tensorflow:global_step/sec: 0.174083 +INFO:tensorflow:step = 10801, loss = 0.719295, precision = 0.859375 (574.439 sec) +INFO:tensorflow:global_step/sec: 0.173628 +INFO:tensorflow:step = 10901, loss = 0.623673, precision = 0.914062 (575.946 sec) +Saved checkpoint after 28 epoch(s) to data/resnet164/checkpoints/00028... +INFO:tensorflow:global_step/sec: 0.178235 +INFO:tensorflow:step = 11001, loss = 0.758532, precision = 0.851562 (561.058 sec) +INFO:tensorflow:global_step/sec: 0.175609 +INFO:tensorflow:step = 11101, loss = 0.718379, precision = 0.851562 (569.445 sec) +INFO:tensorflow:global_step/sec: 0.181429 +INFO:tensorflow:step = 11201, loss = 0.598584, precision = 0.90625 (551.181 sec) +INFO:tensorflow:global_step/sec: 0.182731 +INFO:tensorflow:step = 11301, loss = 0.824421, precision = 0.820312 (547.251 sec) +Saved checkpoint after 29 epoch(s) to data/resnet164/checkpoints/00029... +INFO:tensorflow:global_step/sec: 0.181548 +INFO:tensorflow:step = 11401, loss = 0.759378, precision = 0.84375 (550.819 sec) +INFO:tensorflow:global_step/sec: 0.180787 +INFO:tensorflow:step = 11501, loss = 0.556471, precision = 0.90625 (553.138 sec) +INFO:tensorflow:global_step/sec: 0.179219 +INFO:tensorflow:step = 11601, loss = 0.886335, precision = 0.757812 (557.976 sec) +INFO:tensorflow:global_step/sec: 0.173393 +INFO:tensorflow:step = 11701, loss = 0.739943, precision = 0.851562 (576.724 sec) +Saved checkpoint after 30 epoch(s) to data/resnet164/checkpoints/00030... +INFO:tensorflow:global_step/sec: 0.177435 +INFO:tensorflow:step = 11801, loss = 0.645257, precision = 0.882812 (563.586 sec) +INFO:tensorflow:global_step/sec: 0.191676 +INFO:tensorflow:step = 11901, loss = 0.703388, precision = 0.859375 (521.713 sec) +INFO:tensorflow:global_step/sec: 0.189384 +INFO:tensorflow:step = 12001, loss = 0.656425, precision = 0.890625 (528.027 sec) +INFO:tensorflow:global_step/sec: 0.189252 +INFO:tensorflow:step = 12101, loss = 0.644028, precision = 0.882812 (528.395 sec) +Saved checkpoint after 31 epoch(s) to data/resnet164/checkpoints/00031... +INFO:tensorflow:global_step/sec: 0.188294 +INFO:tensorflow:step = 12201, loss = 0.771068, precision = 0.84375 (531.084 sec) +INFO:tensorflow:global_step/sec: 0.192075 +INFO:tensorflow:step = 12301, loss = 0.774135, precision = 0.835938 (520.629 sec) +INFO:tensorflow:global_step/sec: 0.186608 +INFO:tensorflow:step = 12401, loss = 0.833949, precision = 0.828125 (535.884 sec) +INFO:tensorflow:global_step/sec: 0.184398 +INFO:tensorflow:step = 12501, loss = 0.75541, precision = 0.828125 (542.305 sec) +Saved checkpoint after 32 epoch(s) to data/resnet164/checkpoints/00032... +INFO:tensorflow:global_step/sec: 0.19117 +INFO:tensorflow:step = 12601, loss = 0.72205, precision = 0.835938 (523.093 sec) +INFO:tensorflow:global_step/sec: 0.191842 +INFO:tensorflow:step = 12701, loss = 0.714484, precision = 0.867188 (521.262 sec) +INFO:tensorflow:global_step/sec: 0.194918 +INFO:tensorflow:step = 12801, loss = 0.646285, precision = 0.875 (513.036 sec) +INFO:tensorflow:global_step/sec: 0.192846 +INFO:tensorflow:step = 12901, loss = 0.67915, precision = 0.859375 (518.549 sec) +Saved checkpoint after 33 epoch(s) to data/resnet164/checkpoints/00033... +INFO:tensorflow:global_step/sec: 0.188429 +INFO:tensorflow:step = 13001, loss = 0.660963, precision = 0.867188 (530.704 sec) +INFO:tensorflow:global_step/sec: 0.194318 +INFO:tensorflow:step = 13101, loss = 0.711504, precision = 0.867188 (514.621 sec) +INFO:tensorflow:global_step/sec: 0.18141 +INFO:tensorflow:step = 13201, loss = 0.670327, precision = 0.898438 (551.237 sec) +Saved checkpoint after 34 epoch(s) to data/resnet164/checkpoints/00034... +INFO:tensorflow:global_step/sec: 0.182596 +INFO:tensorflow:step = 13301, loss = 0.672101, precision = 0.890625 (547.657 sec) +INFO:tensorflow:global_step/sec: 0.187152 +INFO:tensorflow:step = 13401, loss = 0.801785, precision = 0.804688 (534.324 sec) +INFO:tensorflow:global_step/sec: 0.18414 +INFO:tensorflow:step = 13501, loss = 0.598704, precision = 0.90625 (543.065 sec) +INFO:tensorflow:global_step/sec: 0.186072 +INFO:tensorflow:step = 13601, loss = 0.568188, precision = 0.921875 (537.425 sec) +Saved checkpoint after 35 epoch(s) to data/resnet164/checkpoints/00035... +INFO:tensorflow:global_step/sec: 0.187075 +INFO:tensorflow:step = 13701, loss = 0.561004, precision = 0.90625 (534.546 sec) +INFO:tensorflow:global_step/sec: 0.198304 +INFO:tensorflow:step = 13801, loss = 0.601045, precision = 0.859375 (504.275 sec) +INFO:tensorflow:global_step/sec: 0.194064 +INFO:tensorflow:step = 13901, loss = 0.685094, precision = 0.859375 (515.293 sec) +INFO:tensorflow:global_step/sec: 0.196066 +INFO:tensorflow:step = 14001, loss = 0.64617, precision = 0.90625 (510.032 sec) +Saved checkpoint after 36 epoch(s) to data/resnet164/checkpoints/00036... +INFO:tensorflow:global_step/sec: 0.202156 +INFO:tensorflow:step = 14101, loss = 0.601647, precision = 0.890625 (494.668 sec) +INFO:tensorflow:global_step/sec: 0.200695 +INFO:tensorflow:step = 14201, loss = 0.601021, precision = 0.898438 (498.268 sec) +INFO:tensorflow:global_step/sec: 0.202149 +INFO:tensorflow:step = 14301, loss = 0.523669, precision = 0.921875 (494.685 sec) +INFO:tensorflow:global_step/sec: 0.202512 +INFO:tensorflow:step = 14401, loss = 0.540635, precision = 0.929688 (493.797 sec) +Saved checkpoint after 37 epoch(s) to data/resnet164/checkpoints/00037... +INFO:tensorflow:global_step/sec: 0.198844 +INFO:tensorflow:step = 14501, loss = 0.598075, precision = 0.914062 (502.906 sec) +INFO:tensorflow:global_step/sec: 0.199105 +INFO:tensorflow:step = 14601, loss = 0.736154, precision = 0.84375 (502.247 sec) +INFO:tensorflow:global_step/sec: 0.196832 +INFO:tensorflow:step = 14701, loss = 0.602998, precision = 0.890625 (508.048 sec) +INFO:tensorflow:global_step/sec: 0.19013 +INFO:tensorflow:step = 14801, loss = 0.811775, precision = 0.875 (525.956 sec) +Saved checkpoint after 38 epoch(s) to data/resnet164/checkpoints/00038... +INFO:tensorflow:global_step/sec: 0.188634 +INFO:tensorflow:step = 14901, loss = 0.657464, precision = 0.898438 (530.126 sec) +INFO:tensorflow:global_step/sec: 0.190207 +INFO:tensorflow:step = 15001, loss = 0.73313, precision = 0.867188 (525.744 sec) +INFO:tensorflow:global_step/sec: 0.192724 +INFO:tensorflow:step = 15101, loss = 0.698175, precision = 0.859375 (518.877 sec) +INFO:tensorflow:global_step/sec: 0.188333 +INFO:tensorflow:step = 15201, loss = 0.676217, precision = 0.859375 (530.975 sec) +Saved checkpoint after 39 epoch(s) to data/resnet164/checkpoints/00039... +INFO:tensorflow:global_step/sec: 0.202414 +INFO:tensorflow:step = 15301, loss = 0.566839, precision = 0.929688 (494.037 sec) +INFO:tensorflow:global_step/sec: 0.203897 +INFO:tensorflow:step = 15401, loss = 0.737164, precision = 0.84375 (490.443 sec) +INFO:tensorflow:global_step/sec: 0.203992 +INFO:tensorflow:step = 15501, loss = 0.692471, precision = 0.867188 (490.216 sec) +INFO:tensorflow:global_step/sec: 0.205254 +INFO:tensorflow:step = 15601, loss = 0.651677, precision = 0.875 (487.200 sec) +Saved checkpoint after 40 epoch(s) to data/resnet164/checkpoints/00040... +INFO:tensorflow:global_step/sec: 0.205929 +INFO:tensorflow:step = 15701, loss = 0.642259, precision = 0.890625 (485.603 sec) +INFO:tensorflow:global_step/sec: 0.205521 +INFO:tensorflow:step = 15801, loss = 0.764301, precision = 0.820312 (486.569 sec) +INFO:tensorflow:global_step/sec: 0.195047 +INFO:tensorflow:step = 15901, loss = 0.645264, precision = 0.882812 (512.696 sec) +INFO:tensorflow:global_step/sec: 0.199903 +INFO:tensorflow:step = 16001, loss = 0.538016, precision = 0.9375 (500.243 sec) +Saved checkpoint after 41 epoch(s) to data/resnet164/checkpoints/00041... +INFO:tensorflow:global_step/sec: 0.194644 +INFO:tensorflow:step = 16101, loss = 0.681326, precision = 0.875 (513.760 sec) +INFO:tensorflow:global_step/sec: 0.195575 +INFO:tensorflow:step = 16201, loss = 0.659171, precision = 0.867188 (511.314 sec) +INFO:tensorflow:global_step/sec: 0.197283 +INFO:tensorflow:step = 16301, loss = 0.619946, precision = 0.882812 (506.885 sec) +INFO:tensorflow:global_step/sec: 0.195901 +INFO:tensorflow:step = 16401, loss = 0.708509, precision = 0.875 (510.462 sec) +Saved checkpoint after 42 epoch(s) to data/resnet164/checkpoints/00042... +INFO:tensorflow:global_step/sec: 0.199569 +INFO:tensorflow:step = 16501, loss = 0.619853, precision = 0.90625 (501.079 sec) +INFO:tensorflow:global_step/sec: 0.192985 +INFO:tensorflow:step = 16601, loss = 0.762388, precision = 0.851562 (518.175 sec) +INFO:tensorflow:global_step/sec: 0.195857 +INFO:tensorflow:step = 16701, loss = 0.753461, precision = 0.835938 (510.578 sec) +INFO:tensorflow:global_step/sec: 0.202478 +INFO:tensorflow:step = 16801, loss = 0.802803, precision = 0.820312 (493.882 sec) +Saved checkpoint after 43 epoch(s) to data/resnet164/checkpoints/00043... +INFO:tensorflow:global_step/sec: 0.199868 +INFO:tensorflow:step = 16901, loss = 0.715555, precision = 0.828125 (500.330 sec) +INFO:tensorflow:global_step/sec: 0.205064 +INFO:tensorflow:step = 17001, loss = 0.76915, precision = 0.835938 (487.653 sec) +INFO:tensorflow:global_step/sec: 0.207247 +INFO:tensorflow:step = 17101, loss = 0.669807, precision = 0.84375 (482.516 sec) +INFO:tensorflow:global_step/sec: 0.205939 +INFO:tensorflow:step = 17201, loss = 0.743939, precision = 0.859375 (485.582 sec) +Saved checkpoint after 44 epoch(s) to data/resnet164/checkpoints/00044... +INFO:tensorflow:global_step/sec: 0.206547 +INFO:tensorflow:step = 17301, loss = 0.737821, precision = 0.84375 (484.151 sec) +INFO:tensorflow:global_step/sec: 0.206308 +INFO:tensorflow:step = 17401, loss = 0.698151, precision = 0.875 (484.711 sec) +INFO:tensorflow:global_step/sec: 0.20006 +INFO:tensorflow:step = 17501, loss = 0.604448, precision = 0.898438 (499.851 sec) +Saved checkpoint after 45 epoch(s) to data/resnet164/checkpoints/00045... +INFO:tensorflow:global_step/sec: 0.197834 +INFO:tensorflow:step = 17601, loss = 0.559886, precision = 0.914062 (505.476 sec) +INFO:tensorflow:global_step/sec: 0.208447 +INFO:tensorflow:step = 17701, loss = 0.751739, precision = 0.851562 (479.737 sec) +INFO:tensorflow:global_step/sec: 0.207433 +INFO:tensorflow:step = 17801, loss = 0.501656, precision = 0.945312 (482.084 sec) +INFO:tensorflow:global_step/sec: 0.205432 +INFO:tensorflow:step = 17901, loss = 0.73049, precision = 0.875 (486.779 sec) +Saved checkpoint after 46 epoch(s) to data/resnet164/checkpoints/00046... +INFO:tensorflow:global_step/sec: 0.204093 +INFO:tensorflow:step = 18001, loss = 0.498915, precision = 0.914062 (489.973 sec) +INFO:tensorflow:global_step/sec: 0.200853 +INFO:tensorflow:step = 18101, loss = 0.862912, precision = 0.796875 (497.876 sec) +INFO:tensorflow:global_step/sec: 0.203435 +INFO:tensorflow:step = 18201, loss = 0.699058, precision = 0.835938 (491.558 sec) +INFO:tensorflow:global_step/sec: 0.204303 +INFO:tensorflow:step = 18301, loss = 0.603139, precision = 0.90625 (489.469 sec) +Saved checkpoint after 47 epoch(s) to data/resnet164/checkpoints/00047... +INFO:tensorflow:global_step/sec: 0.201798 +INFO:tensorflow:step = 18401, loss = 0.643236, precision = 0.851562 (495.545 sec) +INFO:tensorflow:global_step/sec: 0.202579 +INFO:tensorflow:step = 18501, loss = 0.680994, precision = 0.882812 (493.634 sec) +INFO:tensorflow:global_step/sec: 0.207647 +INFO:tensorflow:step = 18601, loss = 0.860807, precision = 0.84375 (481.586 sec) +INFO:tensorflow:global_step/sec: 0.210772 +INFO:tensorflow:step = 18701, loss = 0.709771, precision = 0.859375 (474.447 sec) +Saved checkpoint after 48 epoch(s) to data/resnet164/checkpoints/00048... +INFO:tensorflow:global_step/sec: 0.204803 +INFO:tensorflow:step = 18801, loss = 0.626195, precision = 0.890625 (488.273 sec) +INFO:tensorflow:global_step/sec: 0.202784 +INFO:tensorflow:step = 18901, loss = 0.61679, precision = 0.898438 (493.135 sec) +INFO:tensorflow:global_step/sec: 0.203294 +INFO:tensorflow:step = 19001, loss = 0.74275, precision = 0.835938 (491.897 sec) +INFO:tensorflow:global_step/sec: 0.204038 +INFO:tensorflow:step = 19101, loss = 0.61184, precision = 0.914062 (490.105 sec) +Saved checkpoint after 49 epoch(s) to data/resnet164/checkpoints/00049... +INFO:tensorflow:global_step/sec: 0.208319 +INFO:tensorflow:step = 19201, loss = 0.605354, precision = 0.898438 (480.033 sec) +INFO:tensorflow:global_step/sec: 0.207086 +INFO:tensorflow:step = 19301, loss = 0.77637, precision = 0.820312 (482.892 sec) +INFO:tensorflow:global_step/sec: 0.20719 +INFO:tensorflow:step = 19401, loss = 0.67037, precision = 0.882812 (482.648 sec) +INFO:tensorflow:global_step/sec: 0.205918 +INFO:tensorflow:step = 19501, loss = 0.742221, precision = 0.867188 (485.630 sec) +Saved checkpoint after 50 epoch(s) to data/resnet164/checkpoints/00050... +INFO:tensorflow:global_step/sec: 0.200339 +INFO:tensorflow:step = 19601, loss = 0.649998, precision = 0.90625 (499.155 sec) +INFO:tensorflow:global_step/sec: 0.205153 +INFO:tensorflow:step = 19701, loss = 0.669714, precision = 0.867188 (487.440 sec) +INFO:tensorflow:global_step/sec: 0.201974 +INFO:tensorflow:step = 19801, loss = 0.595641, precision = 0.890625 (495.113 sec) +INFO:tensorflow:global_step/sec: 0.206279 +INFO:tensorflow:step = 19901, loss = 0.617033, precision = 0.90625 (484.780 sec) +Saved checkpoint after 51 epoch(s) to data/resnet164/checkpoints/00051... +INFO:tensorflow:global_step/sec: 0.20302 +INFO:tensorflow:step = 20001, loss = 0.684475, precision = 0.882812 (492.563 sec) +INFO:tensorflow:global_step/sec: 0.2057 +INFO:tensorflow:step = 20101, loss = 0.67665, precision = 0.867188 (486.146 sec) +INFO:tensorflow:global_step/sec: 0.207648 +INFO:tensorflow:step = 20201, loss = 0.701941, precision = 0.859375 (481.585 sec) +INFO:tensorflow:global_step/sec: 0.205486 +INFO:tensorflow:step = 20301, loss = 0.697956, precision = 0.875 (486.650 sec) +Saved checkpoint after 52 epoch(s) to data/resnet164/checkpoints/00052... +INFO:tensorflow:global_step/sec: 0.204916 +INFO:tensorflow:step = 20401, loss = 0.635937, precision = 0.898438 (488.005 sec) +INFO:tensorflow:global_step/sec: 0.193016 +INFO:tensorflow:step = 20501, loss = 0.70637, precision = 0.84375 (518.092 sec) +INFO:tensorflow:global_step/sec: 0.195116 +INFO:tensorflow:step = 20601, loss = 0.60955, precision = 0.882812 (512.515 sec) +INFO:tensorflow:global_step/sec: 0.194312 +INFO:tensorflow:step = 20701, loss = 0.685261, precision = 0.867188 (514.635 sec) +Saved checkpoint after 53 epoch(s) to data/resnet164/checkpoints/00053... +INFO:tensorflow:global_step/sec: 0.195731 +INFO:tensorflow:step = 20801, loss = 0.55689, precision = 0.929688 (510.906 sec) +INFO:tensorflow:global_step/sec: 0.198553 +INFO:tensorflow:step = 20901, loss = 0.669336, precision = 0.859375 (503.642 sec) +INFO:tensorflow:global_step/sec: 0.20032 +INFO:tensorflow:step = 21001, loss = 0.647151, precision = 0.875 (499.200 sec) +INFO:tensorflow:global_step/sec: 0.202368 +INFO:tensorflow:step = 21101, loss = 0.654005, precision = 0.90625 (494.149 sec) +Saved checkpoint after 54 epoch(s) to data/resnet164/checkpoints/00054... +INFO:tensorflow:global_step/sec: 0.202316 +INFO:tensorflow:step = 21201, loss = 0.629503, precision = 0.882812 (494.277 sec) +INFO:tensorflow:global_step/sec: 0.20352 +INFO:tensorflow:step = 21301, loss = 0.553433, precision = 0.9375 (491.351 sec) +INFO:tensorflow:global_step/sec: 0.206537 +INFO:tensorflow:step = 21401, loss = 0.760917, precision = 0.828125 (484.175 sec) +INFO:tensorflow:global_step/sec: 0.203292 +INFO:tensorflow:step = 21501, loss = 0.638365, precision = 0.914062 (491.904 sec) +Saved checkpoint after 55 epoch(s) to data/resnet164/checkpoints/00055... +INFO:tensorflow:global_step/sec: 0.203102 +INFO:tensorflow:step = 21601, loss = 0.706559, precision = 0.882812 (492.364 sec) +INFO:tensorflow:global_step/sec: 0.201665 +INFO:tensorflow:step = 21701, loss = 0.687369, precision = 0.875 (495.871 sec) +INFO:tensorflow:global_step/sec: 0.204022 +INFO:tensorflow:step = 21801, loss = 0.620598, precision = 0.875 (490.143 sec) +Saved checkpoint after 56 epoch(s) to data/resnet164/checkpoints/00056... +INFO:tensorflow:global_step/sec: 0.20026 +INFO:tensorflow:step = 21901, loss = 0.598661, precision = 0.9375 (499.351 sec) +INFO:tensorflow:global_step/sec: 0.203733 +INFO:tensorflow:step = 22001, loss = 0.656961, precision = 0.890625 (490.839 sec) +INFO:tensorflow:global_step/sec: 0.206463 +INFO:tensorflow:step = 22101, loss = 0.590518, precision = 0.898438 (484.349 sec) +INFO:tensorflow:global_step/sec: 0.200267 +INFO:tensorflow:step = 22201, loss = 0.548689, precision = 0.929688 (499.333 sec) +Saved checkpoint after 57 epoch(s) to data/resnet164/checkpoints/00057... +INFO:tensorflow:global_step/sec: 0.197654 +INFO:tensorflow:step = 22301, loss = 0.51665, precision = 0.953125 (505.935 sec) +INFO:tensorflow:global_step/sec: 0.201203 +INFO:tensorflow:step = 22401, loss = 0.701949, precision = 0.859375 (497.009 sec) +INFO:tensorflow:global_step/sec: 0.202542 +INFO:tensorflow:step = 22501, loss = 0.632833, precision = 0.867188 (493.725 sec) +INFO:tensorflow:global_step/sec: 0.199555 +INFO:tensorflow:step = 22601, loss = 0.768551, precision = 0.859375 (501.116 sec) +Saved checkpoint after 58 epoch(s) to data/resnet164/checkpoints/00058... +INFO:tensorflow:global_step/sec: 0.197919 +INFO:tensorflow:step = 22701, loss = 0.5982, precision = 0.898438 (505.257 sec) +INFO:tensorflow:global_step/sec: 0.201674 +INFO:tensorflow:step = 22801, loss = 0.712559, precision = 0.859375 (495.849 sec) +INFO:tensorflow:global_step/sec: 0.202247 +INFO:tensorflow:step = 22901, loss = 0.564889, precision = 0.90625 (494.445 sec) +INFO:tensorflow:global_step/sec: 0.200142 +INFO:tensorflow:step = 23001, loss = 0.828347, precision = 0.828125 (499.645 sec) +Saved checkpoint after 59 epoch(s) to data/resnet164/checkpoints/00059... +INFO:tensorflow:global_step/sec: 0.196877 +INFO:tensorflow:step = 23101, loss = 0.850917, precision = 0.820312 (507.930 sec) +INFO:tensorflow:global_step/sec: 0.195062 +INFO:tensorflow:step = 23201, loss = 0.601493, precision = 0.898438 (512.658 sec) +INFO:tensorflow:global_step/sec: 0.198518 +INFO:tensorflow:step = 23301, loss = 0.66397, precision = 0.890625 (503.732 sec) +INFO:tensorflow:global_step/sec: 0.198395 +INFO:tensorflow:step = 23401, loss = 0.712116, precision = 0.859375 (504.046 sec) +Saved checkpoint after 60 epoch(s) to data/resnet164/checkpoints/00060... +INFO:tensorflow:global_step/sec: 0.197818 +INFO:tensorflow:step = 23501, loss = 0.594214, precision = 0.914062 (505.516 sec) +INFO:tensorflow:global_step/sec: 0.18887 +INFO:tensorflow:step = 23601, loss = 0.594376, precision = 0.929688 (529.463 sec) +INFO:tensorflow:global_step/sec: 0.193744 +INFO:tensorflow:step = 23701, loss = 0.732888, precision = 0.851562 (516.146 sec) +INFO:tensorflow:global_step/sec: 0.190701 +INFO:tensorflow:step = 23801, loss = 0.568494, precision = 0.945312 (524.382 sec) +Saved checkpoint after 61 epoch(s) to data/resnet164/checkpoints/00061... +INFO:tensorflow:global_step/sec: 0.19604 +INFO:tensorflow:step = 23901, loss = 0.751475, precision = 0.851562 (510.099 sec) +INFO:tensorflow:global_step/sec: 0.196093 +INFO:tensorflow:step = 24001, loss = 0.616328, precision = 0.875 (509.963 sec) +INFO:tensorflow:global_step/sec: 0.19494 +INFO:tensorflow:step = 24101, loss = 0.910781, precision = 0.804688 (512.978 sec) +INFO:tensorflow:global_step/sec: 0.180571 +INFO:tensorflow:step = 24201, loss = 0.655336, precision = 0.882812 (553.799 sec) +Saved checkpoint after 62 epoch(s) to data/resnet164/checkpoints/00062... +INFO:tensorflow:global_step/sec: 0.172634 +INFO:tensorflow:step = 24301, loss = 0.596541, precision = 0.90625 (579.262 sec) +INFO:tensorflow:global_step/sec: 0.195037 +INFO:tensorflow:step = 24401, loss = 0.556198, precision = 0.929688 (512.723 sec) +INFO:tensorflow:global_step/sec: 0.199602 +INFO:tensorflow:step = 24501, loss = 0.675434, precision = 0.875 (500.998 sec) +INFO:tensorflow:global_step/sec: 0.200569 +INFO:tensorflow:step = 24601, loss = 0.589795, precision = 0.890625 (498.581 sec) +Saved checkpoint after 63 epoch(s) to data/resnet164/checkpoints/00063... +INFO:tensorflow:global_step/sec: 0.196279 +INFO:tensorflow:step = 24701, loss = 0.556234, precision = 0.90625 (509.478 sec) +INFO:tensorflow:global_step/sec: 0.196896 +INFO:tensorflow:step = 24801, loss = 0.780173, precision = 0.828125 (507.883 sec) +INFO:tensorflow:global_step/sec: 0.195265 +INFO:tensorflow:step = 24901, loss = 0.613658, precision = 0.882812 (512.124 sec) +INFO:tensorflow:global_step/sec: 0.196647 +INFO:tensorflow:step = 25001, loss = 0.540738, precision = 0.90625 (508.526 sec) +Saved checkpoint after 64 epoch(s) to data/resnet164/checkpoints/00064... +INFO:tensorflow:global_step/sec: 0.183481 +INFO:tensorflow:step = 25101, loss = 0.760777, precision = 0.882812 (545.016 sec) +INFO:tensorflow:global_step/sec: 0.191401 +INFO:tensorflow:step = 25201, loss = 0.542077, precision = 0.929688 (522.463 sec) +INFO:tensorflow:global_step/sec: 0.189906 +INFO:tensorflow:step = 25301, loss = 0.59628, precision = 0.914062 (526.575 sec) +INFO:tensorflow:global_step/sec: 0.197505 +INFO:tensorflow:step = 25401, loss = 0.457168, precision = 0.960938 (506.316 sec) +Saved checkpoint after 65 epoch(s) to data/resnet164/checkpoints/00065... +INFO:tensorflow:global_step/sec: 0.196221 +INFO:tensorflow:step = 25501, loss = 0.619157, precision = 0.921875 (509.630 sec) +INFO:tensorflow:global_step/sec: 0.193126 +INFO:tensorflow:step = 25601, loss = 0.617576, precision = 0.898438 (517.795 sec) +INFO:tensorflow:global_step/sec: 0.190785 +INFO:tensorflow:step = 25701, loss = 0.647143, precision = 0.882812 (524.149 sec) +INFO:tensorflow:global_step/sec: 0.194716 +INFO:tensorflow:step = 25801, loss = 0.684796, precision = 0.882812 (513.567 sec) +Saved checkpoint after 66 epoch(s) to data/resnet164/checkpoints/00066... +INFO:tensorflow:global_step/sec: 0.187287 +INFO:tensorflow:step = 25901, loss = 0.619565, precision = 0.90625 (533.939 sec) +INFO:tensorflow:global_step/sec: 0.192845 +INFO:tensorflow:step = 26001, loss = 0.677533, precision = 0.914062 (518.550 sec) +INFO:tensorflow:global_step/sec: 0.201774 +INFO:tensorflow:step = 26101, loss = 0.719016, precision = 0.859375 (495.604 sec) +Saved checkpoint after 67 epoch(s) to data/resnet164/checkpoints/00067... +INFO:tensorflow:global_step/sec: 0.194366 +INFO:tensorflow:step = 26201, loss = 0.578726, precision = 0.890625 (514.493 sec) +INFO:tensorflow:global_step/sec: 0.187562 +INFO:tensorflow:step = 26301, loss = 0.690693, precision = 0.875 (533.157 sec) +INFO:tensorflow:global_step/sec: 0.190078 +INFO:tensorflow:step = 26401, loss = 0.632395, precision = 0.90625 (526.099 sec) +INFO:tensorflow:global_step/sec: 0.193426 +INFO:tensorflow:step = 26501, loss = 0.709299, precision = 0.867188 (516.995 sec) +Saved checkpoint after 68 epoch(s) to data/resnet164/checkpoints/00068... +INFO:tensorflow:global_step/sec: 0.195095 +INFO:tensorflow:step = 26601, loss = 0.519556, precision = 0.921875 (512.571 sec) +INFO:tensorflow:global_step/sec: 0.187383 +INFO:tensorflow:step = 26701, loss = 0.599217, precision = 0.882812 (533.666 sec) +INFO:tensorflow:global_step/sec: 0.153513 +INFO:tensorflow:step = 26801, loss = 0.609239, precision = 0.929688 (651.410 sec) +INFO:tensorflow:global_step/sec: 0.185221 +INFO:tensorflow:step = 26901, loss = 0.524366, precision = 0.929688 (539.896 sec) +Saved checkpoint after 69 epoch(s) to data/resnet164/checkpoints/00069... +INFO:tensorflow:global_step/sec: 0.192063 +INFO:tensorflow:step = 27001, loss = 0.607935, precision = 0.890625 (520.662 sec) +INFO:tensorflow:global_step/sec: 0.194362 +INFO:tensorflow:step = 27101, loss = 0.707572, precision = 0.851562 (514.505 sec) +INFO:tensorflow:global_step/sec: 0.187212 +INFO:tensorflow:step = 27201, loss = 0.625669, precision = 0.890625 (534.155 sec) +INFO:tensorflow:global_step/sec: 0.183903 +INFO:tensorflow:step = 27301, loss = 0.614064, precision = 0.898438 (543.764 sec) +Saved checkpoint after 70 epoch(s) to data/resnet164/checkpoints/00070... +INFO:tensorflow:global_step/sec: 0.162902 +INFO:tensorflow:step = 27401, loss = 0.591253, precision = 0.90625 (613.868 sec) +INFO:tensorflow:global_step/sec: 0.151417 +INFO:tensorflow:step = 27501, loss = 0.558233, precision = 0.921875 (660.428 sec) +INFO:tensorflow:global_step/sec: 0.154045 +INFO:tensorflow:step = 27601, loss = 0.639343, precision = 0.875 (649.161 sec) +INFO:tensorflow:global_step/sec: 0.18636 +INFO:tensorflow:step = 27701, loss = 0.569213, precision = 0.929688 (536.597 sec) +Saved checkpoint after 71 epoch(s) to data/resnet164/checkpoints/00071... +INFO:tensorflow:global_step/sec: 0.178085 +INFO:tensorflow:step = 27801, loss = 0.71887, precision = 0.828125 (561.530 sec) +INFO:tensorflow:global_step/sec: 0.19851 +INFO:tensorflow:step = 27901, loss = 0.614689, precision = 0.890625 (503.752 sec) +INFO:tensorflow:global_step/sec: 0.192524 +INFO:tensorflow:step = 28001, loss = 0.56975, precision = 0.921875 (519.416 sec) +INFO:tensorflow:global_step/sec: 0.181601 +INFO:tensorflow:step = 28101, loss = 0.669932, precision = 0.84375 (550.658 sec) +Saved checkpoint after 72 epoch(s) to data/resnet164/checkpoints/00072... +INFO:tensorflow:global_step/sec: 0.186556 +INFO:tensorflow:step = 28201, loss = 0.573512, precision = 0.929688 (536.032 sec) +INFO:tensorflow:global_step/sec: 0.189937 +INFO:tensorflow:step = 28301, loss = 0.664863, precision = 0.890625 (526.490 sec) +INFO:tensorflow:global_step/sec: 0.201212 +INFO:tensorflow:step = 28401, loss = 0.502067, precision = 0.945312 (496.988 sec) +INFO:tensorflow:global_step/sec: 0.197911 +INFO:tensorflow:step = 28501, loss = 0.736538, precision = 0.859375 (505.277 sec) +Saved checkpoint after 73 epoch(s) to data/resnet164/checkpoints/00073... +INFO:tensorflow:global_step/sec: 0.199191 +INFO:tensorflow:step = 28601, loss = 0.685587, precision = 0.851562 (502.031 sec) +INFO:tensorflow:global_step/sec: 0.191996 +INFO:tensorflow:step = 28701, loss = 0.617995, precision = 0.929688 (520.845 sec) +INFO:tensorflow:global_step/sec: 0.193283 +INFO:tensorflow:step = 28801, loss = 0.575745, precision = 0.898438 (517.375 sec) +INFO:tensorflow:global_step/sec: 0.189858 +INFO:tensorflow:step = 28901, loss = 0.595688, precision = 0.882812 (526.708 sec) +Saved checkpoint after 74 epoch(s) to data/resnet164/checkpoints/00074... +INFO:tensorflow:global_step/sec: 0.192471 +INFO:tensorflow:step = 29001, loss = 0.604407, precision = 0.914062 (519.560 sec) +INFO:tensorflow:global_step/sec: 0.183937 +INFO:tensorflow:step = 29101, loss = 0.618375, precision = 0.898438 (543.665 sec) +INFO:tensorflow:global_step/sec: 0.183386 +INFO:tensorflow:step = 29201, loss = 0.659823, precision = 0.882812 (545.298 sec) +INFO:tensorflow:global_step/sec: 0.19523 +INFO:tensorflow:step = 29301, loss = 0.632211, precision = 0.890625 (512.215 sec) +Saved checkpoint after 75 epoch(s) to data/resnet164/checkpoints/00075... +INFO:tensorflow:global_step/sec: 0.193427 +INFO:tensorflow:step = 29401, loss = 0.544096, precision = 0.921875 (516.992 sec) +INFO:tensorflow:global_step/sec: 0.189267 +INFO:tensorflow:step = 29501, loss = 0.585775, precision = 0.898438 (528.354 sec) +INFO:tensorflow:global_step/sec: 0.19701 +INFO:tensorflow:step = 29601, loss = 0.609668, precision = 0.890625 (507.588 sec) +INFO:tensorflow:global_step/sec: 0.194084 +INFO:tensorflow:step = 29701, loss = 0.590667, precision = 0.914062 (515.240 sec) +Saved checkpoint after 76 epoch(s) to data/resnet164/checkpoints/00076... +INFO:tensorflow:global_step/sec: 0.191136 +INFO:tensorflow:step = 29801, loss = 0.706386, precision = 0.859375 (523.187 sec) +INFO:tensorflow:global_step/sec: 0.180651 +INFO:tensorflow:step = 29901, loss = 0.666568, precision = 0.898438 (553.552 sec) +INFO:tensorflow:global_step/sec: 0.189427 +INFO:tensorflow:step = 30001, loss = 0.584404, precision = 0.890625 (527.907 sec) +INFO:tensorflow:global_step/sec: 0.195055 +INFO:tensorflow:step = 30101, loss = 0.60677, precision = 0.90625 (512.677 sec) +Saved checkpoint after 77 epoch(s) to data/resnet164/checkpoints/00077... +INFO:tensorflow:global_step/sec: 0.192448 +INFO:tensorflow:step = 30201, loss = 0.6133, precision = 0.90625 (519.621 sec) +INFO:tensorflow:global_step/sec: 0.188498 +INFO:tensorflow:step = 30301, loss = 0.901219, precision = 0.8125 (530.509 sec) +INFO:tensorflow:global_step/sec: 0.183771 +INFO:tensorflow:step = 30401, loss = 0.558876, precision = 0.914062 (544.156 sec) +Saved checkpoint after 78 epoch(s) to data/resnet164/checkpoints/00078... +INFO:tensorflow:global_step/sec: 0.191754 +INFO:tensorflow:step = 30501, loss = 0.553231, precision = 0.921875 (521.502 sec) +INFO:tensorflow:global_step/sec: 0.190752 +INFO:tensorflow:step = 30601, loss = 0.555134, precision = 0.921875 (524.241 sec) +INFO:tensorflow:global_step/sec: 0.190687 +INFO:tensorflow:step = 30701, loss = 0.584475, precision = 0.90625 (524.418 sec) +INFO:tensorflow:global_step/sec: 0.201629 +INFO:tensorflow:step = 30801, loss = 0.517568, precision = 0.9375 (495.960 sec) +Saved checkpoint after 79 epoch(s) to data/resnet164/checkpoints/00079... +INFO:tensorflow:global_step/sec: 0.189894 +INFO:tensorflow:step = 30901, loss = 0.57697, precision = 0.882812 (526.609 sec) +INFO:tensorflow:global_step/sec: 0.190332 +INFO:tensorflow:step = 31001, loss = 0.584669, precision = 0.898438 (525.399 sec) +INFO:tensorflow:global_step/sec: 0.197827 +INFO:tensorflow:step = 31101, loss = 0.647562, precision = 0.898438 (505.493 sec) +INFO:tensorflow:global_step/sec: 0.203469 +INFO:tensorflow:step = 31201, loss = 0.577458, precision = 0.929688 (491.474 sec) +Saved checkpoint after 80 epoch(s) to data/resnet164/checkpoints/00080... +INFO:tensorflow:global_step/sec: 0.206369 +INFO:tensorflow:step = 31301, loss = 0.668297, precision = 0.890625 (484.570 sec) +INFO:tensorflow:global_step/sec: 0.205864 +INFO:tensorflow:step = 31401, loss = 0.605316, precision = 0.882812 (485.757 sec) +INFO:tensorflow:global_step/sec: 0.196004 +INFO:tensorflow:step = 31501, loss = 0.818181, precision = 0.828125 (510.194 sec) +INFO:tensorflow:global_step/sec: 0.194292 +INFO:tensorflow:step = 31601, loss = 0.614573, precision = 0.921875 (514.688 sec) +Saved checkpoint after 81 epoch(s) to data/resnet164/checkpoints/00081... +INFO:tensorflow:global_step/sec: 0.186357 +INFO:tensorflow:step = 31701, loss = 0.636042, precision = 0.898438 (536.606 sec) +INFO:tensorflow:global_step/sec: 0.189176 +INFO:tensorflow:step = 31801, loss = 0.586951, precision = 0.929688 (528.608 sec) +INFO:tensorflow:global_step/sec: 0.186676 +INFO:tensorflow:step = 31901, loss = 0.652054, precision = 0.890625 (535.687 sec) +INFO:tensorflow:global_step/sec: 0.189284 +INFO:tensorflow:step = 32001, loss = 0.649933, precision = 0.882812 (528.307 sec) +Saved checkpoint after 82 epoch(s) to data/resnet164/checkpoints/00082... +INFO:tensorflow:global_step/sec: 0.190998 +INFO:tensorflow:step = 32101, loss = 0.620821, precision = 0.898438 (523.567 sec) +INFO:tensorflow:global_step/sec: 0.196745 +INFO:tensorflow:step = 32201, loss = 0.645892, precision = 0.890625 (508.272 sec) +INFO:tensorflow:global_step/sec: 0.199783 +INFO:tensorflow:step = 32301, loss = 0.630566, precision = 0.921875 (500.542 sec) +INFO:tensorflow:global_step/sec: 0.202036 +INFO:tensorflow:step = 32401, loss = 0.541263, precision = 0.90625 (494.962 sec) +Saved checkpoint after 83 epoch(s) to data/resnet164/checkpoints/00083... +INFO:tensorflow:global_step/sec: 0.193579 +INFO:tensorflow:step = 32501, loss = 0.661151, precision = 0.898438 (516.584 sec) +INFO:tensorflow:global_step/sec: 0.20095 +INFO:tensorflow:step = 32601, loss = 0.76557, precision = 0.851562 (497.636 sec) +INFO:tensorflow:global_step/sec: 0.201486 +INFO:tensorflow:step = 32701, loss = 0.55761, precision = 0.929688 (496.313 sec) +INFO:tensorflow:global_step/sec: 0.200759 +INFO:tensorflow:step = 32801, loss = 0.558845, precision = 0.921875 (498.108 sec) +Saved checkpoint after 84 epoch(s) to data/resnet164/checkpoints/00084... +INFO:tensorflow:global_step/sec: 0.200233 +INFO:tensorflow:step = 32901, loss = 0.747049, precision = 0.851562 (499.418 sec) +INFO:tensorflow:global_step/sec: 0.197892 +INFO:tensorflow:step = 33001, loss = 0.601725, precision = 0.90625 (505.326 sec) +INFO:tensorflow:global_step/sec: 0.202129 +INFO:tensorflow:step = 33101, loss = 0.524847, precision = 0.921875 (494.733 sec) +INFO:tensorflow:global_step/sec: 0.205846 +INFO:tensorflow:step = 33201, loss = 0.599357, precision = 0.890625 (485.801 sec) +Saved checkpoint after 85 epoch(s) to data/resnet164/checkpoints/00085... +INFO:tensorflow:global_step/sec: 0.204318 +INFO:tensorflow:step = 33301, loss = 0.67647, precision = 0.875 (489.433 sec) +INFO:tensorflow:global_step/sec: 0.210089 +INFO:tensorflow:step = 33401, loss = 0.528065, precision = 0.945312 (475.988 sec) +INFO:tensorflow:global_step/sec: 0.2056 +INFO:tensorflow:step = 33501, loss = 0.583156, precision = 0.914062 (486.381 sec) +INFO:tensorflow:global_step/sec: 0.204218 +INFO:tensorflow:step = 33601, loss = 0.53612, precision = 0.929688 (489.673 sec) +Saved checkpoint after 86 epoch(s) to data/resnet164/checkpoints/00086... +INFO:tensorflow:global_step/sec: 0.209584 +INFO:tensorflow:step = 33701, loss = 0.701073, precision = 0.890625 (477.136 sec) +INFO:tensorflow:global_step/sec: 0.204198 +INFO:tensorflow:step = 33801, loss = 0.644502, precision = 0.90625 (489.722 sec) +INFO:tensorflow:global_step/sec: 0.206061 +INFO:tensorflow:step = 33901, loss = 0.612429, precision = 0.890625 (485.294 sec) +INFO:tensorflow:global_step/sec: 0.205518 +INFO:tensorflow:step = 34001, loss = 0.637029, precision = 0.875 (486.575 sec) +Saved checkpoint after 87 epoch(s) to data/resnet164/checkpoints/00087... +INFO:tensorflow:global_step/sec: 0.204936 +INFO:tensorflow:step = 34101, loss = 0.743055, precision = 0.851562 (487.959 sec) +INFO:tensorflow:global_step/sec: 0.210254 +INFO:tensorflow:step = 34201, loss = 0.67827, precision = 0.875 (475.616 sec) +INFO:tensorflow:global_step/sec: 0.208168 +INFO:tensorflow:step = 34301, loss = 0.778541, precision = 0.828125 (480.381 sec) +INFO:tensorflow:global_step/sec: 0.208597 +INFO:tensorflow:step = 34401, loss = 0.602823, precision = 0.914062 (479.394 sec) +Saved checkpoint after 88 epoch(s) to data/resnet164/checkpoints/00088... +INFO:tensorflow:global_step/sec: 0.210618 +INFO:tensorflow:step = 34501, loss = 0.583303, precision = 0.914062 (474.794 sec) +INFO:tensorflow:global_step/sec: 0.209704 +INFO:tensorflow:step = 34601, loss = 0.606781, precision = 0.875 (476.863 sec) +INFO:tensorflow:global_step/sec: 0.211989 +INFO:tensorflow:step = 34701, loss = 0.590863, precision = 0.890625 (471.723 sec) +Saved checkpoint after 89 epoch(s) to data/resnet164/checkpoints/00089... +INFO:tensorflow:global_step/sec: 0.211295 +INFO:tensorflow:step = 34801, loss = 0.700416, precision = 0.859375 (473.272 sec) +INFO:tensorflow:global_step/sec: 0.207777 +INFO:tensorflow:step = 34901, loss = 0.661891, precision = 0.867188 (481.284 sec) +INFO:tensorflow:global_step/sec: 0.207097 +INFO:tensorflow:step = 35001, loss = 0.58084, precision = 0.90625 (482.866 sec) +INFO:tensorflow:global_step/sec: 0.203352 +INFO:tensorflow:step = 35101, loss = 0.705136, precision = 0.867188 (491.759 sec) +Saved checkpoint after 90 epoch(s) to data/resnet164/checkpoints/00090... +INFO:tensorflow:global_step/sec: 0.206242 +INFO:tensorflow:step = 35201, loss = 0.605627, precision = 0.898438 (484.866 sec) +INFO:tensorflow:global_step/sec: 0.200577 +INFO:tensorflow:step = 35301, loss = 0.714373, precision = 0.859375 (498.560 sec) +INFO:tensorflow:global_step/sec: 0.20744 +INFO:tensorflow:step = 35401, loss = 0.553531, precision = 0.921875 (482.067 sec) +INFO:tensorflow:global_step/sec: 0.205585 +INFO:tensorflow:step = 35501, loss = 0.575813, precision = 0.90625 (486.416 sec) +Saved checkpoint after 91 epoch(s) to data/resnet164/checkpoints/00091... +INFO:tensorflow:global_step/sec: 0.20402 +INFO:tensorflow:step = 35601, loss = 0.641281, precision = 0.914062 (490.147 sec) +INFO:tensorflow:global_step/sec: 0.204066 +INFO:tensorflow:step = 35701, loss = 0.446543, precision = 0.953125 (490.038 sec) +INFO:tensorflow:global_step/sec: 0.203472 +INFO:tensorflow:step = 35801, loss = 0.575262, precision = 0.9375 (491.468 sec) +INFO:tensorflow:global_step/sec: 0.210637 +INFO:tensorflow:step = 35901, loss = 0.518979, precision = 0.9375 (474.749 sec) +Saved checkpoint after 92 epoch(s) to data/resnet164/checkpoints/00092... +INFO:tensorflow:global_step/sec: 0.214877 +INFO:tensorflow:step = 36001, loss = 0.395348, precision = 0.976562 (465.382 sec) +INFO:tensorflow:global_step/sec: 0.21817 +INFO:tensorflow:step = 36101, loss = 0.469471, precision = 0.945312 (458.357 sec) +INFO:tensorflow:global_step/sec: 0.218505 +INFO:tensorflow:step = 36201, loss = 0.440518, precision = 0.953125 (457.656 sec) +INFO:tensorflow:global_step/sec: 0.214178 +INFO:tensorflow:step = 36301, loss = 0.382113, precision = 0.976562 (466.902 sec) +Saved checkpoint after 93 epoch(s) to data/resnet164/checkpoints/00093... +INFO:tensorflow:global_step/sec: 0.217078 +INFO:tensorflow:step = 36401, loss = 0.343698, precision = 0.992188 (460.664 sec) +INFO:tensorflow:global_step/sec: 0.215736 +INFO:tensorflow:step = 36501, loss = 0.393518, precision = 0.976562 (463.529 sec) +INFO:tensorflow:global_step/sec: 0.218267 +INFO:tensorflow:step = 36601, loss = 0.41184, precision = 0.960938 (458.154 sec) +INFO:tensorflow:global_step/sec: 0.218408 +INFO:tensorflow:step = 36701, loss = 0.367417, precision = 0.976562 (457.858 sec) +Saved checkpoint after 94 epoch(s) to data/resnet164/checkpoints/00094... +INFO:tensorflow:global_step/sec: 0.215318 +INFO:tensorflow:step = 36801, loss = 0.397605, precision = 0.960938 (464.430 sec) +INFO:tensorflow:global_step/sec: 0.218935 +INFO:tensorflow:step = 36901, loss = 0.366975, precision = 0.976562 (456.757 sec) +INFO:tensorflow:global_step/sec: 0.21529 +INFO:tensorflow:step = 37001, loss = 0.420121, precision = 0.945312 (464.489 sec) +INFO:tensorflow:global_step/sec: 0.21818 +INFO:tensorflow:step = 37101, loss = 0.341655, precision = 0.984375 (458.337 sec) +Saved checkpoint after 95 epoch(s) to data/resnet164/checkpoints/00095... +INFO:tensorflow:global_step/sec: 0.217546 +INFO:tensorflow:step = 37201, loss = 0.390435, precision = 0.976562 (459.672 sec) +INFO:tensorflow:global_step/sec: 0.219568 +INFO:tensorflow:step = 37301, loss = 0.405988, precision = 0.96875 (455.439 sec) +INFO:tensorflow:global_step/sec: 0.216071 +INFO:tensorflow:step = 37401, loss = 0.392559, precision = 0.96875 (462.810 sec) +INFO:tensorflow:global_step/sec: 0.213724 +INFO:tensorflow:step = 37501, loss = 0.356852, precision = 0.960938 (467.893 sec) +Saved checkpoint after 96 epoch(s) to data/resnet164/checkpoints/00096... +INFO:tensorflow:global_step/sec: 0.209524 +INFO:tensorflow:step = 37601, loss = 0.337277, precision = 0.976562 (477.273 sec) +INFO:tensorflow:global_step/sec: 0.213598 +INFO:tensorflow:step = 37701, loss = 0.35482, precision = 0.976562 (468.169 sec) +INFO:tensorflow:global_step/sec: 0.213563 +INFO:tensorflow:step = 37801, loss = 0.371139, precision = 0.953125 (468.246 sec) +INFO:tensorflow:global_step/sec: 0.213172 +INFO:tensorflow:step = 37901, loss = 0.32545, precision = 0.976562 (469.104 sec) +Saved checkpoint after 97 epoch(s) to data/resnet164/checkpoints/00097... +INFO:tensorflow:global_step/sec: 0.212772 +INFO:tensorflow:step = 38001, loss = 0.339658, precision = 0.96875 (469.987 sec) +INFO:tensorflow:global_step/sec: 0.21266 +INFO:tensorflow:step = 38101, loss = 0.303672, precision = 0.984375 (470.234 sec) +INFO:tensorflow:global_step/sec: 0.213549 +INFO:tensorflow:step = 38201, loss = 0.31964, precision = 0.96875 (468.277 sec) +INFO:tensorflow:global_step/sec: 0.213898 +INFO:tensorflow:step = 38301, loss = 0.358681, precision = 0.921875 (467.512 sec) +Saved checkpoint after 98 epoch(s) to data/resnet164/checkpoints/00098... +INFO:tensorflow:global_step/sec: 0.214921 +INFO:tensorflow:step = 38401, loss = 0.348351, precision = 0.992188 (465.287 sec) +INFO:tensorflow:global_step/sec: 0.215791 +INFO:tensorflow:step = 38501, loss = 0.299072, precision = 0.992188 (463.412 sec) +INFO:tensorflow:global_step/sec: 0.21396 +INFO:tensorflow:step = 38601, loss = 0.326074, precision = 0.96875 (467.377 sec) +INFO:tensorflow:global_step/sec: 0.215405 +INFO:tensorflow:step = 38701, loss = 0.30258, precision = 0.976562 (464.242 sec) +Saved checkpoint after 99 epoch(s) to data/resnet164/checkpoints/00099... +INFO:tensorflow:global_step/sec: 0.208893 +INFO:tensorflow:step = 38801, loss = 0.311752, precision = 0.984375 (478.714 sec) +INFO:tensorflow:global_step/sec: 0.210239 +INFO:tensorflow:step = 38901, loss = 0.34474, precision = 0.976562 (475.649 sec) +INFO:tensorflow:global_step/sec: 0.206651 +INFO:tensorflow:step = 39001, loss = 0.338047, precision = 0.960938 (483.907 sec) +Saved checkpoint after 100 epoch(s) to data/resnet164/checkpoints/00100... +INFO:tensorflow:global_step/sec: 0.208035 +INFO:tensorflow:step = 39101, loss = 0.270482, precision = 0.992188 (480.689 sec) +INFO:tensorflow:global_step/sec: 0.212043 +INFO:tensorflow:step = 39201, loss = 0.312663, precision = 0.976562 (471.602 sec) +INFO:tensorflow:global_step/sec: 0.210176 +INFO:tensorflow:step = 39301, loss = 0.332679, precision = 0.976562 (475.792 sec) +INFO:tensorflow:global_step/sec: 0.208314 +INFO:tensorflow:step = 39401, loss = 0.332767, precision = 0.96875 (480.044 sec) +Saved checkpoint after 101 epoch(s) to data/resnet164/checkpoints/00101... +INFO:tensorflow:global_step/sec: 0.205185 +INFO:tensorflow:step = 39501, loss = 0.356896, precision = 0.953125 (487.366 sec) +INFO:tensorflow:global_step/sec: 0.204869 +INFO:tensorflow:step = 39601, loss = 0.267102, precision = 0.992188 (488.116 sec) +INFO:tensorflow:global_step/sec: 0.203195 +INFO:tensorflow:step = 39701, loss = 0.283068, precision = 0.984375 (492.138 sec) +INFO:tensorflow:global_step/sec: 0.20118 +INFO:tensorflow:step = 39801, loss = 0.294336, precision = 0.976562 (497.068 sec) +Saved checkpoint after 102 epoch(s) to data/resnet164/checkpoints/00102... +INFO:tensorflow:global_step/sec: 0.202953 +INFO:tensorflow:step = 39901, loss = 0.278781, precision = 0.976562 (492.726 sec) +INFO:tensorflow:global_step/sec: 0.202194 +INFO:tensorflow:step = 40001, loss = 0.305332, precision = 0.960938 (494.575 sec) +INFO:tensorflow:global_step/sec: 0.20515 +INFO:tensorflow:step = 40101, loss = 0.346145, precision = 0.960938 (487.448 sec) +INFO:tensorflow:global_step/sec: 0.205449 +INFO:tensorflow:step = 40201, loss = 0.289178, precision = 0.984375 (486.739 sec) +Saved checkpoint after 103 epoch(s) to data/resnet164/checkpoints/00103... +INFO:tensorflow:global_step/sec: 0.205885 +INFO:tensorflow:step = 40301, loss = 0.270441, precision = 0.96875 (485.709 sec) +INFO:tensorflow:global_step/sec: 0.205197 +INFO:tensorflow:step = 40401, loss = 0.247675, precision = 0.984375 (487.337 sec) +INFO:tensorflow:global_step/sec: 0.207816 +INFO:tensorflow:step = 40501, loss = 0.267138, precision = 0.976562 (481.195 sec) +INFO:tensorflow:global_step/sec: 0.206316 +INFO:tensorflow:step = 40601, loss = 0.253681, precision = 0.984375 (484.693 sec) +Saved checkpoint after 104 epoch(s) to data/resnet164/checkpoints/00104... +INFO:tensorflow:global_step/sec: 0.202359 +INFO:tensorflow:step = 40701, loss = 0.266186, precision = 0.984375 (494.172 sec) +INFO:tensorflow:global_step/sec: 0.203177 +INFO:tensorflow:step = 40801, loss = 0.247645, precision = 1.0 (492.183 sec) +INFO:tensorflow:global_step/sec: 0.203455 +INFO:tensorflow:step = 40901, loss = 0.280346, precision = 0.96875 (491.510 sec) +INFO:tensorflow:global_step/sec: 0.203034 +INFO:tensorflow:step = 41001, loss = 0.285189, precision = 0.984375 (492.530 sec) +Saved checkpoint after 105 epoch(s) to data/resnet164/checkpoints/00105... +INFO:tensorflow:global_step/sec: 0.19457 +INFO:tensorflow:step = 41101, loss = 0.241947, precision = 1.0 (513.953 sec) +INFO:tensorflow:global_step/sec: 0.195279 +INFO:tensorflow:step = 41201, loss = 0.245312, precision = 0.984375 (512.086 sec) +INFO:tensorflow:global_step/sec: 0.200186 +INFO:tensorflow:step = 41301, loss = 0.232765, precision = 0.992188 (499.535 sec) +INFO:tensorflow:global_step/sec: 0.204116 +INFO:tensorflow:step = 41401, loss = 0.236068, precision = 1.0 (489.919 sec) +Saved checkpoint after 106 epoch(s) to data/resnet164/checkpoints/00106... +INFO:tensorflow:global_step/sec: 0.204916 +INFO:tensorflow:step = 41501, loss = 0.231827, precision = 0.984375 (488.005 sec) +INFO:tensorflow:global_step/sec: 0.204842 +INFO:tensorflow:step = 41601, loss = 0.227627, precision = 1.0 (488.180 sec) +INFO:tensorflow:global_step/sec: 0.200601 +INFO:tensorflow:step = 41701, loss = 0.241882, precision = 0.992188 (498.503 sec) +INFO:tensorflow:global_step/sec: 0.199084 +INFO:tensorflow:step = 41801, loss = 0.282412, precision = 0.96875 (502.300 sec) +Saved checkpoint after 107 epoch(s) to data/resnet164/checkpoints/00107... +INFO:tensorflow:global_step/sec: 0.201587 +INFO:tensorflow:step = 41901, loss = 0.221301, precision = 1.0 (496.063 sec) +INFO:tensorflow:global_step/sec: 0.199253 +INFO:tensorflow:step = 42001, loss = 0.215127, precision = 1.0 (501.875 sec) +INFO:tensorflow:global_step/sec: 0.202227 +INFO:tensorflow:step = 42101, loss = 0.256441, precision = 0.984375 (494.495 sec) +INFO:tensorflow:global_step/sec: 0.195305 +INFO:tensorflow:step = 42201, loss = 0.2467, precision = 0.992188 (512.020 sec) +Saved checkpoint after 108 epoch(s) to data/resnet164/checkpoints/00108... +INFO:tensorflow:global_step/sec: 0.199956 +INFO:tensorflow:step = 42301, loss = 0.260856, precision = 0.976562 (500.109 sec) +INFO:tensorflow:global_step/sec: 0.199777 +INFO:tensorflow:step = 42401, loss = 0.28236, precision = 0.96875 (500.558 sec) +INFO:tensorflow:global_step/sec: 0.200212 +INFO:tensorflow:step = 42501, loss = 0.245394, precision = 0.96875 (499.471 sec) +INFO:tensorflow:global_step/sec: 0.200622 +INFO:tensorflow:step = 42601, loss = 0.221777, precision = 0.992188 (498.450 sec) +Saved checkpoint after 109 epoch(s) to data/resnet164/checkpoints/00109... +INFO:tensorflow:global_step/sec: 0.200508 +INFO:tensorflow:step = 42701, loss = 0.236763, precision = 0.984375 (498.733 sec) +INFO:tensorflow:global_step/sec: 0.198375 +INFO:tensorflow:step = 42801, loss = 0.224234, precision = 0.984375 (504.094 sec) +INFO:tensorflow:global_step/sec: 0.201178 +INFO:tensorflow:step = 42901, loss = 0.275249, precision = 0.976562 (497.072 sec) +INFO:tensorflow:global_step/sec: 0.199692 +INFO:tensorflow:step = 43001, loss = 0.261851, precision = 0.960938 (500.772 sec) +Saved checkpoint after 110 epoch(s) to data/resnet164/checkpoints/00110... +INFO:tensorflow:global_step/sec: 0.196331 +INFO:tensorflow:step = 43101, loss = 0.209544, precision = 0.992188 (509.345 sec) +INFO:tensorflow:global_step/sec: 0.195873 +INFO:tensorflow:step = 43201, loss = 0.216409, precision = 0.984375 (510.536 sec) +INFO:tensorflow:global_step/sec: 0.18315 +INFO:tensorflow:step = 43301, loss = 0.21768, precision = 0.992188 (546.002 sec) +Saved checkpoint after 111 epoch(s) to data/resnet164/checkpoints/00111... +INFO:tensorflow:global_step/sec: 0.196684 +INFO:tensorflow:step = 43401, loss = 0.184752, precision = 1.0 (508.429 sec) +INFO:tensorflow:global_step/sec: 0.197005 +INFO:tensorflow:step = 43501, loss = 0.213675, precision = 0.984375 (507.601 sec) +INFO:tensorflow:global_step/sec: 0.2006 +INFO:tensorflow:step = 43601, loss = 0.228017, precision = 0.984375 (498.504 sec) +INFO:tensorflow:global_step/sec: 0.199881 +INFO:tensorflow:step = 43701, loss = 0.289575, precision = 0.96875 (500.299 sec) +Saved checkpoint after 112 epoch(s) to data/resnet164/checkpoints/00112... +INFO:tensorflow:global_step/sec: 0.198087 +INFO:tensorflow:step = 43801, loss = 0.251848, precision = 0.960938 (504.828 sec) +INFO:tensorflow:global_step/sec: 0.19564 +INFO:tensorflow:step = 43901, loss = 0.245282, precision = 0.976562 (511.143 sec) +INFO:tensorflow:global_step/sec: 0.185404 +INFO:tensorflow:step = 44001, loss = 0.187767, precision = 1.0 (539.362 sec) +INFO:tensorflow:global_step/sec: 0.188126 +INFO:tensorflow:step = 44101, loss = 0.193054, precision = 1.0 (531.558 sec) +Saved checkpoint after 113 epoch(s) to data/resnet164/checkpoints/00113... +INFO:tensorflow:global_step/sec: 0.197333 +INFO:tensorflow:step = 44201, loss = 0.205098, precision = 0.992188 (506.759 sec) +INFO:tensorflow:global_step/sec: 0.194281 +INFO:tensorflow:step = 44301, loss = 0.197556, precision = 0.992188 (514.718 sec) +INFO:tensorflow:global_step/sec: 0.160767 +INFO:tensorflow:step = 44401, loss = 0.265703, precision = 0.96875 (622.017 sec) +INFO:tensorflow:global_step/sec: 0.164397 +INFO:tensorflow:step = 44501, loss = 0.193309, precision = 0.992188 (608.284 sec) +Saved checkpoint after 114 epoch(s) to data/resnet164/checkpoints/00114... +INFO:tensorflow:global_step/sec: 0.158584 +INFO:tensorflow:step = 44601, loss = 0.212256, precision = 0.992188 (630.579 sec) +INFO:tensorflow:global_step/sec: 0.16159 +INFO:tensorflow:step = 44701, loss = 0.179067, precision = 1.0 (618.852 sec) +INFO:tensorflow:global_step/sec: 0.178484 +INFO:tensorflow:step = 44801, loss = 0.242142, precision = 0.960938 (560.275 sec) +INFO:tensorflow:global_step/sec: 0.19413 +INFO:tensorflow:step = 44901, loss = 0.246567, precision = 0.984375 (515.120 sec) +Saved checkpoint after 115 epoch(s) to data/resnet164/checkpoints/00115... +INFO:tensorflow:global_step/sec: 0.156339 +INFO:tensorflow:step = 45001, loss = 0.208577, precision = 0.992188 (639.636 sec) +INFO:tensorflow:global_step/sec: 0.161322 +INFO:tensorflow:step = 45101, loss = 0.180785, precision = 1.0 (619.879 sec) +INFO:tensorflow:global_step/sec: 0.159837 +INFO:tensorflow:step = 45201, loss = 0.227578, precision = 0.984375 (625.637 sec) +INFO:tensorflow:global_step/sec: 0.163865 +INFO:tensorflow:step = 45301, loss = 0.187193, precision = 1.0 (610.258 sec) +Saved checkpoint after 116 epoch(s) to data/resnet164/checkpoints/00116... +INFO:tensorflow:global_step/sec: 0.163386 +INFO:tensorflow:step = 45401, loss = 0.193279, precision = 1.0 (612.048 sec) +INFO:tensorflow:global_step/sec: 0.154396 +INFO:tensorflow:step = 45501, loss = 0.219143, precision = 0.984375 (647.683 sec) +INFO:tensorflow:global_step/sec: 0.157167 +INFO:tensorflow:step = 45601, loss = 0.23861, precision = 0.976562 (636.267 sec) +INFO:tensorflow:global_step/sec: 0.165503 +INFO:tensorflow:step = 45701, loss = 0.183445, precision = 1.0 (604.218 sec) +Saved checkpoint after 117 epoch(s) to data/resnet164/checkpoints/00117... +INFO:tensorflow:global_step/sec: 0.161313 +INFO:tensorflow:step = 45801, loss = 0.199912, precision = 0.984375 (619.915 sec) +INFO:tensorflow:global_step/sec: 0.162111 +INFO:tensorflow:step = 45901, loss = 0.174026, precision = 1.0 (616.861 sec) +INFO:tensorflow:global_step/sec: 0.166605 +INFO:tensorflow:step = 46001, loss = 0.17914, precision = 1.0 (600.224 sec) +INFO:tensorflow:global_step/sec: 0.16734 +INFO:tensorflow:step = 46101, loss = 0.210231, precision = 0.992188 (597.584 sec) +Saved checkpoint after 118 epoch(s) to data/resnet164/checkpoints/00118... +INFO:tensorflow:global_step/sec: 0.164192 +INFO:tensorflow:step = 46201, loss = 0.194966, precision = 0.984375 (609.044 sec) +INFO:tensorflow:global_step/sec: 0.161806 +INFO:tensorflow:step = 46301, loss = 0.173763, precision = 1.0 (618.024 sec) +INFO:tensorflow:global_step/sec: 0.160605 +INFO:tensorflow:step = 46401, loss = 0.220583, precision = 0.984375 (622.648 sec) +INFO:tensorflow:global_step/sec: 0.161448 +INFO:tensorflow:step = 46501, loss = 0.187266, precision = 0.992188 (619.394 sec) +Saved checkpoint after 119 epoch(s) to data/resnet164/checkpoints/00119... +INFO:tensorflow:global_step/sec: 0.162867 +INFO:tensorflow:step = 46601, loss = 0.245929, precision = 0.976562 (614.000 sec) +INFO:tensorflow:global_step/sec: 0.162671 +INFO:tensorflow:step = 46701, loss = 0.225677, precision = 0.960938 (614.739 sec) +INFO:tensorflow:global_step/sec: 0.163841 +INFO:tensorflow:step = 46801, loss = 0.169515, precision = 1.0 (610.348 sec) +INFO:tensorflow:global_step/sec: 0.163131 +INFO:tensorflow:step = 46901, loss = 0.216681, precision = 0.984375 (613.005 sec) +Saved checkpoint after 120 epoch(s) to data/resnet164/checkpoints/00120... +INFO:tensorflow:global_step/sec: 0.162421 +INFO:tensorflow:step = 47001, loss = 0.24834, precision = 0.976562 (615.685 sec) +INFO:tensorflow:global_step/sec: 0.162582 +INFO:tensorflow:step = 47101, loss = 0.177567, precision = 0.984375 (615.074 sec) +INFO:tensorflow:global_step/sec: 0.162536 +INFO:tensorflow:step = 47201, loss = 0.206886, precision = 0.984375 (615.250 sec) +INFO:tensorflow:global_step/sec: 0.162518 +INFO:tensorflow:step = 47301, loss = 0.194428, precision = 0.992188 (615.315 sec) +Saved checkpoint after 121 epoch(s) to data/resnet164/checkpoints/00121... +INFO:tensorflow:global_step/sec: 0.161829 +INFO:tensorflow:step = 47401, loss = 0.251472, precision = 0.96875 (617.936 sec) +INFO:tensorflow:global_step/sec: 0.168712 +INFO:tensorflow:step = 47501, loss = 0.234722, precision = 0.976562 (592.726 sec) +INFO:tensorflow:global_step/sec: 0.163509 +INFO:tensorflow:step = 47601, loss = 0.196874, precision = 0.976562 (611.586 sec) +INFO:tensorflow:global_step/sec: 0.161214 +INFO:tensorflow:step = 47701, loss = 0.197534, precision = 0.992188 (620.293 sec) +Saved checkpoint after 122 epoch(s) to data/resnet164/checkpoints/00122... +INFO:tensorflow:global_step/sec: 0.161666 +INFO:tensorflow:step = 47801, loss = 0.205915, precision = 0.984375 (618.560 sec) +INFO:tensorflow:global_step/sec: 0.163846 +INFO:tensorflow:step = 47901, loss = 0.235132, precision = 0.96875 (610.330 sec) +INFO:tensorflow:global_step/sec: 0.169156 +INFO:tensorflow:step = 48001, loss = 0.272107, precision = 0.953125 (591.169 sec) +Saved checkpoint after 123 epoch(s) to data/resnet164/checkpoints/00123... +INFO:tensorflow:global_step/sec: 0.168218 +INFO:tensorflow:step = 48101, loss = 0.205409, precision = 0.976562 (594.468 sec) +INFO:tensorflow:global_step/sec: 0.169818 +INFO:tensorflow:step = 48201, loss = 0.177534, precision = 0.992188 (588.865 sec) +INFO:tensorflow:global_step/sec: 0.165698 +INFO:tensorflow:step = 48301, loss = 0.175125, precision = 1.0 (603.506 sec) +INFO:tensorflow:global_step/sec: 0.167585 +INFO:tensorflow:step = 48401, loss = 0.198325, precision = 0.976562 (596.714 sec) +Saved checkpoint after 124 epoch(s) to data/resnet164/checkpoints/00124... +INFO:tensorflow:global_step/sec: 0.171653 +INFO:tensorflow:step = 48501, loss = 0.17414, precision = 0.992188 (582.570 sec) +INFO:tensorflow:global_step/sec: 0.170697 +INFO:tensorflow:step = 48601, loss = 0.169501, precision = 0.992188 (585.833 sec) +INFO:tensorflow:global_step/sec: 0.171675 +INFO:tensorflow:step = 48701, loss = 0.201067, precision = 0.976562 (582.497 sec) +INFO:tensorflow:global_step/sec: 0.17012 +INFO:tensorflow:step = 48801, loss = 0.22606, precision = 0.976562 (587.821 sec) +Saved checkpoint after 125 epoch(s) to data/resnet164/checkpoints/00125... +INFO:tensorflow:global_step/sec: 0.166879 +INFO:tensorflow:step = 48901, loss = 0.178611, precision = 0.992188 (599.238 sec) +INFO:tensorflow:global_step/sec: 0.166888 +INFO:tensorflow:step = 49001, loss = 0.223888, precision = 0.976562 (599.204 sec) +INFO:tensorflow:global_step/sec: 0.165093 +INFO:tensorflow:step = 49101, loss = 0.163666, precision = 1.0 (605.717 sec) +INFO:tensorflow:global_step/sec: 0.16806 +INFO:tensorflow:step = 49201, loss = 0.268793, precision = 0.9375 (595.027 sec) +Saved checkpoint after 126 epoch(s) to data/resnet164/checkpoints/00126... +INFO:tensorflow:global_step/sec: 0.166442 +INFO:tensorflow:step = 49301, loss = 0.194236, precision = 0.984375 (600.812 sec) +INFO:tensorflow:global_step/sec: 0.172176 +INFO:tensorflow:step = 49401, loss = 0.184374, precision = 0.984375 (580.802 sec) +INFO:tensorflow:global_step/sec: 0.171251 +INFO:tensorflow:step = 49501, loss = 0.203193, precision = 0.984375 (583.938 sec) +INFO:tensorflow:global_step/sec: 0.172128 +INFO:tensorflow:step = 49601, loss = 0.186931, precision = 0.984375 (580.965 sec) +Saved checkpoint after 127 epoch(s) to data/resnet164/checkpoints/00127... +INFO:tensorflow:global_step/sec: 0.172708 +INFO:tensorflow:step = 49701, loss = 0.169161, precision = 1.0 (579.013 sec) +INFO:tensorflow:global_step/sec: 0.172502 +INFO:tensorflow:step = 49801, loss = 0.215088, precision = 0.976562 (579.705 sec) +INFO:tensorflow:global_step/sec: 0.169543 +INFO:tensorflow:step = 49901, loss = 0.232127, precision = 0.976562 (589.821 sec) +INFO:tensorflow:global_step/sec: 0.171912 +INFO:tensorflow:step = 50001, loss = 0.168962, precision = 0.992188 (581.691 sec) +Saved checkpoint after 128 epoch(s) to data/resnet164/checkpoints/00128... +INFO:tensorflow:global_step/sec: 0.17505 +INFO:tensorflow:step = 50101, loss = 0.269898, precision = 0.953125 (571.265 sec) +INFO:tensorflow:global_step/sec: 0.173661 +INFO:tensorflow:step = 50201, loss = 0.217865, precision = 0.976562 (575.836 sec) +INFO:tensorflow:global_step/sec: 0.172991 +INFO:tensorflow:step = 50301, loss = 0.171764, precision = 0.992188 (578.063 sec) +INFO:tensorflow:global_step/sec: 0.172385 +INFO:tensorflow:step = 50401, loss = 0.175336, precision = 0.992188 (580.096 sec) +Saved checkpoint after 129 epoch(s) to data/resnet164/checkpoints/00129... +INFO:tensorflow:global_step/sec: 0.175372 +INFO:tensorflow:step = 50501, loss = 0.184864, precision = 0.976562 (570.217 sec) +INFO:tensorflow:global_step/sec: 0.17448 +INFO:tensorflow:step = 50601, loss = 0.233478, precision = 0.976562 (573.133 sec) +INFO:tensorflow:global_step/sec: 0.169171 +INFO:tensorflow:step = 50701, loss = 0.209685, precision = 0.976562 (591.116 sec) +INFO:tensorflow:global_step/sec: 0.17443 +INFO:tensorflow:step = 50801, loss = 0.187804, precision = 0.984375 (573.296 sec) +Saved checkpoint after 130 epoch(s) to data/resnet164/checkpoints/00130... +INFO:tensorflow:global_step/sec: 0.173816 +INFO:tensorflow:step = 50901, loss = 0.181285, precision = 0.984375 (575.322 sec) +INFO:tensorflow:global_step/sec: 0.166783 +INFO:tensorflow:step = 51001, loss = 0.18758, precision = 0.984375 (599.580 sec) +INFO:tensorflow:global_step/sec: 0.166152 +INFO:tensorflow:step = 51101, loss = 0.2218, precision = 0.960938 (601.859 sec) +INFO:tensorflow:global_step/sec: 0.167036 +INFO:tensorflow:step = 51201, loss = 0.1811, precision = 0.984375 (598.675 sec) +Saved checkpoint after 131 epoch(s) to data/resnet164/checkpoints/00131... +INFO:tensorflow:global_step/sec: 0.171665 +INFO:tensorflow:step = 51301, loss = 0.190896, precision = 0.976562 (582.531 sec) +INFO:tensorflow:global_step/sec: 0.177076 +INFO:tensorflow:step = 51401, loss = 0.21587, precision = 0.976562 (564.728 sec) +INFO:tensorflow:global_step/sec: 0.173225 +INFO:tensorflow:step = 51501, loss = 0.212346, precision = 0.96875 (577.285 sec) +INFO:tensorflow:global_step/sec: 0.167679 +INFO:tensorflow:step = 51601, loss = 0.169703, precision = 0.984375 (596.379 sec) +Saved checkpoint after 132 epoch(s) to data/resnet164/checkpoints/00132... +INFO:tensorflow:global_step/sec: 0.16789 +INFO:tensorflow:step = 51701, loss = 0.296754, precision = 0.929688 (595.629 sec) +INFO:tensorflow:global_step/sec: 0.173906 +INFO:tensorflow:step = 51801, loss = 0.222379, precision = 0.96875 (575.024 sec) +INFO:tensorflow:global_step/sec: 0.175746 +INFO:tensorflow:step = 51901, loss = 0.211144, precision = 0.96875 (569.003 sec) +INFO:tensorflow:global_step/sec: 0.166252 +INFO:tensorflow:step = 52001, loss = 0.192982, precision = 0.96875 (601.497 sec) +Saved checkpoint after 133 epoch(s) to data/resnet164/checkpoints/00133... +INFO:tensorflow:global_step/sec: 0.165752 +INFO:tensorflow:step = 52101, loss = 0.260266, precision = 0.960938 (603.313 sec) +INFO:tensorflow:global_step/sec: 0.166654 +INFO:tensorflow:step = 52201, loss = 0.21482, precision = 0.96875 (600.046 sec) +INFO:tensorflow:global_step/sec: 0.167275 +INFO:tensorflow:step = 52301, loss = 0.200132, precision = 0.984375 (597.819 sec) +Saved checkpoint after 134 epoch(s) to data/resnet164/checkpoints/00134... +INFO:tensorflow:global_step/sec: 0.17011 +INFO:tensorflow:step = 52401, loss = 0.183047, precision = 0.984375 (587.854 sec) +INFO:tensorflow:global_step/sec: 0.172445 +INFO:tensorflow:step = 52501, loss = 0.165739, precision = 1.0 (579.895 sec) +INFO:tensorflow:global_step/sec: 0.170386 +INFO:tensorflow:step = 52601, loss = 0.207892, precision = 0.976562 (586.902 sec) +INFO:tensorflow:global_step/sec: 0.167904 +INFO:tensorflow:step = 52701, loss = 0.211984, precision = 0.976562 (595.580 sec) +Saved checkpoint after 135 epoch(s) to data/resnet164/checkpoints/00135... +INFO:tensorflow:global_step/sec: 0.166213 +INFO:tensorflow:step = 52801, loss = 0.220833, precision = 0.960938 (601.637 sec) +INFO:tensorflow:global_step/sec: 0.166391 +INFO:tensorflow:step = 52901, loss = 0.178655, precision = 0.984375 (600.996 sec) +INFO:tensorflow:global_step/sec: 0.169371 +INFO:tensorflow:step = 53001, loss = 0.191057, precision = 0.984375 (590.420 sec) +INFO:tensorflow:global_step/sec: 0.167756 +INFO:tensorflow:step = 53101, loss = 0.184296, precision = 0.976562 (596.104 sec) +Saved checkpoint after 136 epoch(s) to data/resnet164/checkpoints/00136... +INFO:tensorflow:global_step/sec: 0.167729 +INFO:tensorflow:step = 53201, loss = 0.205579, precision = 0.976562 (596.201 sec) +INFO:tensorflow:global_step/sec: 0.165971 +INFO:tensorflow:step = 53301, loss = 0.176558, precision = 0.992188 (602.516 sec) +INFO:tensorflow:global_step/sec: 0.163492 +INFO:tensorflow:step = 53401, loss = 0.157301, precision = 0.992188 (611.650 sec) +INFO:tensorflow:global_step/sec: 0.163523 +INFO:tensorflow:step = 53501, loss = 0.169925, precision = 0.984375 (611.536 sec) +Saved checkpoint after 137 epoch(s) to data/resnet164/checkpoints/00137... +INFO:tensorflow:global_step/sec: 0.16415 +INFO:tensorflow:step = 53601, loss = 0.147915, precision = 1.0 (609.197 sec) +INFO:tensorflow:global_step/sec: 0.167746 +INFO:tensorflow:step = 53701, loss = 0.164523, precision = 0.992188 (596.141 sec) +INFO:tensorflow:global_step/sec: 0.164634 +INFO:tensorflow:step = 53801, loss = 0.169396, precision = 0.992188 (607.408 sec) +INFO:tensorflow:global_step/sec: 0.166266 +INFO:tensorflow:step = 53901, loss = 0.159112, precision = 0.992188 (601.444 sec) +Saved checkpoint after 138 epoch(s) to data/resnet164/checkpoints/00138... +INFO:tensorflow:global_step/sec: 0.16685 +INFO:tensorflow:step = 54001, loss = 0.151686, precision = 0.992188 (599.342 sec) +INFO:tensorflow:global_step/sec: 0.168446 +INFO:tensorflow:step = 54101, loss = 0.145383, precision = 1.0 (593.662 sec) +INFO:tensorflow:global_step/sec: 0.165542 +INFO:tensorflow:step = 54201, loss = 0.150761, precision = 1.0 (604.075 sec) +INFO:tensorflow:global_step/sec: 0.161501 +INFO:tensorflow:step = 54301, loss = 0.141963, precision = 1.0 (619.190 sec) +Saved checkpoint after 139 epoch(s) to data/resnet164/checkpoints/00139... +INFO:tensorflow:global_step/sec: 0.158431 +INFO:tensorflow:step = 54401, loss = 0.142247, precision = 1.0 (631.190 sec) +INFO:tensorflow:global_step/sec: 0.159172 +INFO:tensorflow:step = 54501, loss = 0.16016, precision = 0.984375 (628.251 sec) +INFO:tensorflow:global_step/sec: 0.159588 +INFO:tensorflow:step = 54601, loss = 0.139602, precision = 1.0 (626.614 sec) +INFO:tensorflow:global_step/sec: 0.160593 +INFO:tensorflow:step = 54701, loss = 0.142486, precision = 1.0 (622.694 sec) +Saved checkpoint after 140 epoch(s) to data/resnet164/checkpoints/00140... +INFO:tensorflow:global_step/sec: 0.162845 +INFO:tensorflow:step = 54801, loss = 0.14201, precision = 1.0 (614.082 sec) +INFO:tensorflow:global_step/sec: 0.158506 +INFO:tensorflow:step = 54901, loss = 0.144896, precision = 1.0 (630.891 sec) +INFO:tensorflow:global_step/sec: 0.163246 +INFO:tensorflow:step = 55001, loss = 0.138739, precision = 1.0 (612.573 sec) +INFO:tensorflow:global_step/sec: 0.165264 +INFO:tensorflow:step = 55101, loss = 0.138772, precision = 1.0 (605.091 sec) +Saved checkpoint after 141 epoch(s) to data/resnet164/checkpoints/00141... +INFO:tensorflow:global_step/sec: 0.165163 +INFO:tensorflow:step = 55201, loss = 0.137307, precision = 1.0 (605.463 sec) +INFO:tensorflow:global_step/sec: 0.164893 +INFO:tensorflow:step = 55301, loss = 0.142073, precision = 1.0 (606.455 sec) +INFO:tensorflow:global_step/sec: 0.167235 +INFO:tensorflow:step = 55401, loss = 0.14611, precision = 0.992188 (597.959 sec) +INFO:tensorflow:global_step/sec: 0.170792 +INFO:tensorflow:step = 55501, loss = 0.146614, precision = 0.992188 (585.508 sec) +Saved checkpoint after 142 epoch(s) to data/resnet164/checkpoints/00142... +INFO:tensorflow:global_step/sec: 0.16514 +INFO:tensorflow:step = 55601, loss = 0.138788, precision = 1.0 (605.545 sec) +INFO:tensorflow:global_step/sec: 0.163703 +INFO:tensorflow:step = 55701, loss = 0.137471, precision = 1.0 (610.864 sec) +INFO:tensorflow:global_step/sec: 0.16086 +INFO:tensorflow:step = 55801, loss = 0.137769, precision = 1.0 (621.657 sec) +INFO:tensorflow:global_step/sec: 0.169353 +INFO:tensorflow:step = 55901, loss = 0.136643, precision = 1.0 (590.482 sec) +Saved checkpoint after 143 epoch(s) to data/resnet164/checkpoints/00143... +INFO:tensorflow:global_step/sec: 0.168804 +INFO:tensorflow:step = 56001, loss = 0.136646, precision = 1.0 (592.402 sec) +INFO:tensorflow:global_step/sec: 0.167768 +INFO:tensorflow:step = 56101, loss = 0.137916, precision = 1.0 (596.063 sec) +INFO:tensorflow:global_step/sec: 0.168261 +INFO:tensorflow:step = 56201, loss = 0.138764, precision = 1.0 (594.315 sec) +INFO:tensorflow:global_step/sec: 0.173029 +INFO:tensorflow:step = 56301, loss = 0.14287, precision = 1.0 (577.939 sec) +Saved checkpoint after 144 epoch(s) to data/resnet164/checkpoints/00144... +INFO:tensorflow:global_step/sec: 0.16453 +INFO:tensorflow:step = 56401, loss = 0.143107, precision = 1.0 (607.790 sec) +INFO:tensorflow:global_step/sec: 0.16775 +INFO:tensorflow:step = 56501, loss = 0.137004, precision = 1.0 (596.125 sec) +INFO:tensorflow:global_step/sec: 0.170014 +INFO:tensorflow:step = 56601, loss = 0.151359, precision = 0.992188 (588.188 sec) +Saved checkpoint after 145 epoch(s) to data/resnet164/checkpoints/00145... +INFO:tensorflow:global_step/sec: 0.168365 +INFO:tensorflow:step = 56701, loss = 0.139897, precision = 1.0 (593.946 sec) +INFO:tensorflow:global_step/sec: 0.177649 +INFO:tensorflow:step = 56801, loss = 0.136814, precision = 1.0 (562.906 sec) +INFO:tensorflow:global_step/sec: 0.168089 +INFO:tensorflow:step = 56901, loss = 0.136149, precision = 1.0 (594.922 sec) +INFO:tensorflow:global_step/sec: 0.16553 +INFO:tensorflow:step = 57001, loss = 0.134206, precision = 1.0 (604.121 sec) +Saved checkpoint after 146 epoch(s) to data/resnet164/checkpoints/00146... +INFO:tensorflow:global_step/sec: 0.170546 +INFO:tensorflow:step = 57101, loss = 0.152236, precision = 0.992188 (586.353 sec) +INFO:tensorflow:global_step/sec: 0.15779 +INFO:tensorflow:step = 57201, loss = 0.136463, precision = 1.0 (633.752 sec) +INFO:tensorflow:global_step/sec: 0.159433 +INFO:tensorflow:step = 57301, loss = 0.139814, precision = 1.0 (627.223 sec) +INFO:tensorflow:global_step/sec: 0.163949 +INFO:tensorflow:step = 57401, loss = 0.145433, precision = 0.992188 (609.946 sec) +Saved checkpoint after 147 epoch(s) to data/resnet164/checkpoints/00147... +INFO:tensorflow:global_step/sec: 0.163002 +INFO:tensorflow:step = 57501, loss = 0.134084, precision = 1.0 (613.489 sec) +INFO:tensorflow:global_step/sec: 0.164089 +INFO:tensorflow:step = 57601, loss = 0.139204, precision = 1.0 (609.424 sec) +INFO:tensorflow:global_step/sec: 0.159071 +INFO:tensorflow:step = 57701, loss = 0.146182, precision = 0.992188 (628.651 sec) +INFO:tensorflow:global_step/sec: 0.162734 +INFO:tensorflow:step = 57801, loss = 0.142919, precision = 1.0 (614.498 sec) +Saved checkpoint after 148 epoch(s) to data/resnet164/checkpoints/00148... +INFO:tensorflow:global_step/sec: 0.16461 +INFO:tensorflow:step = 57901, loss = 0.135809, precision = 1.0 (607.496 sec) +INFO:tensorflow:global_step/sec: 0.16845 +INFO:tensorflow:step = 58001, loss = 0.134559, precision = 1.0 (593.648 sec) +INFO:tensorflow:global_step/sec: 0.165697 +INFO:tensorflow:step = 58101, loss = 0.133724, precision = 1.0 (603.512 sec) +INFO:tensorflow:global_step/sec: 0.166661 +INFO:tensorflow:step = 58201, loss = 0.136079, precision = 1.0 (600.019 sec) +Saved checkpoint after 149 epoch(s) to data/resnet164/checkpoints/00149... +INFO:tensorflow:global_step/sec: 0.168924 +INFO:tensorflow:step = 58301, loss = 0.136274, precision = 1.0 (591.983 sec) +INFO:tensorflow:global_step/sec: 0.169389 +INFO:tensorflow:step = 58401, loss = 0.193987, precision = 0.992188 (590.358 sec) +INFO:tensorflow:global_step/sec: 0.166361 +INFO:tensorflow:step = 58501, loss = 0.134455, precision = 1.0 (601.103 sec) +INFO:tensorflow:global_step/sec: 0.16671 +INFO:tensorflow:step = 58601, loss = 0.133711, precision = 1.0 (599.846 sec) +Saved checkpoint after 150 epoch(s) to data/resnet164/checkpoints/00150... +INFO:tensorflow:global_step/sec: 0.1661 +INFO:tensorflow:step = 58701, loss = 0.131755, precision = 1.0 (602.045 sec) +INFO:tensorflow:global_step/sec: 0.168455 +INFO:tensorflow:step = 58801, loss = 0.137042, precision = 1.0 (593.631 sec) +INFO:tensorflow:global_step/sec: 0.166456 +INFO:tensorflow:step = 58901, loss = 0.137165, precision = 0.992188 (600.758 sec) +INFO:tensorflow:global_step/sec: 0.16653 +INFO:tensorflow:step = 59001, loss = 0.135594, precision = 1.0 (600.491 sec) +Saved checkpoint after 151 epoch(s) to data/resnet164/checkpoints/00151... +INFO:tensorflow:global_step/sec: 0.164641 +INFO:tensorflow:step = 59101, loss = 0.131853, precision = 1.0 (607.380 sec) +INFO:tensorflow:global_step/sec: 0.168546 +INFO:tensorflow:step = 59201, loss = 0.141404, precision = 0.992188 (593.309 sec) +INFO:tensorflow:global_step/sec: 0.167509 +INFO:tensorflow:step = 59301, loss = 0.130931, precision = 1.0 (596.984 sec) +INFO:tensorflow:global_step/sec: 0.160551 +INFO:tensorflow:step = 59401, loss = 0.130825, precision = 1.0 (622.856 sec) +Saved checkpoint after 152 epoch(s) to data/resnet164/checkpoints/00152... +INFO:tensorflow:global_step/sec: 0.16798 +INFO:tensorflow:step = 59501, loss = 0.131231, precision = 1.0 (595.310 sec) +INFO:tensorflow:global_step/sec: 0.164182 +INFO:tensorflow:step = 59601, loss = 0.130949, precision = 1.0 (609.081 sec) +INFO:tensorflow:global_step/sec: 0.166738 +INFO:tensorflow:step = 59701, loss = 0.137796, precision = 1.0 (599.742 sec) +INFO:tensorflow:global_step/sec: 0.161398 +INFO:tensorflow:step = 59801, loss = 0.132359, precision = 1.0 (619.586 sec) +Saved checkpoint after 153 epoch(s) to data/resnet164/checkpoints/00153... +INFO:tensorflow:global_step/sec: 0.167592 +INFO:tensorflow:step = 59901, loss = 0.131357, precision = 1.0 (596.688 sec) +INFO:tensorflow:global_step/sec: 0.15785 +INFO:tensorflow:step = 60001, loss = 0.1331, precision = 1.0 (633.511 sec) +INFO:tensorflow:global_step/sec: 0.158282 +INFO:tensorflow:step = 60101, loss = 0.138149, precision = 0.992188 (631.782 sec) +INFO:tensorflow:global_step/sec: 0.161246 +INFO:tensorflow:step = 60201, loss = 0.136003, precision = 1.0 (620.172 sec) +Saved checkpoint after 154 epoch(s) to data/resnet164/checkpoints/00154... +INFO:tensorflow:global_step/sec: 0.16528 +INFO:tensorflow:step = 60301, loss = 0.130213, precision = 1.0 (605.033 sec) +INFO:tensorflow:global_step/sec: 0.161683 +INFO:tensorflow:step = 60401, loss = 0.13062, precision = 1.0 (618.493 sec) +INFO:tensorflow:global_step/sec: 0.164099 +INFO:tensorflow:step = 60501, loss = 0.13408, precision = 1.0 (609.389 sec) +INFO:tensorflow:global_step/sec: 0.164615 +INFO:tensorflow:step = 60601, loss = 0.129603, precision = 1.0 (607.479 sec) +Saved checkpoint after 155 epoch(s) to data/resnet164/checkpoints/00155... +INFO:tensorflow:global_step/sec: 0.16082 +INFO:tensorflow:step = 60701, loss = 0.130191, precision = 1.0 (621.813 sec) +INFO:tensorflow:global_step/sec: 0.16049 +INFO:tensorflow:step = 60801, loss = 0.130163, precision = 1.0 (623.092 sec) +INFO:tensorflow:global_step/sec: 0.159255 +INFO:tensorflow:step = 60901, loss = 0.130489, precision = 1.0 (627.926 sec) +Saved checkpoint after 156 epoch(s) to data/resnet164/checkpoints/00156... +INFO:tensorflow:global_step/sec: 0.164623 +INFO:tensorflow:step = 61001, loss = 0.136189, precision = 1.0 (607.447 sec) +INFO:tensorflow:global_step/sec: 0.160761 +INFO:tensorflow:step = 61101, loss = 0.129851, precision = 1.0 (622.041 sec) +INFO:tensorflow:global_step/sec: 0.160093 +INFO:tensorflow:step = 61201, loss = 0.128919, precision = 1.0 (624.639 sec) +INFO:tensorflow:global_step/sec: 0.158147 +INFO:tensorflow:step = 61301, loss = 0.129699, precision = 1.0 (632.324 sec) +Saved checkpoint after 157 epoch(s) to data/resnet164/checkpoints/00157... +INFO:tensorflow:global_step/sec: 0.156021 +INFO:tensorflow:step = 61401, loss = 0.129869, precision = 1.0 (640.940 sec) +INFO:tensorflow:global_step/sec: 0.153563 +INFO:tensorflow:step = 61501, loss = 0.128252, precision = 1.0 (651.200 sec) +INFO:tensorflow:global_step/sec: 0.161288 +INFO:tensorflow:step = 61601, loss = 0.1303, precision = 1.0 (620.011 sec) +INFO:tensorflow:global_step/sec: 0.150774 +INFO:tensorflow:step = 61701, loss = 0.133797, precision = 1.0 (663.245 sec) +Saved checkpoint after 158 epoch(s) to data/resnet164/checkpoints/00158... +INFO:tensorflow:global_step/sec: 0.152109 +INFO:tensorflow:step = 61801, loss = 0.127714, precision = 1.0 (657.422 sec) +INFO:tensorflow:global_step/sec: 0.166283 +INFO:tensorflow:step = 61901, loss = 0.132208, precision = 1.0 (601.384 sec) +INFO:tensorflow:global_step/sec: 0.166957 +INFO:tensorflow:step = 62001, loss = 0.128571, precision = 1.0 (598.958 sec) +INFO:tensorflow:global_step/sec: 0.155432 +INFO:tensorflow:step = 62101, loss = 0.130172, precision = 1.0 (643.366 sec) +Saved checkpoint after 159 epoch(s) to data/resnet164/checkpoints/00159... +INFO:tensorflow:global_step/sec: 0.164631 +INFO:tensorflow:step = 62201, loss = 0.128104, precision = 1.0 (607.419 sec) +INFO:tensorflow:global_step/sec: 0.167145 +INFO:tensorflow:step = 62301, loss = 0.140699, precision = 0.992188 (598.282 sec) +INFO:tensorflow:global_step/sec: 0.167747 +INFO:tensorflow:step = 62401, loss = 0.127167, precision = 1.0 (596.137 sec) +INFO:tensorflow:global_step/sec: 0.167976 +INFO:tensorflow:step = 62501, loss = 0.139835, precision = 0.992188 (595.325 sec) +Saved checkpoint after 160 epoch(s) to data/resnet164/checkpoints/00160... +INFO:tensorflow:global_step/sec: 0.168143 +INFO:tensorflow:step = 62601, loss = 0.127671, precision = 1.0 (594.733 sec) +INFO:tensorflow:global_step/sec: 0.16869 +INFO:tensorflow:step = 62701, loss = 0.128985, precision = 1.0 (592.802 sec) +INFO:tensorflow:global_step/sec: 0.170706 +INFO:tensorflow:step = 62801, loss = 0.12946, precision = 1.0 (585.801 sec) +INFO:tensorflow:global_step/sec: 0.17068 +INFO:tensorflow:step = 62901, loss = 0.128875, precision = 1.0 (585.891 sec) +Saved checkpoint after 161 epoch(s) to data/resnet164/checkpoints/00161... +INFO:tensorflow:global_step/sec: 0.170355 +INFO:tensorflow:step = 63001, loss = 0.126664, precision = 1.0 (587.011 sec) +INFO:tensorflow:global_step/sec: 0.170974 +INFO:tensorflow:step = 63101, loss = 0.136611, precision = 0.992188 (584.885 sec) +INFO:tensorflow:global_step/sec: 0.171519 +INFO:tensorflow:step = 63201, loss = 0.126933, precision = 1.0 (583.026 sec) +INFO:tensorflow:global_step/sec: 0.171213 +INFO:tensorflow:step = 63301, loss = 0.12622, precision = 1.0 (584.067 sec) +Saved checkpoint after 162 epoch(s) to data/resnet164/checkpoints/00162... +INFO:tensorflow:global_step/sec: 0.16978 +INFO:tensorflow:step = 63401, loss = 0.128186, precision = 1.0 (588.998 sec) +INFO:tensorflow:global_step/sec: 0.169805 +INFO:tensorflow:step = 63501, loss = 0.126354, precision = 1.0 (588.910 sec) +INFO:tensorflow:global_step/sec: 0.165589 +INFO:tensorflow:step = 63601, loss = 0.12597, precision = 1.0 (603.907 sec) +INFO:tensorflow:global_step/sec: 0.169724 +INFO:tensorflow:step = 63701, loss = 0.125902, precision = 1.0 (589.192 sec) +Saved checkpoint after 163 epoch(s) to data/resnet164/checkpoints/00163... +INFO:tensorflow:global_step/sec: 0.169876 +INFO:tensorflow:step = 63801, loss = 0.125287, precision = 1.0 (588.664 sec) +INFO:tensorflow:global_step/sec: 0.170811 +INFO:tensorflow:step = 63901, loss = 0.127511, precision = 1.0 (585.442 sec) +INFO:tensorflow:global_step/sec: 0.171867 +INFO:tensorflow:step = 64001, loss = 0.125638, precision = 1.0 (581.845 sec) +INFO:tensorflow:global_step/sec: 0.168696 +INFO:tensorflow:step = 64101, loss = 0.127926, precision = 1.0 (592.783 sec) +Saved checkpoint after 164 epoch(s) to data/resnet164/checkpoints/00164... +INFO:tensorflow:global_step/sec: 0.165452 +INFO:tensorflow:step = 64201, loss = 0.125479, precision = 1.0 (604.404 sec) +INFO:tensorflow:global_step/sec: 0.166294 +INFO:tensorflow:step = 64301, loss = 0.132034, precision = 1.0 (601.343 sec) +INFO:tensorflow:global_step/sec: 0.166631 +INFO:tensorflow:step = 64401, loss = 0.125772, precision = 1.0 (600.127 sec) +INFO:tensorflow:global_step/sec: 0.170779 +INFO:tensorflow:step = 64501, loss = 0.125138, precision = 1.0 (585.552 sec) +Saved checkpoint after 165 epoch(s) to data/resnet164/checkpoints/00165... +INFO:tensorflow:global_step/sec: 0.170943 +INFO:tensorflow:step = 64601, loss = 0.125224, precision = 1.0 (584.990 sec) +INFO:tensorflow:global_step/sec: 0.169927 +INFO:tensorflow:step = 64701, loss = 0.139476, precision = 0.992188 (588.488 sec) +INFO:tensorflow:global_step/sec: 0.170401 +INFO:tensorflow:step = 64801, loss = 0.1257, precision = 1.0 (586.852 sec) +INFO:tensorflow:global_step/sec: 0.170488 +INFO:tensorflow:step = 64901, loss = 0.123669, precision = 1.0 (586.553 sec) +Saved checkpoint after 166 epoch(s) to data/resnet164/checkpoints/00166... +INFO:tensorflow:global_step/sec: 0.170079 +INFO:tensorflow:step = 65001, loss = 0.12382, precision = 1.0 (587.961 sec) +INFO:tensorflow:global_step/sec: 0.167263 +INFO:tensorflow:step = 65101, loss = 0.123987, precision = 1.0 (597.861 sec) +INFO:tensorflow:global_step/sec: 0.164749 +INFO:tensorflow:step = 65201, loss = 0.128943, precision = 1.0 (606.983 sec) +Saved checkpoint after 167 epoch(s) to data/resnet164/checkpoints/00167... +INFO:tensorflow:global_step/sec: 0.165994 +INFO:tensorflow:step = 65301, loss = 0.124334, precision = 1.0 (602.431 sec) +INFO:tensorflow:global_step/sec: 0.167552 +INFO:tensorflow:step = 65401, loss = 0.125926, precision = 1.0 (596.829 sec) +INFO:tensorflow:global_step/sec: 0.168366 +INFO:tensorflow:step = 65501, loss = 0.124995, precision = 1.0 (593.945 sec) +INFO:tensorflow:global_step/sec: 0.170892 +INFO:tensorflow:step = 65601, loss = 0.123871, precision = 1.0 (585.166 sec) +Saved checkpoint after 168 epoch(s) to data/resnet164/checkpoints/00168... +INFO:tensorflow:global_step/sec: 0.167743 +INFO:tensorflow:step = 65701, loss = 0.125523, precision = 1.0 (596.152 sec) +INFO:tensorflow:global_step/sec: 0.172142 +INFO:tensorflow:step = 65801, loss = 0.125982, precision = 1.0 (580.917 sec) +INFO:tensorflow:global_step/sec: 0.172981 +INFO:tensorflow:step = 65901, loss = 0.130753, precision = 1.0 (578.097 sec) +INFO:tensorflow:global_step/sec: 0.176297 +INFO:tensorflow:step = 66001, loss = 0.123623, precision = 1.0 (567.225 sec) +Saved checkpoint after 169 epoch(s) to data/resnet164/checkpoints/00169... +INFO:tensorflow:global_step/sec: 0.175561 +INFO:tensorflow:step = 66101, loss = 0.122946, precision = 1.0 (569.602 sec) +INFO:tensorflow:global_step/sec: 0.173334 +INFO:tensorflow:step = 66201, loss = 0.124555, precision = 1.0 (576.921 sec) +INFO:tensorflow:global_step/sec: 0.169121 +INFO:tensorflow:step = 66301, loss = 0.129624, precision = 0.992188 (591.291 sec) +INFO:tensorflow:global_step/sec: 0.168286 +INFO:tensorflow:step = 66401, loss = 0.122186, precision = 1.0 (594.225 sec) +Saved checkpoint after 170 epoch(s) to data/resnet164/checkpoints/00170... +INFO:tensorflow:global_step/sec: 0.173427 +INFO:tensorflow:step = 66501, loss = 0.122199, precision = 1.0 (576.612 sec) +INFO:tensorflow:global_step/sec: 0.17579 +INFO:tensorflow:step = 66601, loss = 0.122057, precision = 1.0 (568.859 sec) +INFO:tensorflow:global_step/sec: 0.175329 +INFO:tensorflow:step = 66701, loss = 0.121654, precision = 1.0 (570.356 sec) +INFO:tensorflow:global_step/sec: 0.174318 +INFO:tensorflow:step = 66801, loss = 0.122633, precision = 1.0 (573.664 sec) +Saved checkpoint after 171 epoch(s) to data/resnet164/checkpoints/00171... +INFO:tensorflow:global_step/sec: 0.172593 +INFO:tensorflow:step = 66901, loss = 0.131395, precision = 0.992188 (579.398 sec) +INFO:tensorflow:global_step/sec: 0.172098 +INFO:tensorflow:step = 67001, loss = 0.132049, precision = 0.992188 (581.066 sec) +INFO:tensorflow:global_step/sec: 0.171692 +INFO:tensorflow:step = 67101, loss = 0.126531, precision = 1.0 (582.439 sec) +INFO:tensorflow:global_step/sec: 0.171669 +INFO:tensorflow:step = 67201, loss = 0.122965, precision = 1.0 (582.518 sec) +Saved checkpoint after 172 epoch(s) to data/resnet164/checkpoints/00172... +INFO:tensorflow:global_step/sec: 0.170588 +INFO:tensorflow:step = 67301, loss = 0.136456, precision = 0.992188 (586.207 sec) +INFO:tensorflow:global_step/sec: 0.171221 +INFO:tensorflow:step = 67401, loss = 0.125502, precision = 1.0 (584.041 sec) +INFO:tensorflow:global_step/sec: 0.167081 +INFO:tensorflow:step = 67501, loss = 0.120653, precision = 1.0 (598.514 sec) +INFO:tensorflow:global_step/sec: 0.164924 +INFO:tensorflow:step = 67601, loss = 0.121524, precision = 1.0 (606.339 sec) +Saved checkpoint after 173 epoch(s) to data/resnet164/checkpoints/00173... +INFO:tensorflow:global_step/sec: 0.169638 +INFO:tensorflow:step = 67701, loss = 0.121157, precision = 1.0 (589.491 sec) +INFO:tensorflow:global_step/sec: 0.171164 +INFO:tensorflow:step = 67801, loss = 0.123708, precision = 1.0 (584.235 sec) +INFO:tensorflow:global_step/sec: 0.172727 +INFO:tensorflow:step = 67901, loss = 0.121378, precision = 1.0 (578.949 sec) +INFO:tensorflow:global_step/sec: 0.171215 +INFO:tensorflow:step = 68001, loss = 0.121801, precision = 1.0 (584.062 sec) +Saved checkpoint after 174 epoch(s) to data/resnet164/checkpoints/00174... +INFO:tensorflow:global_step/sec: 0.168448 +INFO:tensorflow:step = 68101, loss = 0.120701, precision = 1.0 (593.656 sec) +INFO:tensorflow:global_step/sec: 0.169327 +INFO:tensorflow:step = 68201, loss = 0.122105, precision = 1.0 (590.574 sec) +INFO:tensorflow:global_step/sec: 0.168869 +INFO:tensorflow:step = 68301, loss = 0.122639, precision = 1.0 (592.175 sec) +INFO:tensorflow:global_step/sec: 0.167544 +INFO:tensorflow:step = 68401, loss = 0.119868, precision = 1.0 (596.858 sec) +Saved checkpoint after 175 epoch(s) to data/resnet164/checkpoints/00175... +INFO:tensorflow:global_step/sec: 0.167476 +INFO:tensorflow:step = 68501, loss = 0.121182, precision = 1.0 (597.102 sec) +INFO:tensorflow:global_step/sec: 0.171978 +INFO:tensorflow:step = 68601, loss = 0.119446, precision = 1.0 (581.469 sec) +INFO:tensorflow:global_step/sec: 0.171438 +INFO:tensorflow:step = 68701, loss = 0.122059, precision = 1.0 (583.301 sec) +INFO:tensorflow:global_step/sec: 0.170946 +INFO:tensorflow:step = 68801, loss = 0.119128, precision = 1.0 (584.980 sec) +Saved checkpoint after 176 epoch(s) to data/resnet164/checkpoints/00176... +INFO:tensorflow:global_step/sec: 0.172481 +INFO:tensorflow:step = 68901, loss = 0.11995, precision = 1.0 (579.775 sec) +INFO:tensorflow:global_step/sec: 0.174509 +INFO:tensorflow:step = 69001, loss = 0.120326, precision = 1.0 (573.036 sec) +INFO:tensorflow:global_step/sec: 0.174083 +INFO:tensorflow:step = 69101, loss = 0.120129, precision = 1.0 (574.440 sec) +INFO:tensorflow:global_step/sec: 0.171245 +INFO:tensorflow:step = 69201, loss = 0.130367, precision = 0.992188 (583.958 sec) +Saved checkpoint after 177 epoch(s) to data/resnet164/checkpoints/00177... +INFO:tensorflow:global_step/sec: 0.165114 +INFO:tensorflow:step = 69301, loss = 0.119188, precision = 1.0 (605.641 sec) +INFO:tensorflow:global_step/sec: 0.16419 +INFO:tensorflow:step = 69401, loss = 0.121885, precision = 1.0 (609.051 sec) +INFO:tensorflow:global_step/sec: 0.165023 +INFO:tensorflow:step = 69501, loss = 0.118558, precision = 1.0 (605.976 sec) +Saved checkpoint after 178 epoch(s) to data/resnet164/checkpoints/00178... +INFO:tensorflow:global_step/sec: 0.163273 +INFO:tensorflow:step = 69601, loss = 0.119653, precision = 1.0 (612.470 sec) +INFO:tensorflow:global_step/sec: 0.165833 +INFO:tensorflow:step = 69701, loss = 0.121428, precision = 1.0 (603.018 sec) +INFO:tensorflow:global_step/sec: 0.176964 +INFO:tensorflow:step = 69801, loss = 0.150091, precision = 0.984375 (565.088 sec) +INFO:tensorflow:global_step/sec: 0.178026 +INFO:tensorflow:step = 69901, loss = 0.119523, precision = 1.0 (561.715 sec) +Saved checkpoint after 179 epoch(s) to data/resnet164/checkpoints/00179... +INFO:tensorflow:global_step/sec: 0.177395 +INFO:tensorflow:step = 70001, loss = 0.118036, precision = 1.0 (563.714 sec) +INFO:tensorflow:global_step/sec: 0.176974 +INFO:tensorflow:step = 70101, loss = 0.11833, precision = 1.0 (565.053 sec) +INFO:tensorflow:global_step/sec: 0.168733 +INFO:tensorflow:step = 70201, loss = 0.11954, precision = 1.0 (592.653 sec) +INFO:tensorflow:global_step/sec: 0.171346 +INFO:tensorflow:step = 70301, loss = 0.119128, precision = 1.0 (583.613 sec) +Saved checkpoint after 180 epoch(s) to data/resnet164/checkpoints/00180... +INFO:tensorflow:global_step/sec: 0.194394 +INFO:tensorflow:step = 70401, loss = 0.118902, precision = 1.0 (514.419 sec) +INFO:tensorflow:global_step/sec: 0.243473 +INFO:tensorflow:step = 70501, loss = 0.117848, precision = 1.0 (410.724 sec) +INFO:tensorflow:global_step/sec: 0.24555 +INFO:tensorflow:step = 70601, loss = 0.149297, precision = 0.992188 (407.249 sec) +INFO:tensorflow:global_step/sec: 0.245123 +INFO:tensorflow:step = 70701, loss = 0.117407, precision = 1.0 (407.958 sec) +Saved checkpoint after 181 epoch(s) to data/resnet164/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet20_train.log b/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet20_train.log new file mode 100644 index 0000000..e89f3b4 --- /dev/null +++ b/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet20_train.log @@ -0,0 +1,1713 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/269.03k params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/10.38b flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + +======================End of Report========================== +2017-07-31 22:28:00.219979: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 16 visible devices +2017-07-31 22:28:00.228734: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x4ff94f0 executing computations on platform Host. Devices: +2017-07-31 22:28:00.228848: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +INFO:tensorflow:step = 1, loss = 2.68739, precision = 0.140625 +INFO:tensorflow:global_step/sec: 1.50751 +INFO:tensorflow:step = 101, loss = 2.11201, precision = 0.367188 (66.335 sec) +INFO:tensorflow:global_step/sec: 1.56967 +INFO:tensorflow:step = 201, loss = 1.8836, precision = 0.445312 (63.708 sec) +INFO:tensorflow:global_step/sec: 1.58839 +INFO:tensorflow:step = 301, loss = 1.77321, precision = 0.5 (62.957 sec) +total_params: 269034 +Saved checkpoint after 1 epoch(s) to data/resnet20/checkpoints/00001... +INFO:tensorflow:global_step/sec: 1.58882 +INFO:tensorflow:step = 401, loss = 1.89524, precision = 0.445312 (62.940 sec) +INFO:tensorflow:global_step/sec: 1.59028 +INFO:tensorflow:step = 501, loss = 1.72784, precision = 0.539062 (62.882 sec) +INFO:tensorflow:global_step/sec: 1.61074 +INFO:tensorflow:step = 601, loss = 1.40923, precision = 0.5625 (62.083 sec) +INFO:tensorflow:global_step/sec: 1.59717 +INFO:tensorflow:step = 701, loss = 1.37298, precision = 0.640625 (62.611 sec) +Saved checkpoint after 2 epoch(s) to data/resnet20/checkpoints/00002... +INFO:tensorflow:global_step/sec: 1.61964 +INFO:tensorflow:step = 801, loss = 1.3398, precision = 0.570312 (61.742 sec) +INFO:tensorflow:global_step/sec: 1.63165 +INFO:tensorflow:step = 901, loss = 1.1919, precision = 0.65625 (61.287 sec) +INFO:tensorflow:global_step/sec: 1.62037 +INFO:tensorflow:step = 1001, loss = 1.15625, precision = 0.640625 (61.714 sec) +INFO:tensorflow:global_step/sec: 1.58832 +INFO:tensorflow:step = 1101, loss = 1.23805, precision = 0.625 (62.960 sec) +Saved checkpoint after 3 epoch(s) to data/resnet20/checkpoints/00003... +INFO:tensorflow:global_step/sec: 1.63132 +INFO:tensorflow:step = 1201, loss = 1.05062, precision = 0.695312 (61.300 sec) +INFO:tensorflow:global_step/sec: 1.66836 +INFO:tensorflow:step = 1301, loss = 0.97333, precision = 0.757812 (59.939 sec) +INFO:tensorflow:global_step/sec: 1.60065 +INFO:tensorflow:step = 1401, loss = 1.01184, precision = 0.742188 (62.475 sec) +INFO:tensorflow:global_step/sec: 1.60104 +INFO:tensorflow:step = 1501, loss = 0.994624, precision = 0.726562 (62.459 sec) +Saved checkpoint after 4 epoch(s) to data/resnet20/checkpoints/00004... +INFO:tensorflow:global_step/sec: 1.59792 +INFO:tensorflow:step = 1601, loss = 0.841178, precision = 0.8125 (62.581 sec) +INFO:tensorflow:global_step/sec: 1.63965 +INFO:tensorflow:step = 1701, loss = 0.850485, precision = 0.789062 (60.988 sec) +INFO:tensorflow:global_step/sec: 1.65218 +INFO:tensorflow:step = 1801, loss = 0.880397, precision = 0.765625 (60.526 sec) +INFO:tensorflow:global_step/sec: 1.60301 +INFO:tensorflow:step = 1901, loss = 0.956733, precision = 0.71875 (62.382 sec) +Saved checkpoint after 5 epoch(s) to data/resnet20/checkpoints/00005... +INFO:tensorflow:global_step/sec: 1.591 +INFO:tensorflow:step = 2001, loss = 1.0818, precision = 0.6875 (62.854 sec) +INFO:tensorflow:global_step/sec: 1.59334 +INFO:tensorflow:step = 2101, loss = 0.932756, precision = 0.78125 (62.761 sec) +INFO:tensorflow:global_step/sec: 1.58238 +INFO:tensorflow:step = 2201, loss = 0.889963, precision = 0.773438 (63.196 sec) +INFO:tensorflow:global_step/sec: 1.59999 +INFO:tensorflow:step = 2301, loss = 0.884394, precision = 0.78125 (62.500 sec) +Saved checkpoint after 6 epoch(s) to data/resnet20/checkpoints/00006... +INFO:tensorflow:global_step/sec: 1.59277 +INFO:tensorflow:step = 2401, loss = 0.742575, precision = 0.804688 (62.784 sec) +INFO:tensorflow:global_step/sec: 1.61955 +INFO:tensorflow:step = 2501, loss = 0.989218, precision = 0.742188 (61.745 sec) +INFO:tensorflow:global_step/sec: 1.62491 +INFO:tensorflow:step = 2601, loss = 0.800529, precision = 0.765625 (61.542 sec) +INFO:tensorflow:global_step/sec: 1.62666 +INFO:tensorflow:step = 2701, loss = 0.824227, precision = 0.796875 (61.475 sec) +Saved checkpoint after 7 epoch(s) to data/resnet20/checkpoints/00007... +INFO:tensorflow:global_step/sec: 1.61356 +INFO:tensorflow:step = 2801, loss = 0.81128, precision = 0.78125 (61.975 sec) +INFO:tensorflow:global_step/sec: 1.6579 +INFO:tensorflow:step = 2901, loss = 0.921155, precision = 0.78125 (60.317 sec) +INFO:tensorflow:global_step/sec: 1.63652 +INFO:tensorflow:step = 3001, loss = 0.742619, precision = 0.8125 (61.105 sec) +INFO:tensorflow:global_step/sec: 1.66427 +INFO:tensorflow:step = 3101, loss = 1.05875, precision = 0.710938 (60.086 sec) +Saved checkpoint after 8 epoch(s) to data/resnet20/checkpoints/00008... +INFO:tensorflow:global_step/sec: 1.68024 +INFO:tensorflow:step = 3201, loss = 0.768693, precision = 0.8125 (59.515 sec) +INFO:tensorflow:global_step/sec: 1.68902 +INFO:tensorflow:step = 3301, loss = 0.637507, precision = 0.875 (59.206 sec) +INFO:tensorflow:global_step/sec: 1.65528 +INFO:tensorflow:step = 3401, loss = 0.73582, precision = 0.8125 (60.413 sec) +INFO:tensorflow:global_step/sec: 1.67962 +INFO:tensorflow:step = 3501, loss = 0.848241, precision = 0.789062 (59.537 sec) +Saved checkpoint after 9 epoch(s) to data/resnet20/checkpoints/00009... +INFO:tensorflow:global_step/sec: 1.68907 +INFO:tensorflow:step = 3601, loss = 0.67343, precision = 0.820312 (59.204 sec) +INFO:tensorflow:global_step/sec: 1.67895 +INFO:tensorflow:step = 3701, loss = 0.87014, precision = 0.789062 (59.561 sec) +INFO:tensorflow:global_step/sec: 1.71983 +INFO:tensorflow:step = 3801, loss = 0.794875, precision = 0.804688 (58.145 sec) +INFO:tensorflow:global_step/sec: 1.69097 +INFO:tensorflow:step = 3901, loss = 0.822678, precision = 0.804688 (59.138 sec) +Saved checkpoint after 10 epoch(s) to data/resnet20/checkpoints/00010... +INFO:tensorflow:global_step/sec: 1.67492 +INFO:tensorflow:step = 4001, loss = 0.717259, precision = 0.828125 (59.705 sec) +INFO:tensorflow:global_step/sec: 1.69399 +INFO:tensorflow:step = 4101, loss = 0.707552, precision = 0.859375 (59.032 sec) +INFO:tensorflow:global_step/sec: 1.69495 +INFO:tensorflow:step = 4201, loss = 0.65864, precision = 0.84375 (58.999 sec) +Saved checkpoint after 11 epoch(s) to data/resnet20/checkpoints/00011... +INFO:tensorflow:global_step/sec: 1.6473 +INFO:tensorflow:step = 4301, loss = 0.690976, precision = 0.828125 (60.705 sec) +INFO:tensorflow:global_step/sec: 1.61365 +INFO:tensorflow:step = 4401, loss = 0.785919, precision = 0.78125 (61.972 sec) +INFO:tensorflow:global_step/sec: 1.61278 +INFO:tensorflow:step = 4501, loss = 0.691547, precision = 0.859375 (62.005 sec) +INFO:tensorflow:global_step/sec: 1.67931 +INFO:tensorflow:step = 4601, loss = 0.778251, precision = 0.78125 (59.548 sec) +Saved checkpoint after 12 epoch(s) to data/resnet20/checkpoints/00012... +INFO:tensorflow:global_step/sec: 1.67415 +INFO:tensorflow:step = 4701, loss = 0.835066, precision = 0.773438 (59.732 sec) +INFO:tensorflow:global_step/sec: 1.6563 +INFO:tensorflow:step = 4801, loss = 0.785123, precision = 0.796875 (60.376 sec) +INFO:tensorflow:global_step/sec: 1.62377 +INFO:tensorflow:step = 4901, loss = 0.72395, precision = 0.835938 (61.585 sec) +INFO:tensorflow:global_step/sec: 1.62132 +INFO:tensorflow:step = 5001, loss = 0.774649, precision = 0.796875 (61.678 sec) +Saved checkpoint after 13 epoch(s) to data/resnet20/checkpoints/00013... +INFO:tensorflow:global_step/sec: 1.70354 +INFO:tensorflow:step = 5101, loss = 0.738118, precision = 0.84375 (58.701 sec) +INFO:tensorflow:global_step/sec: 1.71367 +INFO:tensorflow:step = 5201, loss = 0.720436, precision = 0.851562 (58.354 sec) +INFO:tensorflow:global_step/sec: 1.71551 +INFO:tensorflow:step = 5301, loss = 0.88972, precision = 0.757812 (58.292 sec) +INFO:tensorflow:global_step/sec: 1.68098 +INFO:tensorflow:step = 5401, loss = 0.577394, precision = 0.867188 (59.489 sec) +Saved checkpoint after 14 epoch(s) to data/resnet20/checkpoints/00014... +INFO:tensorflow:global_step/sec: 1.63475 +INFO:tensorflow:step = 5501, loss = 0.619628, precision = 0.875 (61.172 sec) +INFO:tensorflow:global_step/sec: 1.6507 +INFO:tensorflow:step = 5601, loss = 0.816636, precision = 0.804688 (60.580 sec) +INFO:tensorflow:global_step/sec: 1.58641 +INFO:tensorflow:step = 5701, loss = 0.843703, precision = 0.796875 (63.036 sec) +INFO:tensorflow:global_step/sec: 1.54643 +INFO:tensorflow:step = 5801, loss = 0.777115, precision = 0.820312 (64.665 sec) +Saved checkpoint after 15 epoch(s) to data/resnet20/checkpoints/00015... +INFO:tensorflow:global_step/sec: 1.55674 +INFO:tensorflow:step = 5901, loss = 0.830447, precision = 0.820312 (64.237 sec) +INFO:tensorflow:global_step/sec: 1.52388 +INFO:tensorflow:step = 6001, loss = 0.809877, precision = 0.796875 (65.622 sec) +INFO:tensorflow:global_step/sec: 1.53082 +INFO:tensorflow:step = 6101, loss = 0.770401, precision = 0.820312 (65.325 sec) +INFO:tensorflow:global_step/sec: 1.53816 +INFO:tensorflow:step = 6201, loss = 0.893268, precision = 0.78125 (65.013 sec) +Saved checkpoint after 16 epoch(s) to data/resnet20/checkpoints/00016... +INFO:tensorflow:global_step/sec: 1.52312 +INFO:tensorflow:step = 6301, loss = 0.65625, precision = 0.882812 (65.655 sec) +INFO:tensorflow:global_step/sec: 1.50358 +INFO:tensorflow:step = 6401, loss = 0.747932, precision = 0.828125 (66.508 sec) +INFO:tensorflow:global_step/sec: 1.58133 +INFO:tensorflow:step = 6501, loss = 0.705926, precision = 0.835938 (63.238 sec) +INFO:tensorflow:global_step/sec: 1.5678 +INFO:tensorflow:step = 6601, loss = 0.67982, precision = 0.859375 (63.783 sec) +Saved checkpoint after 17 epoch(s) to data/resnet20/checkpoints/00017... +INFO:tensorflow:global_step/sec: 1.53863 +INFO:tensorflow:step = 6701, loss = 0.647462, precision = 0.835938 (64.993 sec) +INFO:tensorflow:global_step/sec: 1.5959 +INFO:tensorflow:step = 6801, loss = 0.682353, precision = 0.828125 (62.660 sec) +INFO:tensorflow:global_step/sec: 1.54736 +INFO:tensorflow:step = 6901, loss = 0.74936, precision = 0.828125 (64.626 sec) +INFO:tensorflow:global_step/sec: 1.58364 +INFO:tensorflow:step = 7001, loss = 0.847003, precision = 0.742188 (63.146 sec) +Saved checkpoint after 18 epoch(s) to data/resnet20/checkpoints/00018... +INFO:tensorflow:global_step/sec: 1.56799 +INFO:tensorflow:step = 7101, loss = 0.591658, precision = 0.875 (63.776 sec) +INFO:tensorflow:global_step/sec: 1.5901 +INFO:tensorflow:step = 7201, loss = 0.59565, precision = 0.90625 (62.889 sec) +INFO:tensorflow:global_step/sec: 1.5802 +INFO:tensorflow:step = 7301, loss = 0.585775, precision = 0.859375 (63.283 sec) +INFO:tensorflow:global_step/sec: 1.55779 +INFO:tensorflow:step = 7401, loss = 0.627907, precision = 0.84375 (64.193 sec) +Saved checkpoint after 19 epoch(s) to data/resnet20/checkpoints/00019... +INFO:tensorflow:global_step/sec: 1.56328 +INFO:tensorflow:step = 7501, loss = 0.766165, precision = 0.820312 (63.968 sec) +INFO:tensorflow:global_step/sec: 1.58512 +INFO:tensorflow:step = 7601, loss = 0.773539, precision = 0.8125 (63.087 sec) +INFO:tensorflow:global_step/sec: 1.52955 +INFO:tensorflow:step = 7701, loss = 0.586248, precision = 0.898438 (65.379 sec) +INFO:tensorflow:global_step/sec: 1.56949 +INFO:tensorflow:step = 7801, loss = 0.525517, precision = 0.914062 (63.715 sec) +Saved checkpoint after 20 epoch(s) to data/resnet20/checkpoints/00020... +INFO:tensorflow:global_step/sec: 1.58504 +INFO:tensorflow:step = 7901, loss = 0.532694, precision = 0.921875 (63.090 sec) +INFO:tensorflow:global_step/sec: 1.58535 +INFO:tensorflow:step = 8001, loss = 0.812557, precision = 0.796875 (63.077 sec) +INFO:tensorflow:global_step/sec: 1.60212 +INFO:tensorflow:step = 8101, loss = 0.824661, precision = 0.8125 (62.417 sec) +INFO:tensorflow:global_step/sec: 1.55228 +INFO:tensorflow:step = 8201, loss = 0.593777, precision = 0.914062 (64.421 sec) +Saved checkpoint after 21 epoch(s) to data/resnet20/checkpoints/00021... +INFO:tensorflow:global_step/sec: 1.54962 +INFO:tensorflow:step = 8301, loss = 0.672527, precision = 0.84375 (64.532 sec) +INFO:tensorflow:global_step/sec: 1.54866 +INFO:tensorflow:step = 8401, loss = 0.701752, precision = 0.851562 (64.572 sec) +INFO:tensorflow:global_step/sec: 1.58019 +INFO:tensorflow:step = 8501, loss = 0.595407, precision = 0.875 (63.284 sec) +INFO:tensorflow:global_step/sec: 1.55869 +INFO:tensorflow:step = 8601, loss = 0.679889, precision = 0.859375 (64.156 sec) +Saved checkpoint after 22 epoch(s) to data/resnet20/checkpoints/00022... +INFO:tensorflow:global_step/sec: 1.56342 +INFO:tensorflow:step = 8701, loss = 0.687805, precision = 0.820312 (63.962 sec) +INFO:tensorflow:global_step/sec: 1.58366 +INFO:tensorflow:step = 8801, loss = 0.653187, precision = 0.851562 (63.145 sec) +INFO:tensorflow:global_step/sec: 1.60529 +INFO:tensorflow:step = 8901, loss = 0.792771, precision = 0.8125 (62.294 sec) +Saved checkpoint after 23 epoch(s) to data/resnet20/checkpoints/00023... +INFO:tensorflow:global_step/sec: 1.58898 +INFO:tensorflow:step = 9001, loss = 0.663949, precision = 0.835938 (62.933 sec) +INFO:tensorflow:global_step/sec: 1.53495 +INFO:tensorflow:step = 9101, loss = 0.744811, precision = 0.835938 (65.149 sec) +INFO:tensorflow:global_step/sec: 1.54189 +INFO:tensorflow:step = 9201, loss = 0.591931, precision = 0.890625 (64.855 sec) +INFO:tensorflow:global_step/sec: 1.57369 +INFO:tensorflow:step = 9301, loss = 0.75786, precision = 0.8125 (63.545 sec) +Saved checkpoint after 24 epoch(s) to data/resnet20/checkpoints/00024... +INFO:tensorflow:global_step/sec: 1.57337 +INFO:tensorflow:step = 9401, loss = 0.818034, precision = 0.820312 (63.558 sec) +INFO:tensorflow:global_step/sec: 1.5877 +INFO:tensorflow:step = 9501, loss = 0.618598, precision = 0.882812 (62.984 sec) +INFO:tensorflow:global_step/sec: 1.56895 +INFO:tensorflow:step = 9601, loss = 0.723511, precision = 0.84375 (63.737 sec) +INFO:tensorflow:global_step/sec: 1.57899 +INFO:tensorflow:step = 9701, loss = 0.622386, precision = 0.882812 (63.332 sec) +Saved checkpoint after 25 epoch(s) to data/resnet20/checkpoints/00025... +INFO:tensorflow:global_step/sec: 1.55029 +INFO:tensorflow:step = 9801, loss = 0.622629, precision = 0.890625 (64.504 sec) +INFO:tensorflow:global_step/sec: 1.56089 +INFO:tensorflow:step = 9901, loss = 0.647721, precision = 0.890625 (64.066 sec) +INFO:tensorflow:global_step/sec: 1.59496 +INFO:tensorflow:step = 10001, loss = 0.736271, precision = 0.804688 (62.698 sec) +INFO:tensorflow:global_step/sec: 1.59261 +INFO:tensorflow:step = 10101, loss = 0.72703, precision = 0.84375 (62.790 sec) +Saved checkpoint after 26 epoch(s) to data/resnet20/checkpoints/00026... +INFO:tensorflow:global_step/sec: 1.55251 +INFO:tensorflow:step = 10201, loss = 0.631039, precision = 0.882812 (64.412 sec) +INFO:tensorflow:global_step/sec: 1.55711 +INFO:tensorflow:step = 10301, loss = 0.583181, precision = 0.875 (64.221 sec) +INFO:tensorflow:global_step/sec: 1.5834 +INFO:tensorflow:step = 10401, loss = 0.62236, precision = 0.875 (63.155 sec) +INFO:tensorflow:global_step/sec: 1.59641 +INFO:tensorflow:step = 10501, loss = 0.655183, precision = 0.875 (62.640 sec) +Saved checkpoint after 27 epoch(s) to data/resnet20/checkpoints/00027... +INFO:tensorflow:global_step/sec: 1.5639 +INFO:tensorflow:step = 10601, loss = 0.688655, precision = 0.867188 (63.943 sec) +INFO:tensorflow:global_step/sec: 1.58347 +INFO:tensorflow:step = 10701, loss = 0.732867, precision = 0.8125 (63.152 sec) +INFO:tensorflow:global_step/sec: 1.57655 +INFO:tensorflow:step = 10801, loss = 0.642431, precision = 0.875 (63.430 sec) +INFO:tensorflow:global_step/sec: 1.58249 +INFO:tensorflow:step = 10901, loss = 0.661798, precision = 0.898438 (63.192 sec) +Saved checkpoint after 28 epoch(s) to data/resnet20/checkpoints/00028... +INFO:tensorflow:global_step/sec: 1.50361 +INFO:tensorflow:step = 11001, loss = 0.679659, precision = 0.835938 (66.507 sec) +INFO:tensorflow:global_step/sec: 1.52511 +INFO:tensorflow:step = 11101, loss = 0.736789, precision = 0.84375 (65.569 sec) +INFO:tensorflow:global_step/sec: 1.41277 +INFO:tensorflow:step = 11201, loss = 0.654128, precision = 0.851562 (70.782 sec) +INFO:tensorflow:global_step/sec: 1.50534 +INFO:tensorflow:step = 11301, loss = 0.627872, precision = 0.882812 (66.430 sec) +Saved checkpoint after 29 epoch(s) to data/resnet20/checkpoints/00029... +INFO:tensorflow:global_step/sec: 1.50699 +INFO:tensorflow:step = 11401, loss = 0.552044, precision = 0.90625 (66.358 sec) +INFO:tensorflow:global_step/sec: 1.53712 +INFO:tensorflow:step = 11501, loss = 0.766983, precision = 0.84375 (65.056 sec) +INFO:tensorflow:global_step/sec: 1.52369 +INFO:tensorflow:step = 11601, loss = 0.735915, precision = 0.804688 (65.630 sec) +INFO:tensorflow:global_step/sec: 1.51396 +INFO:tensorflow:step = 11701, loss = 0.694381, precision = 0.867188 (66.052 sec) +Saved checkpoint after 30 epoch(s) to data/resnet20/checkpoints/00030... +INFO:tensorflow:global_step/sec: 1.51075 +INFO:tensorflow:step = 11801, loss = 0.579636, precision = 0.875 (66.192 sec) +INFO:tensorflow:global_step/sec: 1.4842 +INFO:tensorflow:step = 11901, loss = 0.723832, precision = 0.851562 (67.376 sec) +INFO:tensorflow:global_step/sec: 1.52893 +INFO:tensorflow:step = 12001, loss = 0.684872, precision = 0.851562 (65.405 sec) +INFO:tensorflow:global_step/sec: 1.43934 +INFO:tensorflow:step = 12101, loss = 0.678532, precision = 0.867188 (69.476 sec) +Saved checkpoint after 31 epoch(s) to data/resnet20/checkpoints/00031... +INFO:tensorflow:global_step/sec: 1.50534 +INFO:tensorflow:step = 12201, loss = 0.630737, precision = 0.875 (66.430 sec) +INFO:tensorflow:global_step/sec: 1.50267 +INFO:tensorflow:step = 12301, loss = 0.578136, precision = 0.898438 (66.548 sec) +INFO:tensorflow:global_step/sec: 1.5132 +INFO:tensorflow:step = 12401, loss = 0.716782, precision = 0.875 (66.085 sec) +INFO:tensorflow:global_step/sec: 1.45593 +INFO:tensorflow:step = 12501, loss = 0.566461, precision = 0.929688 (68.685 sec) +Saved checkpoint after 32 epoch(s) to data/resnet20/checkpoints/00032... +INFO:tensorflow:global_step/sec: 1.4938 +INFO:tensorflow:step = 12601, loss = 0.664243, precision = 0.867188 (66.943 sec) +INFO:tensorflow:global_step/sec: 1.61636 +INFO:tensorflow:step = 12701, loss = 0.712351, precision = 0.835938 (61.867 sec) +INFO:tensorflow:global_step/sec: 1.62207 +INFO:tensorflow:step = 12801, loss = 0.537343, precision = 0.921875 (61.650 sec) +INFO:tensorflow:global_step/sec: 1.60055 +INFO:tensorflow:step = 12901, loss = 0.570339, precision = 0.890625 (62.479 sec) +Saved checkpoint after 33 epoch(s) to data/resnet20/checkpoints/00033... +INFO:tensorflow:global_step/sec: 1.59266 +INFO:tensorflow:step = 13001, loss = 0.737637, precision = 0.867188 (62.788 sec) +INFO:tensorflow:global_step/sec: 1.5964 +INFO:tensorflow:step = 13101, loss = 0.703848, precision = 0.828125 (62.641 sec) +INFO:tensorflow:global_step/sec: 1.58342 +INFO:tensorflow:step = 13201, loss = 0.689546, precision = 0.882812 (63.154 sec) +Saved checkpoint after 34 epoch(s) to data/resnet20/checkpoints/00034... +INFO:tensorflow:global_step/sec: 1.49971 +INFO:tensorflow:step = 13301, loss = 0.709373, precision = 0.828125 (66.680 sec) +INFO:tensorflow:global_step/sec: 1.52548 +INFO:tensorflow:step = 13401, loss = 0.547312, precision = 0.890625 (65.553 sec) +INFO:tensorflow:global_step/sec: 1.51845 +INFO:tensorflow:step = 13501, loss = 0.688304, precision = 0.875 (65.857 sec) +INFO:tensorflow:global_step/sec: 1.54422 +INFO:tensorflow:step = 13601, loss = 0.690342, precision = 0.84375 (64.758 sec) +Saved checkpoint after 35 epoch(s) to data/resnet20/checkpoints/00035... +INFO:tensorflow:global_step/sec: 1.45668 +INFO:tensorflow:step = 13701, loss = 0.607068, precision = 0.875 (68.649 sec) +INFO:tensorflow:global_step/sec: 1.52119 +INFO:tensorflow:step = 13801, loss = 0.652032, precision = 0.875 (65.738 sec) +INFO:tensorflow:global_step/sec: 1.45862 +INFO:tensorflow:step = 13901, loss = 0.740276, precision = 0.820312 (68.558 sec) +INFO:tensorflow:global_step/sec: 1.59441 +INFO:tensorflow:step = 14001, loss = 0.71443, precision = 0.820312 (62.719 sec) +Saved checkpoint after 36 epoch(s) to data/resnet20/checkpoints/00036... +INFO:tensorflow:global_step/sec: 1.50058 +INFO:tensorflow:step = 14101, loss = 0.64581, precision = 0.851562 (66.641 sec) +INFO:tensorflow:global_step/sec: 1.50478 +INFO:tensorflow:step = 14201, loss = 0.639704, precision = 0.890625 (66.455 sec) +INFO:tensorflow:global_step/sec: 1.47211 +INFO:tensorflow:step = 14301, loss = 0.731155, precision = 0.835938 (67.930 sec) +INFO:tensorflow:global_step/sec: 1.46921 +INFO:tensorflow:step = 14401, loss = 0.649842, precision = 0.875 (68.064 sec) +Saved checkpoint after 37 epoch(s) to data/resnet20/checkpoints/00037... +INFO:tensorflow:global_step/sec: 1.46584 +INFO:tensorflow:step = 14501, loss = 0.613579, precision = 0.898438 (68.221 sec) +INFO:tensorflow:global_step/sec: 1.43577 +INFO:tensorflow:step = 14601, loss = 0.663998, precision = 0.890625 (69.649 sec) +INFO:tensorflow:global_step/sec: 1.44465 +INFO:tensorflow:step = 14701, loss = 0.620629, precision = 0.84375 (69.221 sec) +INFO:tensorflow:global_step/sec: 1.44185 +INFO:tensorflow:step = 14801, loss = 0.771795, precision = 0.84375 (69.355 sec) +Saved checkpoint after 38 epoch(s) to data/resnet20/checkpoints/00038... +INFO:tensorflow:global_step/sec: 1.52669 +INFO:tensorflow:step = 14901, loss = 0.584798, precision = 0.914062 (65.502 sec) +INFO:tensorflow:global_step/sec: 1.54904 +INFO:tensorflow:step = 15001, loss = 0.837662, precision = 0.789062 (64.556 sec) +INFO:tensorflow:global_step/sec: 1.53878 +INFO:tensorflow:step = 15101, loss = 0.657185, precision = 0.835938 (64.987 sec) +INFO:tensorflow:global_step/sec: 1.55091 +INFO:tensorflow:step = 15201, loss = 0.706353, precision = 0.84375 (64.478 sec) +Saved checkpoint after 39 epoch(s) to data/resnet20/checkpoints/00039... +INFO:tensorflow:global_step/sec: 1.51457 +INFO:tensorflow:step = 15301, loss = 0.752717, precision = 0.851562 (66.025 sec) +INFO:tensorflow:global_step/sec: 1.51273 +INFO:tensorflow:step = 15401, loss = 0.665911, precision = 0.867188 (66.106 sec) +INFO:tensorflow:global_step/sec: 1.54771 +INFO:tensorflow:step = 15501, loss = 0.679165, precision = 0.859375 (64.612 sec) +INFO:tensorflow:global_step/sec: 1.54058 +INFO:tensorflow:step = 15601, loss = 0.672362, precision = 0.859375 (64.910 sec) +Saved checkpoint after 40 epoch(s) to data/resnet20/checkpoints/00040... +INFO:tensorflow:global_step/sec: 1.53658 +INFO:tensorflow:step = 15701, loss = 0.756446, precision = 0.84375 (65.080 sec) +INFO:tensorflow:global_step/sec: 1.54715 +INFO:tensorflow:step = 15801, loss = 0.650268, precision = 0.882812 (64.635 sec) +INFO:tensorflow:global_step/sec: 1.62533 +INFO:tensorflow:step = 15901, loss = 0.899167, precision = 0.789062 (61.526 sec) +INFO:tensorflow:global_step/sec: 1.64138 +INFO:tensorflow:step = 16001, loss = 0.70365, precision = 0.84375 (60.924 sec) +Saved checkpoint after 41 epoch(s) to data/resnet20/checkpoints/00041... +INFO:tensorflow:global_step/sec: 1.59482 +INFO:tensorflow:step = 16101, loss = 0.562205, precision = 0.890625 (62.703 sec) +INFO:tensorflow:global_step/sec: 1.61627 +INFO:tensorflow:step = 16201, loss = 0.640818, precision = 0.859375 (61.871 sec) +INFO:tensorflow:global_step/sec: 1.62449 +INFO:tensorflow:step = 16301, loss = 0.621792, precision = 0.882812 (61.558 sec) +INFO:tensorflow:global_step/sec: 1.6402 +INFO:tensorflow:step = 16401, loss = 0.71432, precision = 0.859375 (60.968 sec) +Saved checkpoint after 42 epoch(s) to data/resnet20/checkpoints/00042... +INFO:tensorflow:global_step/sec: 1.55519 +INFO:tensorflow:step = 16501, loss = 0.614206, precision = 0.851562 (64.301 sec) +INFO:tensorflow:global_step/sec: 1.50461 +INFO:tensorflow:step = 16601, loss = 0.720423, precision = 0.820312 (66.462 sec) +INFO:tensorflow:global_step/sec: 1.47085 +INFO:tensorflow:step = 16701, loss = 0.668331, precision = 0.835938 (67.988 sec) +INFO:tensorflow:global_step/sec: 1.52228 +INFO:tensorflow:step = 16801, loss = 0.660397, precision = 0.867188 (65.691 sec) +Saved checkpoint after 43 epoch(s) to data/resnet20/checkpoints/00043... +INFO:tensorflow:global_step/sec: 1.45342 +INFO:tensorflow:step = 16901, loss = 0.725341, precision = 0.882812 (68.804 sec) +INFO:tensorflow:global_step/sec: 1.48257 +INFO:tensorflow:step = 17001, loss = 0.549321, precision = 0.90625 (67.450 sec) +INFO:tensorflow:global_step/sec: 1.473 +INFO:tensorflow:step = 17101, loss = 0.6658, precision = 0.867188 (67.889 sec) +INFO:tensorflow:global_step/sec: 1.46301 +INFO:tensorflow:step = 17201, loss = 0.69629, precision = 0.851562 (68.352 sec) +Saved checkpoint after 44 epoch(s) to data/resnet20/checkpoints/00044... +INFO:tensorflow:global_step/sec: 1.47147 +INFO:tensorflow:step = 17301, loss = 0.68453, precision = 0.859375 (67.959 sec) +INFO:tensorflow:global_step/sec: 1.50853 +INFO:tensorflow:step = 17401, loss = 0.714991, precision = 0.859375 (66.289 sec) +INFO:tensorflow:global_step/sec: 1.5401 +INFO:tensorflow:step = 17501, loss = 0.653453, precision = 0.859375 (64.931 sec) +Saved checkpoint after 45 epoch(s) to data/resnet20/checkpoints/00045... +INFO:tensorflow:global_step/sec: 1.40336 +INFO:tensorflow:step = 17601, loss = 0.733366, precision = 0.851562 (71.258 sec) +INFO:tensorflow:global_step/sec: 1.4439 +INFO:tensorflow:step = 17701, loss = 0.708236, precision = 0.859375 (69.257 sec) +INFO:tensorflow:global_step/sec: 1.43756 +INFO:tensorflow:step = 17801, loss = 0.855251, precision = 0.804688 (69.562 sec) +INFO:tensorflow:global_step/sec: 1.44408 +INFO:tensorflow:step = 17901, loss = 0.679052, precision = 0.875 (69.248 sec) +Saved checkpoint after 46 epoch(s) to data/resnet20/checkpoints/00046... +INFO:tensorflow:global_step/sec: 1.45101 +INFO:tensorflow:step = 18001, loss = 0.615706, precision = 0.867188 (68.918 sec) +INFO:tensorflow:global_step/sec: 1.45408 +INFO:tensorflow:step = 18101, loss = 0.777204, precision = 0.851562 (68.772 sec) +INFO:tensorflow:global_step/sec: 1.51833 +INFO:tensorflow:step = 18201, loss = 0.778937, precision = 0.851562 (65.861 sec) +INFO:tensorflow:global_step/sec: 1.52288 +INFO:tensorflow:step = 18301, loss = 0.771181, precision = 0.835938 (65.665 sec) +Saved checkpoint after 47 epoch(s) to data/resnet20/checkpoints/00047... +INFO:tensorflow:global_step/sec: 1.55789 +INFO:tensorflow:step = 18401, loss = 0.802125, precision = 0.828125 (64.189 sec) +INFO:tensorflow:global_step/sec: 1.54619 +INFO:tensorflow:step = 18501, loss = 0.637888, precision = 0.875 (64.675 sec) +INFO:tensorflow:global_step/sec: 1.47082 +INFO:tensorflow:step = 18601, loss = 0.66636, precision = 0.882812 (67.989 sec) +INFO:tensorflow:global_step/sec: 1.48268 +INFO:tensorflow:step = 18701, loss = 0.741107, precision = 0.820312 (67.445 sec) +Saved checkpoint after 48 epoch(s) to data/resnet20/checkpoints/00048... +INFO:tensorflow:global_step/sec: 1.51888 +INFO:tensorflow:step = 18801, loss = 0.53541, precision = 0.890625 (65.838 sec) +INFO:tensorflow:global_step/sec: 1.54082 +INFO:tensorflow:step = 18901, loss = 0.695562, precision = 0.835938 (64.900 sec) +INFO:tensorflow:global_step/sec: 1.55845 +INFO:tensorflow:step = 19001, loss = 0.752361, precision = 0.84375 (64.166 sec) +INFO:tensorflow:global_step/sec: 1.5769 +INFO:tensorflow:step = 19101, loss = 0.612695, precision = 0.875 (63.416 sec) +Saved checkpoint after 49 epoch(s) to data/resnet20/checkpoints/00049... +INFO:tensorflow:global_step/sec: 1.57448 +INFO:tensorflow:step = 19201, loss = 0.554885, precision = 0.921875 (63.513 sec) +INFO:tensorflow:global_step/sec: 1.5712 +INFO:tensorflow:step = 19301, loss = 0.609801, precision = 0.882812 (63.646 sec) +INFO:tensorflow:global_step/sec: 1.56276 +INFO:tensorflow:step = 19401, loss = 0.780707, precision = 0.8125 (63.989 sec) +INFO:tensorflow:global_step/sec: 1.56749 +INFO:tensorflow:step = 19501, loss = 0.719959, precision = 0.859375 (63.796 sec) +Saved checkpoint after 50 epoch(s) to data/resnet20/checkpoints/00050... +INFO:tensorflow:global_step/sec: 1.4156 +INFO:tensorflow:step = 19601, loss = 0.700101, precision = 0.84375 (70.641 sec) +INFO:tensorflow:global_step/sec: 1.45525 +INFO:tensorflow:step = 19701, loss = 0.767837, precision = 0.820312 (68.717 sec) +INFO:tensorflow:global_step/sec: 1.27132 +INFO:tensorflow:step = 19801, loss = 0.55373, precision = 0.882812 (78.658 sec) +INFO:tensorflow:global_step/sec: 1.29081 +INFO:tensorflow:step = 19901, loss = 0.78913, precision = 0.84375 (77.470 sec) +Saved checkpoint after 51 epoch(s) to data/resnet20/checkpoints/00051... +INFO:tensorflow:global_step/sec: 1.42817 +INFO:tensorflow:step = 20001, loss = 0.657035, precision = 0.882812 (70.020 sec) +INFO:tensorflow:global_step/sec: 1.34744 +INFO:tensorflow:step = 20101, loss = 0.786068, precision = 0.84375 (74.215 sec) +INFO:tensorflow:global_step/sec: 1.38585 +INFO:tensorflow:step = 20201, loss = 0.794631, precision = 0.8125 (72.158 sec) +INFO:tensorflow:global_step/sec: 1.50771 +INFO:tensorflow:step = 20301, loss = 0.662728, precision = 0.882812 (66.325 sec) +Saved checkpoint after 52 epoch(s) to data/resnet20/checkpoints/00052... +INFO:tensorflow:global_step/sec: 1.4728 +INFO:tensorflow:step = 20401, loss = 0.5844, precision = 0.890625 (67.898 sec) +INFO:tensorflow:global_step/sec: 1.50318 +INFO:tensorflow:step = 20501, loss = 0.6027, precision = 0.90625 (66.526 sec) +INFO:tensorflow:global_step/sec: 1.53175 +INFO:tensorflow:step = 20601, loss = 0.662363, precision = 0.875 (65.285 sec) +INFO:tensorflow:global_step/sec: 1.43749 +INFO:tensorflow:step = 20701, loss = 0.72499, precision = 0.851562 (69.566 sec) +Saved checkpoint after 53 epoch(s) to data/resnet20/checkpoints/00053... +INFO:tensorflow:global_step/sec: 1.52103 +INFO:tensorflow:step = 20801, loss = 0.514132, precision = 0.929688 (65.745 sec) +INFO:tensorflow:global_step/sec: 1.50763 +INFO:tensorflow:step = 20901, loss = 0.713647, precision = 0.835938 (66.329 sec) +INFO:tensorflow:global_step/sec: 1.56795 +INFO:tensorflow:step = 21001, loss = 0.692818, precision = 0.867188 (63.777 sec) +INFO:tensorflow:global_step/sec: 1.58092 +INFO:tensorflow:step = 21101, loss = 0.679634, precision = 0.851562 (63.255 sec) +Saved checkpoint after 54 epoch(s) to data/resnet20/checkpoints/00054... +INFO:tensorflow:global_step/sec: 1.57138 +INFO:tensorflow:step = 21201, loss = 0.708345, precision = 0.875 (63.639 sec) +INFO:tensorflow:global_step/sec: 1.64764 +INFO:tensorflow:step = 21301, loss = 0.76733, precision = 0.835938 (60.693 sec) +INFO:tensorflow:global_step/sec: 1.61011 +INFO:tensorflow:step = 21401, loss = 0.611953, precision = 0.867188 (62.108 sec) +INFO:tensorflow:global_step/sec: 1.62378 +INFO:tensorflow:step = 21501, loss = 0.602955, precision = 0.898438 (61.584 sec) +Saved checkpoint after 55 epoch(s) to data/resnet20/checkpoints/00055... +INFO:tensorflow:global_step/sec: 1.62243 +INFO:tensorflow:step = 21601, loss = 0.61772, precision = 0.859375 (61.636 sec) +INFO:tensorflow:global_step/sec: 1.64097 +INFO:tensorflow:step = 21701, loss = 0.632458, precision = 0.867188 (60.940 sec) +INFO:tensorflow:global_step/sec: 1.64133 +INFO:tensorflow:step = 21801, loss = 0.633773, precision = 0.898438 (60.926 sec) +Saved checkpoint after 56 epoch(s) to data/resnet20/checkpoints/00056... +INFO:tensorflow:global_step/sec: 1.61912 +INFO:tensorflow:step = 21901, loss = 0.659664, precision = 0.835938 (61.762 sec) +INFO:tensorflow:global_step/sec: 1.68661 +INFO:tensorflow:step = 22001, loss = 0.71117, precision = 0.789062 (59.291 sec) +INFO:tensorflow:global_step/sec: 1.67944 +INFO:tensorflow:step = 22101, loss = 0.68984, precision = 0.835938 (59.544 sec) +INFO:tensorflow:global_step/sec: 1.68864 +INFO:tensorflow:step = 22201, loss = 0.609767, precision = 0.882812 (59.219 sec) +Saved checkpoint after 57 epoch(s) to data/resnet20/checkpoints/00057... +INFO:tensorflow:global_step/sec: 1.71551 +INFO:tensorflow:step = 22301, loss = 0.638482, precision = 0.882812 (58.292 sec) +INFO:tensorflow:global_step/sec: 1.62883 +INFO:tensorflow:step = 22401, loss = 0.682938, precision = 0.867188 (61.394 sec) +INFO:tensorflow:global_step/sec: 1.60355 +INFO:tensorflow:step = 22501, loss = 0.723486, precision = 0.851562 (62.362 sec) +INFO:tensorflow:global_step/sec: 1.61357 +INFO:tensorflow:step = 22601, loss = 0.685334, precision = 0.875 (61.974 sec) +Saved checkpoint after 58 epoch(s) to data/resnet20/checkpoints/00058... +INFO:tensorflow:global_step/sec: 1.5514 +INFO:tensorflow:step = 22701, loss = 0.538548, precision = 0.921875 (64.458 sec) +INFO:tensorflow:global_step/sec: 1.53308 +INFO:tensorflow:step = 22801, loss = 0.635249, precision = 0.851562 (65.228 sec) +INFO:tensorflow:global_step/sec: 1.55838 +INFO:tensorflow:step = 22901, loss = 0.718727, precision = 0.835938 (64.169 sec) +INFO:tensorflow:global_step/sec: 1.6213 +INFO:tensorflow:step = 23001, loss = 0.624964, precision = 0.890625 (61.679 sec) +Saved checkpoint after 59 epoch(s) to data/resnet20/checkpoints/00059... +INFO:tensorflow:global_step/sec: 1.70579 +INFO:tensorflow:step = 23101, loss = 0.710213, precision = 0.835938 (58.624 sec) +INFO:tensorflow:global_step/sec: 1.72982 +INFO:tensorflow:step = 23201, loss = 0.771737, precision = 0.859375 (57.809 sec) +INFO:tensorflow:global_step/sec: 1.74104 +INFO:tensorflow:step = 23301, loss = 0.723209, precision = 0.835938 (57.437 sec) +INFO:tensorflow:global_step/sec: 1.72483 +INFO:tensorflow:step = 23401, loss = 0.802945, precision = 0.820312 (57.977 sec) +Saved checkpoint after 60 epoch(s) to data/resnet20/checkpoints/00060... +INFO:tensorflow:global_step/sec: 1.67731 +INFO:tensorflow:step = 23501, loss = 0.678447, precision = 0.867188 (59.619 sec) +INFO:tensorflow:global_step/sec: 1.54424 +INFO:tensorflow:step = 23601, loss = 0.722871, precision = 0.835938 (64.757 sec) +INFO:tensorflow:global_step/sec: 1.45725 +INFO:tensorflow:step = 23701, loss = 0.670434, precision = 0.859375 (68.622 sec) +INFO:tensorflow:global_step/sec: 1.59237 +INFO:tensorflow:step = 23801, loss = 0.696019, precision = 0.84375 (62.799 sec) +Saved checkpoint after 61 epoch(s) to data/resnet20/checkpoints/00061... +INFO:tensorflow:global_step/sec: 1.67329 +INFO:tensorflow:step = 23901, loss = 0.6106, precision = 0.867188 (59.762 sec) +INFO:tensorflow:global_step/sec: 1.62575 +INFO:tensorflow:step = 24001, loss = 0.616202, precision = 0.890625 (61.510 sec) +INFO:tensorflow:global_step/sec: 1.64773 +INFO:tensorflow:step = 24101, loss = 0.711669, precision = 0.851562 (60.689 sec) +INFO:tensorflow:global_step/sec: 1.62974 +INFO:tensorflow:step = 24201, loss = 0.607655, precision = 0.898438 (61.360 sec) +Saved checkpoint after 62 epoch(s) to data/resnet20/checkpoints/00062... +INFO:tensorflow:global_step/sec: 1.60523 +INFO:tensorflow:step = 24301, loss = 0.701967, precision = 0.851562 (62.296 sec) +INFO:tensorflow:global_step/sec: 1.6891 +INFO:tensorflow:step = 24401, loss = 0.628124, precision = 0.890625 (59.203 sec) +INFO:tensorflow:global_step/sec: 1.70972 +INFO:tensorflow:step = 24501, loss = 0.686062, precision = 0.851562 (58.489 sec) +INFO:tensorflow:global_step/sec: 1.74145 +INFO:tensorflow:step = 24601, loss = 0.72426, precision = 0.859375 (57.423 sec) +Saved checkpoint after 63 epoch(s) to data/resnet20/checkpoints/00063... +INFO:tensorflow:global_step/sec: 1.73058 +INFO:tensorflow:step = 24701, loss = 0.587711, precision = 0.890625 (57.784 sec) +INFO:tensorflow:global_step/sec: 1.74069 +INFO:tensorflow:step = 24801, loss = 0.777701, precision = 0.78125 (57.448 sec) +INFO:tensorflow:global_step/sec: 1.72759 +INFO:tensorflow:step = 24901, loss = 0.552325, precision = 0.921875 (57.884 sec) +INFO:tensorflow:global_step/sec: 1.68898 +INFO:tensorflow:step = 25001, loss = 0.674696, precision = 0.867188 (59.207 sec) +Saved checkpoint after 64 epoch(s) to data/resnet20/checkpoints/00064... +INFO:tensorflow:global_step/sec: 1.58272 +INFO:tensorflow:step = 25101, loss = 0.696647, precision = 0.84375 (63.182 sec) +INFO:tensorflow:global_step/sec: 1.67022 +INFO:tensorflow:step = 25201, loss = 0.586556, precision = 0.875 (59.872 sec) +INFO:tensorflow:global_step/sec: 1.72879 +INFO:tensorflow:step = 25301, loss = 0.574885, precision = 0.890625 (57.844 sec) +INFO:tensorflow:global_step/sec: 1.71215 +INFO:tensorflow:step = 25401, loss = 0.505408, precision = 0.929688 (58.406 sec) +Saved checkpoint after 65 epoch(s) to data/resnet20/checkpoints/00065... +INFO:tensorflow:global_step/sec: 1.67682 +INFO:tensorflow:step = 25501, loss = 0.742746, precision = 0.835938 (59.637 sec) +INFO:tensorflow:global_step/sec: 1.62581 +INFO:tensorflow:step = 25601, loss = 0.565486, precision = 0.898438 (61.508 sec) +INFO:tensorflow:global_step/sec: 1.42737 +INFO:tensorflow:step = 25701, loss = 0.750141, precision = 0.851562 (70.059 sec) +INFO:tensorflow:global_step/sec: 1.51489 +INFO:tensorflow:step = 25801, loss = 0.656717, precision = 0.882812 (66.011 sec) +Saved checkpoint after 66 epoch(s) to data/resnet20/checkpoints/00066... +INFO:tensorflow:global_step/sec: 1.55369 +INFO:tensorflow:step = 25901, loss = 0.675905, precision = 0.851562 (64.363 sec) +INFO:tensorflow:global_step/sec: 1.588 +INFO:tensorflow:step = 26001, loss = 0.548083, precision = 0.90625 (62.972 sec) +INFO:tensorflow:global_step/sec: 1.64524 +INFO:tensorflow:step = 26101, loss = 0.671158, precision = 0.867188 (60.781 sec) +Saved checkpoint after 67 epoch(s) to data/resnet20/checkpoints/00067... +INFO:tensorflow:global_step/sec: 1.62036 +INFO:tensorflow:step = 26201, loss = 0.712233, precision = 0.859375 (61.715 sec) +INFO:tensorflow:global_step/sec: 1.69736 +INFO:tensorflow:step = 26301, loss = 0.566034, precision = 0.898438 (58.915 sec) +INFO:tensorflow:global_step/sec: 1.62774 +INFO:tensorflow:step = 26401, loss = 0.676905, precision = 0.859375 (61.435 sec) +INFO:tensorflow:global_step/sec: 1.65409 +INFO:tensorflow:step = 26501, loss = 0.671349, precision = 0.851562 (60.456 sec) +Saved checkpoint after 68 epoch(s) to data/resnet20/checkpoints/00068... +INFO:tensorflow:global_step/sec: 1.62643 +INFO:tensorflow:step = 26601, loss = 0.635986, precision = 0.867188 (61.485 sec) +INFO:tensorflow:global_step/sec: 1.61648 +INFO:tensorflow:step = 26701, loss = 0.662242, precision = 0.890625 (61.863 sec) +INFO:tensorflow:global_step/sec: 1.64013 +INFO:tensorflow:step = 26801, loss = 0.585262, precision = 0.921875 (60.971 sec) +INFO:tensorflow:global_step/sec: 1.6857 +INFO:tensorflow:step = 26901, loss = 0.764096, precision = 0.835938 (59.323 sec) +Saved checkpoint after 69 epoch(s) to data/resnet20/checkpoints/00069... +INFO:tensorflow:global_step/sec: 1.54756 +INFO:tensorflow:step = 27001, loss = 0.627338, precision = 0.898438 (64.618 sec) +INFO:tensorflow:global_step/sec: 1.55526 +INFO:tensorflow:step = 27101, loss = 0.652772, precision = 0.890625 (64.298 sec) +INFO:tensorflow:global_step/sec: 1.55141 +INFO:tensorflow:step = 27201, loss = 0.552271, precision = 0.898438 (64.457 sec) +INFO:tensorflow:global_step/sec: 1.63811 +INFO:tensorflow:step = 27301, loss = 0.560491, precision = 0.890625 (61.047 sec) +Saved checkpoint after 70 epoch(s) to data/resnet20/checkpoints/00070... +INFO:tensorflow:global_step/sec: 1.68691 +INFO:tensorflow:step = 27401, loss = 0.653007, precision = 0.882812 (59.279 sec) +INFO:tensorflow:global_step/sec: 1.63676 +INFO:tensorflow:step = 27501, loss = 0.630428, precision = 0.882812 (61.096 sec) +INFO:tensorflow:global_step/sec: 1.7157 +INFO:tensorflow:step = 27601, loss = 0.764752, precision = 0.835938 (58.285 sec) +INFO:tensorflow:global_step/sec: 1.67999 +INFO:tensorflow:step = 27701, loss = 0.632816, precision = 0.898438 (59.524 sec) +Saved checkpoint after 71 epoch(s) to data/resnet20/checkpoints/00071... +INFO:tensorflow:global_step/sec: 1.71468 +INFO:tensorflow:step = 27801, loss = 0.527472, precision = 0.90625 (58.320 sec) +INFO:tensorflow:global_step/sec: 1.69286 +INFO:tensorflow:step = 27901, loss = 0.675535, precision = 0.882812 (59.072 sec) +INFO:tensorflow:global_step/sec: 1.69444 +INFO:tensorflow:step = 28001, loss = 0.659746, precision = 0.875 (59.016 sec) +INFO:tensorflow:global_step/sec: 1.68775 +INFO:tensorflow:step = 28101, loss = 0.729299, precision = 0.835938 (59.251 sec) +Saved checkpoint after 72 epoch(s) to data/resnet20/checkpoints/00072... +INFO:tensorflow:global_step/sec: 1.6663 +INFO:tensorflow:step = 28201, loss = 0.644823, precision = 0.867188 (60.013 sec) +INFO:tensorflow:global_step/sec: 1.66313 +INFO:tensorflow:step = 28301, loss = 0.596446, precision = 0.90625 (60.128 sec) +INFO:tensorflow:global_step/sec: 1.69736 +INFO:tensorflow:step = 28401, loss = 0.89508, precision = 0.820312 (58.915 sec) +INFO:tensorflow:global_step/sec: 1.69097 +INFO:tensorflow:step = 28501, loss = 0.491487, precision = 0.914062 (59.137 sec) +Saved checkpoint after 73 epoch(s) to data/resnet20/checkpoints/00073... +INFO:tensorflow:global_step/sec: 1.69175 +INFO:tensorflow:step = 28601, loss = 0.512221, precision = 0.90625 (59.111 sec) +INFO:tensorflow:global_step/sec: 1.59886 +INFO:tensorflow:step = 28701, loss = 0.610898, precision = 0.898438 (62.544 sec) +INFO:tensorflow:global_step/sec: 1.32404 +INFO:tensorflow:step = 28801, loss = 0.694949, precision = 0.828125 (75.527 sec) +INFO:tensorflow:global_step/sec: 1.39378 +INFO:tensorflow:step = 28901, loss = 0.770704, precision = 0.835938 (71.747 sec) +Saved checkpoint after 74 epoch(s) to data/resnet20/checkpoints/00074... +INFO:tensorflow:global_step/sec: 1.51681 +INFO:tensorflow:step = 29001, loss = 0.649584, precision = 0.867188 (65.928 sec) +INFO:tensorflow:global_step/sec: 1.55511 +INFO:tensorflow:step = 29101, loss = 0.638853, precision = 0.859375 (64.304 sec) +INFO:tensorflow:global_step/sec: 1.57966 +INFO:tensorflow:step = 29201, loss = 0.661834, precision = 0.859375 (63.305 sec) +INFO:tensorflow:global_step/sec: 1.66981 +INFO:tensorflow:step = 29301, loss = 0.563133, precision = 0.875 (59.887 sec) +Saved checkpoint after 75 epoch(s) to data/resnet20/checkpoints/00075... +INFO:tensorflow:global_step/sec: 1.60488 +INFO:tensorflow:step = 29401, loss = 0.494008, precision = 0.945312 (62.310 sec) +INFO:tensorflow:global_step/sec: 1.58213 +INFO:tensorflow:step = 29501, loss = 0.64821, precision = 0.835938 (63.206 sec) +INFO:tensorflow:global_step/sec: 1.61379 +INFO:tensorflow:step = 29601, loss = 0.582875, precision = 0.882812 (61.966 sec) +INFO:tensorflow:global_step/sec: 1.60333 +INFO:tensorflow:step = 29701, loss = 0.739944, precision = 0.851562 (62.370 sec) +Saved checkpoint after 76 epoch(s) to data/resnet20/checkpoints/00076... +INFO:tensorflow:global_step/sec: 1.59062 +INFO:tensorflow:step = 29801, loss = 0.678566, precision = 0.875 (62.869 sec) +INFO:tensorflow:global_step/sec: 1.55914 +INFO:tensorflow:step = 29901, loss = 0.654501, precision = 0.859375 (64.138 sec) +INFO:tensorflow:global_step/sec: 1.60958 +INFO:tensorflow:step = 30001, loss = 0.533671, precision = 0.914062 (62.128 sec) +INFO:tensorflow:global_step/sec: 1.54997 +INFO:tensorflow:step = 30101, loss = 0.739499, precision = 0.851562 (64.517 sec) +Saved checkpoint after 77 epoch(s) to data/resnet20/checkpoints/00077... +INFO:tensorflow:global_step/sec: 1.43413 +INFO:tensorflow:step = 30201, loss = 0.525179, precision = 0.921875 (69.729 sec) +INFO:tensorflow:global_step/sec: 1.62654 +INFO:tensorflow:step = 30301, loss = 0.597667, precision = 0.882812 (61.480 sec) +INFO:tensorflow:global_step/sec: 1.59012 +INFO:tensorflow:step = 30401, loss = 0.568053, precision = 0.882812 (62.889 sec) +Saved checkpoint after 78 epoch(s) to data/resnet20/checkpoints/00078... +INFO:tensorflow:global_step/sec: 1.64866 +INFO:tensorflow:step = 30501, loss = 0.610836, precision = 0.898438 (60.655 sec) +INFO:tensorflow:global_step/sec: 1.67831 +INFO:tensorflow:step = 30601, loss = 0.727957, precision = 0.828125 (59.584 sec) +INFO:tensorflow:global_step/sec: 1.65684 +INFO:tensorflow:step = 30701, loss = 0.680211, precision = 0.867188 (60.356 sec) +INFO:tensorflow:global_step/sec: 1.67102 +INFO:tensorflow:step = 30801, loss = 0.531496, precision = 0.890625 (59.844 sec) +Saved checkpoint after 79 epoch(s) to data/resnet20/checkpoints/00079... +INFO:tensorflow:global_step/sec: 1.60805 +INFO:tensorflow:step = 30901, loss = 0.696252, precision = 0.875 (62.187 sec) +INFO:tensorflow:global_step/sec: 1.54121 +INFO:tensorflow:step = 31001, loss = 0.649341, precision = 0.859375 (64.884 sec) +INFO:tensorflow:global_step/sec: 1.52912 +INFO:tensorflow:step = 31101, loss = 0.560077, precision = 0.898438 (65.397 sec) +INFO:tensorflow:global_step/sec: 1.48676 +INFO:tensorflow:step = 31201, loss = 0.586556, precision = 0.898438 (67.260 sec) +Saved checkpoint after 80 epoch(s) to data/resnet20/checkpoints/00080... +INFO:tensorflow:global_step/sec: 1.39807 +INFO:tensorflow:step = 31301, loss = 0.633648, precision = 0.859375 (71.527 sec) +INFO:tensorflow:global_step/sec: 1.49604 +INFO:tensorflow:step = 31401, loss = 0.700259, precision = 0.890625 (66.843 sec) +INFO:tensorflow:global_step/sec: 1.49635 +INFO:tensorflow:step = 31501, loss = 0.636934, precision = 0.859375 (66.829 sec) +INFO:tensorflow:global_step/sec: 1.55209 +INFO:tensorflow:step = 31601, loss = 0.628332, precision = 0.875 (64.429 sec) +Saved checkpoint after 81 epoch(s) to data/resnet20/checkpoints/00081... +INFO:tensorflow:global_step/sec: 1.65375 +INFO:tensorflow:step = 31701, loss = 0.6736, precision = 0.875 (60.469 sec) +INFO:tensorflow:global_step/sec: 1.65141 +INFO:tensorflow:step = 31801, loss = 0.571543, precision = 0.890625 (60.554 sec) +INFO:tensorflow:global_step/sec: 1.53952 +INFO:tensorflow:step = 31901, loss = 0.655346, precision = 0.882812 (64.955 sec) +INFO:tensorflow:global_step/sec: 1.59066 +INFO:tensorflow:step = 32001, loss = 0.687948, precision = 0.851562 (62.867 sec) +Saved checkpoint after 82 epoch(s) to data/resnet20/checkpoints/00082... +INFO:tensorflow:global_step/sec: 1.62426 +INFO:tensorflow:step = 32101, loss = 0.713032, precision = 0.84375 (61.567 sec) +INFO:tensorflow:global_step/sec: 1.54262 +INFO:tensorflow:step = 32201, loss = 0.602374, precision = 0.90625 (64.825 sec) +INFO:tensorflow:global_step/sec: 1.49619 +INFO:tensorflow:step = 32301, loss = 0.602867, precision = 0.882812 (66.836 sec) +INFO:tensorflow:global_step/sec: 1.52509 +INFO:tensorflow:step = 32401, loss = 0.737543, precision = 0.835938 (65.570 sec) +Saved checkpoint after 83 epoch(s) to data/resnet20/checkpoints/00083... +INFO:tensorflow:global_step/sec: 1.52844 +INFO:tensorflow:step = 32501, loss = 0.574791, precision = 0.890625 (65.426 sec) +INFO:tensorflow:global_step/sec: 1.65268 +INFO:tensorflow:step = 32601, loss = 0.588808, precision = 0.890625 (60.508 sec) +INFO:tensorflow:global_step/sec: 1.53957 +INFO:tensorflow:step = 32701, loss = 0.585198, precision = 0.890625 (64.953 sec) +INFO:tensorflow:global_step/sec: 1.58748 +INFO:tensorflow:step = 32801, loss = 0.539497, precision = 0.9375 (62.993 sec) +Saved checkpoint after 84 epoch(s) to data/resnet20/checkpoints/00084... +INFO:tensorflow:global_step/sec: 1.58571 +INFO:tensorflow:step = 32901, loss = 0.673112, precision = 0.84375 (63.064 sec) +INFO:tensorflow:global_step/sec: 1.58644 +INFO:tensorflow:step = 33001, loss = 0.845926, precision = 0.820312 (63.034 sec) +INFO:tensorflow:global_step/sec: 1.56564 +INFO:tensorflow:step = 33101, loss = 0.594148, precision = 0.875 (63.872 sec) +INFO:tensorflow:global_step/sec: 1.62414 +INFO:tensorflow:step = 33201, loss = 0.604138, precision = 0.898438 (61.571 sec) +Saved checkpoint after 85 epoch(s) to data/resnet20/checkpoints/00085... +INFO:tensorflow:global_step/sec: 1.63811 +INFO:tensorflow:step = 33301, loss = 0.593228, precision = 0.875 (61.046 sec) +INFO:tensorflow:global_step/sec: 1.64269 +INFO:tensorflow:step = 33401, loss = 0.714943, precision = 0.835938 (60.876 sec) +INFO:tensorflow:global_step/sec: 1.6547 +INFO:tensorflow:step = 33501, loss = 0.662961, precision = 0.882812 (60.434 sec) +INFO:tensorflow:global_step/sec: 1.69141 +INFO:tensorflow:step = 33601, loss = 0.520347, precision = 0.9375 (59.122 sec) +Saved checkpoint after 86 epoch(s) to data/resnet20/checkpoints/00086... +INFO:tensorflow:global_step/sec: 1.6845 +INFO:tensorflow:step = 33701, loss = 0.709338, precision = 0.875 (59.365 sec) +INFO:tensorflow:global_step/sec: 1.70836 +INFO:tensorflow:step = 33801, loss = 0.661667, precision = 0.84375 (58.536 sec) +INFO:tensorflow:global_step/sec: 1.71452 +INFO:tensorflow:step = 33901, loss = 0.732093, precision = 0.820312 (58.325 sec) +INFO:tensorflow:global_step/sec: 1.7139 +INFO:tensorflow:step = 34001, loss = 0.728264, precision = 0.875 (58.347 sec) +Saved checkpoint after 87 epoch(s) to data/resnet20/checkpoints/00087... +INFO:tensorflow:global_step/sec: 1.68318 +INFO:tensorflow:step = 34101, loss = 0.778028, precision = 0.84375 (59.411 sec) +INFO:tensorflow:global_step/sec: 1.68187 +INFO:tensorflow:step = 34201, loss = 0.762506, precision = 0.851562 (59.458 sec) +INFO:tensorflow:global_step/sec: 1.70059 +INFO:tensorflow:step = 34301, loss = 0.728915, precision = 0.84375 (58.803 sec) +INFO:tensorflow:global_step/sec: 1.68039 +INFO:tensorflow:step = 34401, loss = 0.683559, precision = 0.859375 (59.510 sec) +Saved checkpoint after 88 epoch(s) to data/resnet20/checkpoints/00088... +INFO:tensorflow:global_step/sec: 1.68467 +INFO:tensorflow:step = 34501, loss = 0.809355, precision = 0.8125 (59.359 sec) +INFO:tensorflow:global_step/sec: 1.69165 +INFO:tensorflow:step = 34601, loss = 0.65229, precision = 0.882812 (59.114 sec) +INFO:tensorflow:global_step/sec: 1.68701 +INFO:tensorflow:step = 34701, loss = 0.631799, precision = 0.890625 (59.276 sec) +Saved checkpoint after 89 epoch(s) to data/resnet20/checkpoints/00089... +INFO:tensorflow:global_step/sec: 1.70207 +INFO:tensorflow:step = 34801, loss = 0.653298, precision = 0.84375 (58.752 sec) +INFO:tensorflow:global_step/sec: 1.69257 +INFO:tensorflow:step = 34901, loss = 0.55766, precision = 0.890625 (59.082 sec) +INFO:tensorflow:global_step/sec: 1.66176 +INFO:tensorflow:step = 35001, loss = 0.631039, precision = 0.859375 (60.177 sec) +INFO:tensorflow:global_step/sec: 1.66591 +INFO:tensorflow:step = 35101, loss = 0.615555, precision = 0.882812 (60.027 sec) +Saved checkpoint after 90 epoch(s) to data/resnet20/checkpoints/00090... +INFO:tensorflow:global_step/sec: 1.62702 +INFO:tensorflow:step = 35201, loss = 0.56889, precision = 0.898438 (61.462 sec) +INFO:tensorflow:global_step/sec: 1.68581 +INFO:tensorflow:step = 35301, loss = 0.580655, precision = 0.914062 (59.319 sec) +INFO:tensorflow:global_step/sec: 1.65849 +INFO:tensorflow:step = 35401, loss = 0.732215, precision = 0.820312 (60.296 sec) +INFO:tensorflow:global_step/sec: 1.55891 +INFO:tensorflow:step = 35501, loss = 0.682885, precision = 0.851562 (64.147 sec) +Saved checkpoint after 91 epoch(s) to data/resnet20/checkpoints/00091... +INFO:tensorflow:global_step/sec: 1.47138 +INFO:tensorflow:step = 35601, loss = 0.484127, precision = 0.921875 (67.963 sec) +INFO:tensorflow:global_step/sec: 1.60746 +INFO:tensorflow:step = 35701, loss = 0.566095, precision = 0.90625 (62.210 sec) +INFO:tensorflow:global_step/sec: 1.6092 +INFO:tensorflow:step = 35801, loss = 0.482403, precision = 0.9375 (62.142 sec) +INFO:tensorflow:global_step/sec: 1.30392 +INFO:tensorflow:step = 35901, loss = 0.481587, precision = 0.90625 (76.692 sec) +Saved checkpoint after 92 epoch(s) to data/resnet20/checkpoints/00092... +INFO:tensorflow:global_step/sec: 1.48739 +INFO:tensorflow:step = 36001, loss = 0.426908, precision = 0.976562 (67.232 sec) +INFO:tensorflow:global_step/sec: 1.50638 +INFO:tensorflow:step = 36101, loss = 0.446804, precision = 0.9375 (66.384 sec) +INFO:tensorflow:global_step/sec: 1.5573 +INFO:tensorflow:step = 36201, loss = 0.425318, precision = 0.953125 (64.214 sec) +INFO:tensorflow:global_step/sec: 1.53251 +INFO:tensorflow:step = 36301, loss = 0.393316, precision = 0.960938 (65.252 sec) +Saved checkpoint after 93 epoch(s) to data/resnet20/checkpoints/00093... +INFO:tensorflow:global_step/sec: 1.47569 +INFO:tensorflow:step = 36401, loss = 0.408264, precision = 0.96875 (67.765 sec) +INFO:tensorflow:global_step/sec: 1.49888 +INFO:tensorflow:step = 36501, loss = 0.433322, precision = 0.960938 (66.716 sec) +INFO:tensorflow:global_step/sec: 1.5428 +INFO:tensorflow:step = 36601, loss = 0.455872, precision = 0.921875 (64.817 sec) +INFO:tensorflow:global_step/sec: 1.59849 +INFO:tensorflow:step = 36701, loss = 0.466203, precision = 0.9375 (62.559 sec) +Saved checkpoint after 94 epoch(s) to data/resnet20/checkpoints/00094... +INFO:tensorflow:global_step/sec: 1.56738 +INFO:tensorflow:step = 36801, loss = 0.42228, precision = 0.90625 (63.801 sec) +INFO:tensorflow:global_step/sec: 1.59129 +INFO:tensorflow:step = 36901, loss = 0.385611, precision = 0.9375 (62.842 sec) +INFO:tensorflow:global_step/sec: 1.59275 +INFO:tensorflow:step = 37001, loss = 0.466537, precision = 0.90625 (62.784 sec) +INFO:tensorflow:global_step/sec: 1.57289 +INFO:tensorflow:step = 37101, loss = 0.377385, precision = 0.953125 (63.577 sec) +Saved checkpoint after 95 epoch(s) to data/resnet20/checkpoints/00095... +INFO:tensorflow:global_step/sec: 1.5598 +INFO:tensorflow:step = 37201, loss = 0.359912, precision = 0.960938 (64.111 sec) +INFO:tensorflow:global_step/sec: 1.60209 +INFO:tensorflow:step = 37301, loss = 0.405989, precision = 0.953125 (62.418 sec) +INFO:tensorflow:global_step/sec: 1.63764 +INFO:tensorflow:step = 37401, loss = 0.324291, precision = 0.96875 (61.063 sec) +INFO:tensorflow:global_step/sec: 1.56681 +INFO:tensorflow:step = 37501, loss = 0.365355, precision = 0.960938 (63.824 sec) +Saved checkpoint after 96 epoch(s) to data/resnet20/checkpoints/00096... +INFO:tensorflow:global_step/sec: 1.57368 +INFO:tensorflow:step = 37601, loss = 0.360863, precision = 0.960938 (63.545 sec) +INFO:tensorflow:global_step/sec: 1.60884 +INFO:tensorflow:step = 37701, loss = 0.39795, precision = 0.945312 (62.157 sec) +INFO:tensorflow:global_step/sec: 1.60283 +INFO:tensorflow:step = 37801, loss = 0.348515, precision = 0.960938 (62.390 sec) +INFO:tensorflow:global_step/sec: 1.64558 +INFO:tensorflow:step = 37901, loss = 0.402141, precision = 0.945312 (60.769 sec) +Saved checkpoint after 97 epoch(s) to data/resnet20/checkpoints/00097... +INFO:tensorflow:global_step/sec: 1.64635 +INFO:tensorflow:step = 38001, loss = 0.34945, precision = 0.960938 (60.741 sec) +INFO:tensorflow:global_step/sec: 1.67586 +INFO:tensorflow:step = 38101, loss = 0.339284, precision = 0.953125 (59.671 sec) +INFO:tensorflow:global_step/sec: 1.63905 +INFO:tensorflow:step = 38201, loss = 0.304927, precision = 0.992188 (61.011 sec) +INFO:tensorflow:global_step/sec: 1.64464 +INFO:tensorflow:step = 38301, loss = 0.343136, precision = 0.976562 (60.804 sec) +Saved checkpoint after 98 epoch(s) to data/resnet20/checkpoints/00098... +INFO:tensorflow:global_step/sec: 1.63097 +INFO:tensorflow:step = 38401, loss = 0.336744, precision = 0.96875 (61.313 sec) +INFO:tensorflow:global_step/sec: 1.6522 +INFO:tensorflow:step = 38501, loss = 0.412397, precision = 0.921875 (60.526 sec) +INFO:tensorflow:global_step/sec: 1.62792 +INFO:tensorflow:step = 38601, loss = 0.3563, precision = 0.96875 (61.428 sec) +INFO:tensorflow:global_step/sec: 1.58077 +INFO:tensorflow:step = 38701, loss = 0.367107, precision = 0.953125 (63.260 sec) +Saved checkpoint after 99 epoch(s) to data/resnet20/checkpoints/00099... +INFO:tensorflow:global_step/sec: 1.55491 +INFO:tensorflow:step = 38801, loss = 0.317979, precision = 0.953125 (64.312 sec) +INFO:tensorflow:global_step/sec: 1.59216 +INFO:tensorflow:step = 38901, loss = 0.390227, precision = 0.945312 (62.808 sec) +INFO:tensorflow:global_step/sec: 1.63854 +INFO:tensorflow:step = 39001, loss = 0.33497, precision = 0.96875 (61.030 sec) +Saved checkpoint after 100 epoch(s) to data/resnet20/checkpoints/00100... +INFO:tensorflow:global_step/sec: 1.63487 +INFO:tensorflow:step = 39101, loss = 0.360565, precision = 0.9375 (61.167 sec) +INFO:tensorflow:global_step/sec: 1.64347 +INFO:tensorflow:step = 39201, loss = 0.377058, precision = 0.9375 (60.847 sec) +INFO:tensorflow:global_step/sec: 1.57323 +INFO:tensorflow:step = 39301, loss = 0.290508, precision = 0.976562 (63.564 sec) +INFO:tensorflow:global_step/sec: 1.63171 +INFO:tensorflow:step = 39401, loss = 0.365144, precision = 0.945312 (61.285 sec) +Saved checkpoint after 101 epoch(s) to data/resnet20/checkpoints/00101... +INFO:tensorflow:global_step/sec: 1.64887 +INFO:tensorflow:step = 39501, loss = 0.351264, precision = 0.953125 (60.648 sec) +INFO:tensorflow:global_step/sec: 1.60775 +INFO:tensorflow:step = 39601, loss = 0.347081, precision = 0.929688 (62.199 sec) +INFO:tensorflow:global_step/sec: 1.5734 +INFO:tensorflow:step = 39701, loss = 0.281163, precision = 0.976562 (63.557 sec) +INFO:tensorflow:global_step/sec: 1.54816 +INFO:tensorflow:step = 39801, loss = 0.298156, precision = 0.960938 (64.593 sec) +Saved checkpoint after 102 epoch(s) to data/resnet20/checkpoints/00102... +INFO:tensorflow:global_step/sec: 1.53362 +INFO:tensorflow:step = 39901, loss = 0.330979, precision = 0.945312 (65.205 sec) +INFO:tensorflow:global_step/sec: 1.5234 +INFO:tensorflow:step = 40001, loss = 0.319242, precision = 0.960938 (65.643 sec) +INFO:tensorflow:global_step/sec: 1.56033 +INFO:tensorflow:step = 40101, loss = 0.371588, precision = 0.9375 (64.089 sec) +INFO:tensorflow:global_step/sec: 1.53392 +INFO:tensorflow:step = 40201, loss = 0.357117, precision = 0.945312 (65.192 sec) +Saved checkpoint after 103 epoch(s) to data/resnet20/checkpoints/00103... +INFO:tensorflow:global_step/sec: 1.52643 +INFO:tensorflow:step = 40301, loss = 0.385138, precision = 0.882812 (65.513 sec) +INFO:tensorflow:global_step/sec: 1.5813 +INFO:tensorflow:step = 40401, loss = 0.295301, precision = 0.96875 (63.239 sec) +INFO:tensorflow:global_step/sec: 1.60286 +INFO:tensorflow:step = 40501, loss = 0.344314, precision = 0.960938 (62.388 sec) +INFO:tensorflow:global_step/sec: 1.63411 +INFO:tensorflow:step = 40601, loss = 0.313872, precision = 0.945312 (61.195 sec) +Saved checkpoint after 104 epoch(s) to data/resnet20/checkpoints/00104... +INFO:tensorflow:global_step/sec: 1.61033 +INFO:tensorflow:step = 40701, loss = 0.265549, precision = 0.96875 (62.099 sec) +INFO:tensorflow:global_step/sec: 1.63103 +INFO:tensorflow:step = 40801, loss = 0.316212, precision = 0.960938 (61.311 sec) +INFO:tensorflow:global_step/sec: 1.59956 +INFO:tensorflow:step = 40901, loss = 0.307641, precision = 0.96875 (62.517 sec) +INFO:tensorflow:global_step/sec: 1.62447 +INFO:tensorflow:step = 41001, loss = 0.325884, precision = 0.953125 (61.558 sec) +Saved checkpoint after 105 epoch(s) to data/resnet20/checkpoints/00105... +INFO:tensorflow:global_step/sec: 1.57602 +INFO:tensorflow:step = 41101, loss = 0.270561, precision = 0.960938 (63.451 sec) +INFO:tensorflow:global_step/sec: 1.67464 +INFO:tensorflow:step = 41201, loss = 0.255644, precision = 0.976562 (59.714 sec) +INFO:tensorflow:global_step/sec: 1.58622 +INFO:tensorflow:step = 41301, loss = 0.32862, precision = 0.9375 (63.043 sec) +INFO:tensorflow:global_step/sec: 1.51847 +INFO:tensorflow:step = 41401, loss = 0.274544, precision = 0.976562 (65.856 sec) +Saved checkpoint after 106 epoch(s) to data/resnet20/checkpoints/00106... +INFO:tensorflow:global_step/sec: 1.6297 +INFO:tensorflow:step = 41501, loss = 0.292253, precision = 0.960938 (61.361 sec) +INFO:tensorflow:global_step/sec: 1.64103 +INFO:tensorflow:step = 41601, loss = 0.308924, precision = 0.945312 (60.937 sec) +INFO:tensorflow:global_step/sec: 1.65263 +INFO:tensorflow:step = 41701, loss = 0.355481, precision = 0.921875 (60.510 sec) +INFO:tensorflow:global_step/sec: 1.64228 +INFO:tensorflow:step = 41801, loss = 0.25141, precision = 0.976562 (60.891 sec) +Saved checkpoint after 107 epoch(s) to data/resnet20/checkpoints/00107... +INFO:tensorflow:global_step/sec: 1.51338 +INFO:tensorflow:step = 41901, loss = 0.28094, precision = 0.960938 (66.077 sec) +INFO:tensorflow:global_step/sec: 1.64133 +INFO:tensorflow:step = 42001, loss = 0.283499, precision = 0.960938 (60.926 sec) +INFO:tensorflow:global_step/sec: 1.65588 +INFO:tensorflow:step = 42101, loss = 0.381438, precision = 0.9375 (60.391 sec) +INFO:tensorflow:global_step/sec: 1.6924 +INFO:tensorflow:step = 42201, loss = 0.236191, precision = 0.992188 (59.088 sec) +Saved checkpoint after 108 epoch(s) to data/resnet20/checkpoints/00108... +INFO:tensorflow:global_step/sec: 1.67453 +INFO:tensorflow:step = 42301, loss = 0.266886, precision = 0.96875 (59.719 sec) +INFO:tensorflow:global_step/sec: 1.66197 +INFO:tensorflow:step = 42401, loss = 0.336089, precision = 0.9375 (60.170 sec) +INFO:tensorflow:global_step/sec: 1.6364 +INFO:tensorflow:step = 42501, loss = 0.311719, precision = 0.945312 (61.109 sec) +INFO:tensorflow:global_step/sec: 1.64436 +INFO:tensorflow:step = 42601, loss = 0.309929, precision = 0.9375 (60.814 sec) +Saved checkpoint after 109 epoch(s) to data/resnet20/checkpoints/00109... +INFO:tensorflow:global_step/sec: 1.59978 +INFO:tensorflow:step = 42701, loss = 0.25736, precision = 0.976562 (62.509 sec) +INFO:tensorflow:global_step/sec: 1.64067 +INFO:tensorflow:step = 42801, loss = 0.267402, precision = 0.960938 (60.950 sec) +INFO:tensorflow:global_step/sec: 1.66655 +INFO:tensorflow:step = 42901, loss = 0.252441, precision = 0.976562 (60.004 sec) +INFO:tensorflow:global_step/sec: 1.66666 +INFO:tensorflow:step = 43001, loss = 0.331864, precision = 0.945312 (60.000 sec) +Saved checkpoint after 110 epoch(s) to data/resnet20/checkpoints/00110... +INFO:tensorflow:global_step/sec: 1.66035 +INFO:tensorflow:step = 43101, loss = 0.217418, precision = 0.992188 (60.228 sec) +INFO:tensorflow:global_step/sec: 1.62813 +INFO:tensorflow:step = 43201, loss = 0.239731, precision = 0.976562 (61.420 sec) +INFO:tensorflow:global_step/sec: 1.69944 +INFO:tensorflow:step = 43301, loss = 0.252849, precision = 0.976562 (58.843 sec) +Saved checkpoint after 111 epoch(s) to data/resnet20/checkpoints/00111... +INFO:tensorflow:global_step/sec: 1.65047 +INFO:tensorflow:step = 43401, loss = 0.228311, precision = 0.976562 (60.589 sec) +INFO:tensorflow:global_step/sec: 1.66077 +INFO:tensorflow:step = 43501, loss = 0.32464, precision = 0.953125 (60.213 sec) +INFO:tensorflow:global_step/sec: 1.59669 +INFO:tensorflow:step = 43601, loss = 0.281085, precision = 0.960938 (62.629 sec) +INFO:tensorflow:global_step/sec: 1.60703 +INFO:tensorflow:step = 43701, loss = 0.242351, precision = 0.976562 (62.227 sec) +Saved checkpoint after 112 epoch(s) to data/resnet20/checkpoints/00112... +INFO:tensorflow:global_step/sec: 1.58324 +INFO:tensorflow:step = 43801, loss = 0.248166, precision = 0.976562 (63.162 sec) +INFO:tensorflow:global_step/sec: 1.5991 +INFO:tensorflow:step = 43901, loss = 0.274967, precision = 0.953125 (62.535 sec) +INFO:tensorflow:global_step/sec: 1.65612 +INFO:tensorflow:step = 44001, loss = 0.260651, precision = 0.96875 (60.382 sec) +INFO:tensorflow:global_step/sec: 1.69463 +INFO:tensorflow:step = 44101, loss = 0.242082, precision = 0.96875 (59.010 sec) +Saved checkpoint after 113 epoch(s) to data/resnet20/checkpoints/00113... +INFO:tensorflow:global_step/sec: 1.6863 +INFO:tensorflow:step = 44201, loss = 0.295056, precision = 0.9375 (59.302 sec) +INFO:tensorflow:global_step/sec: 1.5826 +INFO:tensorflow:step = 44301, loss = 0.264542, precision = 0.976562 (63.187 sec) +INFO:tensorflow:global_step/sec: 1.6603 +INFO:tensorflow:step = 44401, loss = 0.245796, precision = 0.96875 (60.230 sec) +INFO:tensorflow:global_step/sec: 1.55268 +INFO:tensorflow:step = 44501, loss = 0.316432, precision = 0.945312 (64.405 sec) +Saved checkpoint after 114 epoch(s) to data/resnet20/checkpoints/00114... +INFO:tensorflow:global_step/sec: 1.62294 +INFO:tensorflow:step = 44601, loss = 0.304341, precision = 0.9375 (61.617 sec) +INFO:tensorflow:global_step/sec: 1.6855 +INFO:tensorflow:step = 44701, loss = 0.260245, precision = 0.96875 (59.329 sec) +INFO:tensorflow:global_step/sec: 1.61904 +INFO:tensorflow:step = 44801, loss = 0.274216, precision = 0.960938 (61.765 sec) +INFO:tensorflow:global_step/sec: 1.5551 +INFO:tensorflow:step = 44901, loss = 0.215418, precision = 0.976562 (64.304 sec) +Saved checkpoint after 115 epoch(s) to data/resnet20/checkpoints/00115... +INFO:tensorflow:global_step/sec: 1.56146 +INFO:tensorflow:step = 45001, loss = 0.280278, precision = 0.96875 (64.043 sec) +INFO:tensorflow:global_step/sec: 1.68449 +INFO:tensorflow:step = 45101, loss = 0.369626, precision = 0.921875 (59.365 sec) +INFO:tensorflow:global_step/sec: 1.67702 +INFO:tensorflow:step = 45201, loss = 0.333869, precision = 0.929688 (59.629 sec) +INFO:tensorflow:global_step/sec: 1.65396 +INFO:tensorflow:step = 45301, loss = 0.237899, precision = 0.976562 (60.461 sec) +Saved checkpoint after 116 epoch(s) to data/resnet20/checkpoints/00116... +INFO:tensorflow:global_step/sec: 1.58946 +INFO:tensorflow:step = 45401, loss = 0.240596, precision = 0.960938 (62.914 sec) +INFO:tensorflow:global_step/sec: 1.67867 +INFO:tensorflow:step = 45501, loss = 0.290977, precision = 0.953125 (59.571 sec) +INFO:tensorflow:global_step/sec: 1.64236 +INFO:tensorflow:step = 45601, loss = 0.256581, precision = 0.960938 (60.888 sec) +INFO:tensorflow:global_step/sec: 1.62846 +INFO:tensorflow:step = 45701, loss = 0.310004, precision = 0.945312 (61.408 sec) +Saved checkpoint after 117 epoch(s) to data/resnet20/checkpoints/00117... +INFO:tensorflow:global_step/sec: 1.64568 +INFO:tensorflow:step = 45801, loss = 0.287182, precision = 0.960938 (60.765 sec) +INFO:tensorflow:global_step/sec: 1.7125 +INFO:tensorflow:step = 45901, loss = 0.301148, precision = 0.929688 (58.394 sec) +INFO:tensorflow:global_step/sec: 1.74165 +INFO:tensorflow:step = 46001, loss = 0.252003, precision = 0.960938 (57.417 sec) +INFO:tensorflow:global_step/sec: 1.75066 +INFO:tensorflow:step = 46101, loss = 0.251194, precision = 0.953125 (57.121 sec) +Saved checkpoint after 118 epoch(s) to data/resnet20/checkpoints/00118... +INFO:tensorflow:global_step/sec: 1.70143 +INFO:tensorflow:step = 46201, loss = 0.307219, precision = 0.9375 (58.774 sec) +INFO:tensorflow:global_step/sec: 1.70749 +INFO:tensorflow:step = 46301, loss = 0.295515, precision = 0.945312 (58.566 sec) +INFO:tensorflow:global_step/sec: 1.72197 +INFO:tensorflow:step = 46401, loss = 0.255685, precision = 0.953125 (58.073 sec) +INFO:tensorflow:global_step/sec: 1.69031 +INFO:tensorflow:step = 46501, loss = 0.248941, precision = 0.960938 (59.161 sec) +Saved checkpoint after 119 epoch(s) to data/resnet20/checkpoints/00119... +INFO:tensorflow:global_step/sec: 1.58458 +INFO:tensorflow:step = 46601, loss = 0.315532, precision = 0.945312 (63.108 sec) +INFO:tensorflow:global_step/sec: 1.65317 +INFO:tensorflow:step = 46701, loss = 0.26603, precision = 0.953125 (60.490 sec) +INFO:tensorflow:global_step/sec: 1.67837 +INFO:tensorflow:step = 46801, loss = 0.25236, precision = 0.960938 (59.582 sec) +INFO:tensorflow:global_step/sec: 1.60007 +INFO:tensorflow:step = 46901, loss = 0.242898, precision = 0.960938 (62.497 sec) +Saved checkpoint after 120 epoch(s) to data/resnet20/checkpoints/00120... +INFO:tensorflow:global_step/sec: 1.54296 +INFO:tensorflow:step = 47001, loss = 0.23517, precision = 0.96875 (64.811 sec) +INFO:tensorflow:global_step/sec: 1.64428 +INFO:tensorflow:step = 47101, loss = 0.276869, precision = 0.9375 (60.816 sec) +INFO:tensorflow:global_step/sec: 1.65463 +INFO:tensorflow:step = 47201, loss = 0.28184, precision = 0.953125 (60.437 sec) +INFO:tensorflow:global_step/sec: 1.60327 +INFO:tensorflow:step = 47301, loss = 0.314791, precision = 0.945312 (62.372 sec) +Saved checkpoint after 121 epoch(s) to data/resnet20/checkpoints/00121... +INFO:tensorflow:global_step/sec: 1.6063 +INFO:tensorflow:step = 47401, loss = 0.263736, precision = 0.960938 (62.255 sec) +INFO:tensorflow:global_step/sec: 1.64232 +INFO:tensorflow:step = 47501, loss = 0.222225, precision = 0.960938 (60.890 sec) +INFO:tensorflow:global_step/sec: 1.60774 +INFO:tensorflow:step = 47601, loss = 0.257497, precision = 0.960938 (62.199 sec) +INFO:tensorflow:global_step/sec: 1.49722 +INFO:tensorflow:step = 47701, loss = 0.233177, precision = 0.960938 (66.790 sec) +Saved checkpoint after 122 epoch(s) to data/resnet20/checkpoints/00122... +INFO:tensorflow:global_step/sec: 1.52875 +INFO:tensorflow:step = 47801, loss = 0.257711, precision = 0.9375 (65.413 sec) +INFO:tensorflow:global_step/sec: 1.52427 +INFO:tensorflow:step = 47901, loss = 0.265606, precision = 0.96875 (65.605 sec) +INFO:tensorflow:global_step/sec: 1.56288 +INFO:tensorflow:step = 48001, loss = 0.307257, precision = 0.929688 (63.985 sec) +Saved checkpoint after 123 epoch(s) to data/resnet20/checkpoints/00123... +INFO:tensorflow:global_step/sec: 1.5385 +INFO:tensorflow:step = 48101, loss = 0.197013, precision = 0.976562 (64.999 sec) +INFO:tensorflow:global_step/sec: 1.62513 +INFO:tensorflow:step = 48201, loss = 0.241494, precision = 0.976562 (61.533 sec) +INFO:tensorflow:global_step/sec: 1.57825 +INFO:tensorflow:step = 48301, loss = 0.255441, precision = 0.945312 (63.361 sec) +INFO:tensorflow:global_step/sec: 1.60556 +INFO:tensorflow:step = 48401, loss = 0.286824, precision = 0.953125 (62.284 sec) +Saved checkpoint after 124 epoch(s) to data/resnet20/checkpoints/00124... +INFO:tensorflow:global_step/sec: 1.60324 +INFO:tensorflow:step = 48501, loss = 0.189425, precision = 0.984375 (62.374 sec) +INFO:tensorflow:global_step/sec: 1.61695 +INFO:tensorflow:step = 48601, loss = 0.268965, precision = 0.953125 (61.845 sec) +INFO:tensorflow:global_step/sec: 1.69249 +INFO:tensorflow:step = 48701, loss = 0.32743, precision = 0.921875 (59.084 sec) +INFO:tensorflow:global_step/sec: 1.67772 +INFO:tensorflow:step = 48801, loss = 0.253809, precision = 0.960938 (59.605 sec) +Saved checkpoint after 125 epoch(s) to data/resnet20/checkpoints/00125... +INFO:tensorflow:global_step/sec: 1.6727 +INFO:tensorflow:step = 48901, loss = 0.30741, precision = 0.929688 (59.784 sec) +INFO:tensorflow:global_step/sec: 1.68517 +INFO:tensorflow:step = 49001, loss = 0.245243, precision = 0.960938 (59.341 sec) +INFO:tensorflow:global_step/sec: 1.64129 +INFO:tensorflow:step = 49101, loss = 0.200764, precision = 0.992188 (60.928 sec) +INFO:tensorflow:global_step/sec: 1.67691 +INFO:tensorflow:step = 49201, loss = 0.184797, precision = 0.992188 (59.634 sec) +Saved checkpoint after 126 epoch(s) to data/resnet20/checkpoints/00126... +INFO:tensorflow:global_step/sec: 1.69759 +INFO:tensorflow:step = 49301, loss = 0.264574, precision = 0.953125 (58.907 sec) +INFO:tensorflow:global_step/sec: 1.66707 +INFO:tensorflow:step = 49401, loss = 0.271215, precision = 0.960938 (59.986 sec) +INFO:tensorflow:global_step/sec: 1.59733 +INFO:tensorflow:step = 49501, loss = 0.253923, precision = 0.953125 (62.604 sec) +INFO:tensorflow:global_step/sec: 1.58925 +INFO:tensorflow:step = 49601, loss = 0.260622, precision = 0.960938 (62.923 sec) +Saved checkpoint after 127 epoch(s) to data/resnet20/checkpoints/00127... +INFO:tensorflow:global_step/sec: 1.49552 +INFO:tensorflow:step = 49701, loss = 0.254497, precision = 0.960938 (66.867 sec) +INFO:tensorflow:global_step/sec: 1.57897 +INFO:tensorflow:step = 49801, loss = 0.204944, precision = 0.96875 (63.332 sec) +INFO:tensorflow:global_step/sec: 1.64875 +INFO:tensorflow:step = 49901, loss = 0.231004, precision = 0.960938 (60.652 sec) +INFO:tensorflow:global_step/sec: 1.6484 +INFO:tensorflow:step = 50001, loss = 0.231868, precision = 0.960938 (60.666 sec) +Saved checkpoint after 128 epoch(s) to data/resnet20/checkpoints/00128... +INFO:tensorflow:global_step/sec: 1.58122 +INFO:tensorflow:step = 50101, loss = 0.24302, precision = 0.9375 (63.241 sec) +INFO:tensorflow:global_step/sec: 1.56073 +INFO:tensorflow:step = 50201, loss = 0.194574, precision = 0.992188 (64.072 sec) +INFO:tensorflow:global_step/sec: 1.63564 +INFO:tensorflow:step = 50301, loss = 0.221169, precision = 0.976562 (61.138 sec) +INFO:tensorflow:global_step/sec: 1.6521 +INFO:tensorflow:step = 50401, loss = 0.219385, precision = 0.992188 (60.529 sec) +Saved checkpoint after 129 epoch(s) to data/resnet20/checkpoints/00129... +INFO:tensorflow:global_step/sec: 1.65774 +INFO:tensorflow:step = 50501, loss = 0.262967, precision = 0.960938 (60.323 sec) +INFO:tensorflow:global_step/sec: 1.6741 +INFO:tensorflow:step = 50601, loss = 0.229124, precision = 0.960938 (59.734 sec) +INFO:tensorflow:global_step/sec: 1.67234 +INFO:tensorflow:step = 50701, loss = 0.248846, precision = 0.960938 (59.796 sec) +INFO:tensorflow:global_step/sec: 1.68998 +INFO:tensorflow:step = 50801, loss = 0.220136, precision = 0.96875 (59.172 sec) +Saved checkpoint after 130 epoch(s) to data/resnet20/checkpoints/00130... +INFO:tensorflow:global_step/sec: 1.66907 +INFO:tensorflow:step = 50901, loss = 0.290157, precision = 0.9375 (59.914 sec) +INFO:tensorflow:global_step/sec: 1.56673 +INFO:tensorflow:step = 51001, loss = 0.309163, precision = 0.9375 (63.827 sec) +INFO:tensorflow:global_step/sec: 1.55503 +INFO:tensorflow:step = 51101, loss = 0.240906, precision = 0.960938 (64.307 sec) +INFO:tensorflow:global_step/sec: 1.54664 +INFO:tensorflow:step = 51201, loss = 0.226536, precision = 0.976562 (64.656 sec) +Saved checkpoint after 131 epoch(s) to data/resnet20/checkpoints/00131... +INFO:tensorflow:global_step/sec: 1.58669 +INFO:tensorflow:step = 51301, loss = 0.215821, precision = 0.960938 (63.024 sec) +INFO:tensorflow:global_step/sec: 1.66482 +INFO:tensorflow:step = 51401, loss = 0.185743, precision = 0.984375 (60.067 sec) +INFO:tensorflow:global_step/sec: 1.6183 +INFO:tensorflow:step = 51501, loss = 0.26119, precision = 0.953125 (61.793 sec) +INFO:tensorflow:global_step/sec: 1.67958 +INFO:tensorflow:step = 51601, loss = 0.222166, precision = 0.96875 (59.539 sec) +Saved checkpoint after 132 epoch(s) to data/resnet20/checkpoints/00132... +INFO:tensorflow:global_step/sec: 1.61658 +INFO:tensorflow:step = 51701, loss = 0.200709, precision = 0.976562 (61.859 sec) +INFO:tensorflow:global_step/sec: 1.65208 +INFO:tensorflow:step = 51801, loss = 0.239019, precision = 0.960938 (60.530 sec) +INFO:tensorflow:global_step/sec: 1.69902 +INFO:tensorflow:step = 51901, loss = 0.289613, precision = 0.945312 (58.858 sec) +INFO:tensorflow:global_step/sec: 1.66712 +INFO:tensorflow:step = 52001, loss = 0.249897, precision = 0.960938 (59.984 sec) +Saved checkpoint after 133 epoch(s) to data/resnet20/checkpoints/00133... +INFO:tensorflow:global_step/sec: 1.65348 +INFO:tensorflow:step = 52101, loss = 0.21605, precision = 0.976562 (60.479 sec) +INFO:tensorflow:global_step/sec: 1.65103 +INFO:tensorflow:step = 52201, loss = 0.192032, precision = 0.984375 (60.568 sec) +INFO:tensorflow:global_step/sec: 1.66187 +INFO:tensorflow:step = 52301, loss = 0.293915, precision = 0.945312 (60.173 sec) +Saved checkpoint after 134 epoch(s) to data/resnet20/checkpoints/00134... +INFO:tensorflow:global_step/sec: 1.67147 +INFO:tensorflow:step = 52401, loss = 0.185583, precision = 0.992188 (59.827 sec) +INFO:tensorflow:global_step/sec: 1.70526 +INFO:tensorflow:step = 52501, loss = 0.21345, precision = 0.945312 (58.642 sec) +INFO:tensorflow:global_step/sec: 1.67649 +INFO:tensorflow:step = 52601, loss = 0.254541, precision = 0.953125 (59.649 sec) +INFO:tensorflow:global_step/sec: 1.65846 +INFO:tensorflow:step = 52701, loss = 0.251779, precision = 0.953125 (60.297 sec) +Saved checkpoint after 135 epoch(s) to data/resnet20/checkpoints/00135... +INFO:tensorflow:global_step/sec: 1.6673 +INFO:tensorflow:step = 52801, loss = 0.222814, precision = 0.976562 (59.977 sec) +INFO:tensorflow:global_step/sec: 1.66734 +INFO:tensorflow:step = 52901, loss = 0.236412, precision = 0.976562 (59.976 sec) +INFO:tensorflow:global_step/sec: 1.6194 +INFO:tensorflow:step = 53001, loss = 0.267392, precision = 0.960938 (61.752 sec) +INFO:tensorflow:global_step/sec: 1.66979 +INFO:tensorflow:step = 53101, loss = 0.219929, precision = 0.96875 (59.888 sec) +Saved checkpoint after 136 epoch(s) to data/resnet20/checkpoints/00136... +INFO:tensorflow:global_step/sec: 1.63323 +INFO:tensorflow:step = 53201, loss = 0.233426, precision = 0.953125 (61.228 sec) +INFO:tensorflow:global_step/sec: 1.61775 +INFO:tensorflow:step = 53301, loss = 0.1796, precision = 0.976562 (61.814 sec) +INFO:tensorflow:global_step/sec: 1.63255 +INFO:tensorflow:step = 53401, loss = 0.219874, precision = 0.96875 (61.254 sec) +INFO:tensorflow:global_step/sec: 1.56298 +INFO:tensorflow:step = 53501, loss = 0.170055, precision = 0.992188 (63.980 sec) +Saved checkpoint after 137 epoch(s) to data/resnet20/checkpoints/00137... +INFO:tensorflow:global_step/sec: 1.5746 +INFO:tensorflow:step = 53601, loss = 0.179115, precision = 0.984375 (63.508 sec) +INFO:tensorflow:global_step/sec: 1.67438 +INFO:tensorflow:step = 53701, loss = 0.183916, precision = 1.0 (59.723 sec) +INFO:tensorflow:global_step/sec: 1.6878 +INFO:tensorflow:step = 53801, loss = 0.153993, precision = 1.0 (59.249 sec) +INFO:tensorflow:global_step/sec: 1.61753 +INFO:tensorflow:step = 53901, loss = 0.189131, precision = 0.984375 (61.823 sec) +Saved checkpoint after 138 epoch(s) to data/resnet20/checkpoints/00138... +INFO:tensorflow:global_step/sec: 1.63132 +INFO:tensorflow:step = 54001, loss = 0.178144, precision = 0.96875 (61.300 sec) +INFO:tensorflow:global_step/sec: 1.68077 +INFO:tensorflow:step = 54101, loss = 0.178919, precision = 0.992188 (59.497 sec) +INFO:tensorflow:global_step/sec: 1.63962 +INFO:tensorflow:step = 54201, loss = 0.250199, precision = 0.96875 (60.990 sec) +INFO:tensorflow:global_step/sec: 1.66974 +INFO:tensorflow:step = 54301, loss = 0.193013, precision = 0.976562 (59.890 sec) +Saved checkpoint after 139 epoch(s) to data/resnet20/checkpoints/00139... +INFO:tensorflow:global_step/sec: 1.65995 +INFO:tensorflow:step = 54401, loss = 0.176285, precision = 0.992188 (60.243 sec) +INFO:tensorflow:global_step/sec: 1.69748 +INFO:tensorflow:step = 54501, loss = 0.17778, precision = 0.992188 (58.911 sec) +INFO:tensorflow:global_step/sec: 1.63099 +INFO:tensorflow:step = 54601, loss = 0.172721, precision = 0.984375 (61.313 sec) +INFO:tensorflow:global_step/sec: 1.48771 +INFO:tensorflow:step = 54701, loss = 0.170972, precision = 0.992188 (67.217 sec) +Saved checkpoint after 140 epoch(s) to data/resnet20/checkpoints/00140... +INFO:tensorflow:global_step/sec: 1.23501 +INFO:tensorflow:step = 54801, loss = 0.15492, precision = 0.992188 (80.971 sec) +INFO:tensorflow:global_step/sec: 1.40158 +INFO:tensorflow:step = 54901, loss = 0.165294, precision = 0.992188 (71.348 sec) +INFO:tensorflow:global_step/sec: 1.70921 +INFO:tensorflow:step = 55001, loss = 0.166201, precision = 1.0 (58.507 sec) +INFO:tensorflow:global_step/sec: 1.62412 +INFO:tensorflow:step = 55101, loss = 0.220955, precision = 0.96875 (61.572 sec) +Saved checkpoint after 141 epoch(s) to data/resnet20/checkpoints/00141... +INFO:tensorflow:global_step/sec: 1.69524 +INFO:tensorflow:step = 55201, loss = 0.16271, precision = 1.0 (58.989 sec) +INFO:tensorflow:global_step/sec: 1.65542 +INFO:tensorflow:step = 55301, loss = 0.169836, precision = 0.992188 (60.408 sec) +INFO:tensorflow:global_step/sec: 1.63741 +INFO:tensorflow:step = 55401, loss = 0.182313, precision = 0.992188 (61.072 sec) +INFO:tensorflow:global_step/sec: 1.64282 +INFO:tensorflow:step = 55501, loss = 0.168494, precision = 0.992188 (60.871 sec) +Saved checkpoint after 142 epoch(s) to data/resnet20/checkpoints/00142... +INFO:tensorflow:global_step/sec: 1.60198 +INFO:tensorflow:step = 55601, loss = 0.154382, precision = 1.0 (62.423 sec) +INFO:tensorflow:global_step/sec: 1.60149 +INFO:tensorflow:step = 55701, loss = 0.169324, precision = 0.992188 (62.442 sec) +INFO:tensorflow:global_step/sec: 1.69937 +INFO:tensorflow:step = 55801, loss = 0.1989, precision = 0.976562 (58.845 sec) +INFO:tensorflow:global_step/sec: 1.68621 +INFO:tensorflow:step = 55901, loss = 0.153694, precision = 1.0 (59.304 sec) +Saved checkpoint after 143 epoch(s) to data/resnet20/checkpoints/00143... +INFO:tensorflow:global_step/sec: 1.69133 +INFO:tensorflow:step = 56001, loss = 0.169091, precision = 0.976562 (59.125 sec) +INFO:tensorflow:global_step/sec: 1.64745 +INFO:tensorflow:step = 56101, loss = 0.178017, precision = 0.984375 (60.700 sec) +INFO:tensorflow:global_step/sec: 1.71391 +INFO:tensorflow:step = 56201, loss = 0.161173, precision = 0.984375 (58.346 sec) +INFO:tensorflow:global_step/sec: 1.72599 +INFO:tensorflow:step = 56301, loss = 0.13806, precision = 1.0 (57.938 sec) +Saved checkpoint after 144 epoch(s) to data/resnet20/checkpoints/00144... +INFO:tensorflow:global_step/sec: 1.70014 +INFO:tensorflow:step = 56401, loss = 0.149893, precision = 1.0 (58.819 sec) +INFO:tensorflow:global_step/sec: 1.70723 +INFO:tensorflow:step = 56501, loss = 0.179307, precision = 0.976562 (58.574 sec) +INFO:tensorflow:global_step/sec: 1.69331 +INFO:tensorflow:step = 56601, loss = 0.178593, precision = 0.976562 (59.056 sec) +Saved checkpoint after 145 epoch(s) to data/resnet20/checkpoints/00145... +INFO:tensorflow:global_step/sec: 1.69976 +INFO:tensorflow:step = 56701, loss = 0.170256, precision = 0.992188 (58.832 sec) +INFO:tensorflow:global_step/sec: 1.70356 +INFO:tensorflow:step = 56801, loss = 0.16641, precision = 0.984375 (58.700 sec) +INFO:tensorflow:global_step/sec: 1.70959 +INFO:tensorflow:step = 56901, loss = 0.158838, precision = 1.0 (58.494 sec) +INFO:tensorflow:global_step/sec: 1.70214 +INFO:tensorflow:step = 57001, loss = 0.150166, precision = 0.992188 (58.750 sec) +Saved checkpoint after 146 epoch(s) to data/resnet20/checkpoints/00146... +INFO:tensorflow:global_step/sec: 1.67248 +INFO:tensorflow:step = 57101, loss = 0.149908, precision = 0.992188 (59.791 sec) +INFO:tensorflow:global_step/sec: 1.70514 +INFO:tensorflow:step = 57201, loss = 0.19125, precision = 0.976562 (58.646 sec) +INFO:tensorflow:global_step/sec: 1.7231 +INFO:tensorflow:step = 57301, loss = 0.171065, precision = 0.992188 (58.035 sec) +INFO:tensorflow:global_step/sec: 1.7272 +INFO:tensorflow:step = 57401, loss = 0.162507, precision = 0.984375 (57.897 sec) +Saved checkpoint after 147 epoch(s) to data/resnet20/checkpoints/00147... +INFO:tensorflow:global_step/sec: 1.71144 +INFO:tensorflow:step = 57501, loss = 0.173737, precision = 0.984375 (58.430 sec) +INFO:tensorflow:global_step/sec: 1.74068 +INFO:tensorflow:step = 57601, loss = 0.168904, precision = 0.984375 (57.449 sec) +INFO:tensorflow:global_step/sec: 1.76905 +INFO:tensorflow:step = 57701, loss = 0.181708, precision = 0.976562 (56.527 sec) +INFO:tensorflow:global_step/sec: 1.73675 +INFO:tensorflow:step = 57801, loss = 0.167938, precision = 0.984375 (57.579 sec) +Saved checkpoint after 148 epoch(s) to data/resnet20/checkpoints/00148... +INFO:tensorflow:global_step/sec: 1.72293 +INFO:tensorflow:step = 57901, loss = 0.15228, precision = 0.992188 (58.040 sec) +INFO:tensorflow:global_step/sec: 1.72412 +INFO:tensorflow:step = 58001, loss = 0.149561, precision = 1.0 (58.001 sec) +INFO:tensorflow:global_step/sec: 1.68858 +INFO:tensorflow:step = 58101, loss = 0.135696, precision = 1.0 (59.221 sec) +INFO:tensorflow:global_step/sec: 1.70989 +INFO:tensorflow:step = 58201, loss = 0.18867, precision = 0.96875 (58.483 sec) +Saved checkpoint after 149 epoch(s) to data/resnet20/checkpoints/00149... +INFO:tensorflow:global_step/sec: 1.72795 +INFO:tensorflow:step = 58301, loss = 0.229395, precision = 0.960938 (57.872 sec) +INFO:tensorflow:global_step/sec: 1.7408 +INFO:tensorflow:step = 58401, loss = 0.156363, precision = 0.992188 (57.445 sec) +INFO:tensorflow:global_step/sec: 1.7448 +INFO:tensorflow:step = 58501, loss = 0.192421, precision = 0.96875 (57.313 sec) +INFO:tensorflow:global_step/sec: 1.72826 +INFO:tensorflow:step = 58601, loss = 0.161937, precision = 0.992188 (57.861 sec) +Saved checkpoint after 150 epoch(s) to data/resnet20/checkpoints/00150... +INFO:tensorflow:global_step/sec: 1.70815 +INFO:tensorflow:step = 58701, loss = 0.148259, precision = 1.0 (58.543 sec) +INFO:tensorflow:global_step/sec: 1.71332 +INFO:tensorflow:step = 58801, loss = 0.208087, precision = 0.96875 (58.366 sec) +INFO:tensorflow:global_step/sec: 1.71198 +INFO:tensorflow:step = 58901, loss = 0.144159, precision = 1.0 (58.412 sec) +INFO:tensorflow:global_step/sec: 1.72106 +INFO:tensorflow:step = 59001, loss = 0.149188, precision = 1.0 (58.104 sec) +Saved checkpoint after 151 epoch(s) to data/resnet20/checkpoints/00151... +INFO:tensorflow:global_step/sec: 1.66738 +INFO:tensorflow:step = 59101, loss = 0.185153, precision = 0.992188 (59.975 sec) +INFO:tensorflow:global_step/sec: 1.72362 +INFO:tensorflow:step = 59201, loss = 0.151514, precision = 1.0 (58.017 sec) +INFO:tensorflow:global_step/sec: 1.71509 +INFO:tensorflow:step = 59301, loss = 0.148373, precision = 0.992188 (58.306 sec) +INFO:tensorflow:global_step/sec: 1.71312 +INFO:tensorflow:step = 59401, loss = 0.159607, precision = 0.992188 (58.373 sec) +Saved checkpoint after 152 epoch(s) to data/resnet20/checkpoints/00152... +INFO:tensorflow:global_step/sec: 1.71251 +INFO:tensorflow:step = 59501, loss = 0.198666, precision = 0.984375 (58.394 sec) +INFO:tensorflow:global_step/sec: 1.72839 +INFO:tensorflow:step = 59601, loss = 0.150449, precision = 1.0 (57.857 sec) +INFO:tensorflow:global_step/sec: 1.69409 +INFO:tensorflow:step = 59701, loss = 0.142285, precision = 1.0 (59.029 sec) +INFO:tensorflow:global_step/sec: 1.72607 +INFO:tensorflow:step = 59801, loss = 0.162986, precision = 0.984375 (57.935 sec) +Saved checkpoint after 153 epoch(s) to data/resnet20/checkpoints/00153... +INFO:tensorflow:global_step/sec: 1.73006 +INFO:tensorflow:step = 59901, loss = 0.13686, precision = 1.0 (57.802 sec) +INFO:tensorflow:global_step/sec: 1.72641 +INFO:tensorflow:step = 60001, loss = 0.141562, precision = 1.0 (57.923 sec) +INFO:tensorflow:global_step/sec: 1.70835 +INFO:tensorflow:step = 60101, loss = 0.163626, precision = 0.984375 (58.536 sec) +INFO:tensorflow:global_step/sec: 1.68059 +INFO:tensorflow:step = 60201, loss = 0.156939, precision = 0.984375 (59.503 sec) +Saved checkpoint after 154 epoch(s) to data/resnet20/checkpoints/00154... +INFO:tensorflow:global_step/sec: 1.67216 +INFO:tensorflow:step = 60301, loss = 0.163113, precision = 1.0 (59.803 sec) +INFO:tensorflow:global_step/sec: 1.68968 +INFO:tensorflow:step = 60401, loss = 0.158052, precision = 0.992188 (59.182 sec) +INFO:tensorflow:global_step/sec: 1.71392 +INFO:tensorflow:step = 60501, loss = 0.159477, precision = 0.992188 (58.346 sec) +INFO:tensorflow:global_step/sec: 1.7179 +INFO:tensorflow:step = 60601, loss = 0.153884, precision = 0.992188 (58.212 sec) +Saved checkpoint after 155 epoch(s) to data/resnet20/checkpoints/00155... +INFO:tensorflow:global_step/sec: 1.69041 +INFO:tensorflow:step = 60701, loss = 0.169179, precision = 0.992188 (59.156 sec) +INFO:tensorflow:global_step/sec: 1.73138 +INFO:tensorflow:step = 60801, loss = 0.14879, precision = 0.992188 (57.757 sec) +INFO:tensorflow:global_step/sec: 1.72555 +INFO:tensorflow:step = 60901, loss = 0.168623, precision = 0.992188 (57.953 sec) +Saved checkpoint after 156 epoch(s) to data/resnet20/checkpoints/00156... +INFO:tensorflow:global_step/sec: 1.71863 +INFO:tensorflow:step = 61001, loss = 0.149122, precision = 1.0 (58.186 sec) +INFO:tensorflow:global_step/sec: 1.7176 +INFO:tensorflow:step = 61101, loss = 0.158636, precision = 0.992188 (58.221 sec) +INFO:tensorflow:global_step/sec: 1.73619 +INFO:tensorflow:step = 61201, loss = 0.145181, precision = 0.992188 (57.598 sec) +INFO:tensorflow:global_step/sec: 1.7428 +INFO:tensorflow:step = 61301, loss = 0.159577, precision = 0.992188 (57.379 sec) +Saved checkpoint after 157 epoch(s) to data/resnet20/checkpoints/00157... +INFO:tensorflow:global_step/sec: 1.73083 +INFO:tensorflow:step = 61401, loss = 0.149384, precision = 0.992188 (57.776 sec) +INFO:tensorflow:global_step/sec: 1.58938 +INFO:tensorflow:step = 61501, loss = 0.175008, precision = 0.984375 (62.918 sec) +INFO:tensorflow:global_step/sec: 1.35622 +INFO:tensorflow:step = 61601, loss = 0.151858, precision = 0.992188 (73.734 sec) +INFO:tensorflow:global_step/sec: 1.37154 +INFO:tensorflow:step = 61701, loss = 0.179545, precision = 0.984375 (72.911 sec) +Saved checkpoint after 158 epoch(s) to data/resnet20/checkpoints/00158... +INFO:tensorflow:global_step/sec: 1.43097 +INFO:tensorflow:step = 61801, loss = 0.147914, precision = 0.992188 (69.883 sec) +INFO:tensorflow:global_step/sec: 1.42575 +INFO:tensorflow:step = 61901, loss = 0.148858, precision = 0.984375 (70.139 sec) +INFO:tensorflow:global_step/sec: 1.44002 +INFO:tensorflow:step = 62001, loss = 0.147073, precision = 1.0 (69.443 sec) +INFO:tensorflow:global_step/sec: 1.41161 +INFO:tensorflow:step = 62101, loss = 0.154441, precision = 0.992188 (70.841 sec) +Saved checkpoint after 159 epoch(s) to data/resnet20/checkpoints/00159... +INFO:tensorflow:global_step/sec: 1.43008 +INFO:tensorflow:step = 62201, loss = 0.167266, precision = 0.984375 (69.926 sec) +INFO:tensorflow:global_step/sec: 1.40725 +INFO:tensorflow:step = 62301, loss = 0.165386, precision = 0.992188 (71.061 sec) +INFO:tensorflow:global_step/sec: 1.4129 +INFO:tensorflow:step = 62401, loss = 0.150515, precision = 0.984375 (70.776 sec) +INFO:tensorflow:global_step/sec: 1.37162 +INFO:tensorflow:step = 62501, loss = 0.164386, precision = 0.992188 (72.907 sec) +Saved checkpoint after 160 epoch(s) to data/resnet20/checkpoints/00160... +INFO:tensorflow:global_step/sec: 1.38353 +INFO:tensorflow:step = 62601, loss = 0.146416, precision = 0.992188 (72.279 sec) +INFO:tensorflow:global_step/sec: 1.42794 +INFO:tensorflow:step = 62701, loss = 0.160059, precision = 0.984375 (70.031 sec) +INFO:tensorflow:global_step/sec: 1.42837 +INFO:tensorflow:step = 62801, loss = 0.191638, precision = 0.984375 (70.010 sec) +INFO:tensorflow:global_step/sec: 1.41603 +INFO:tensorflow:step = 62901, loss = 0.149816, precision = 0.992188 (70.620 sec) +Saved checkpoint after 161 epoch(s) to data/resnet20/checkpoints/00161... +INFO:tensorflow:global_step/sec: 1.41118 +INFO:tensorflow:step = 63001, loss = 0.141074, precision = 0.992188 (70.862 sec) +INFO:tensorflow:global_step/sec: 1.42621 +INFO:tensorflow:step = 63101, loss = 0.143543, precision = 1.0 (70.116 sec) +INFO:tensorflow:global_step/sec: 1.4324 +INFO:tensorflow:step = 63201, loss = 0.143842, precision = 1.0 (69.813 sec) +INFO:tensorflow:global_step/sec: 1.43987 +INFO:tensorflow:step = 63301, loss = 0.135739, precision = 1.0 (69.451 sec) +Saved checkpoint after 162 epoch(s) to data/resnet20/checkpoints/00162... +INFO:tensorflow:global_step/sec: 1.39201 +INFO:tensorflow:step = 63401, loss = 0.151771, precision = 1.0 (71.839 sec) +INFO:tensorflow:global_step/sec: 1.42945 +INFO:tensorflow:step = 63501, loss = 0.154043, precision = 0.992188 (69.957 sec) +INFO:tensorflow:global_step/sec: 1.42865 +INFO:tensorflow:step = 63601, loss = 0.134772, precision = 0.992188 (69.996 sec) +INFO:tensorflow:global_step/sec: 1.41144 +INFO:tensorflow:step = 63701, loss = 0.14837, precision = 0.992188 (70.850 sec) +Saved checkpoint after 163 epoch(s) to data/resnet20/checkpoints/00163... +INFO:tensorflow:global_step/sec: 1.40312 +INFO:tensorflow:step = 63801, loss = 0.140243, precision = 0.992188 (71.270 sec) +INFO:tensorflow:global_step/sec: 1.42851 +INFO:tensorflow:step = 63901, loss = 0.150747, precision = 1.0 (70.003 sec) +INFO:tensorflow:global_step/sec: 1.44748 +INFO:tensorflow:step = 64001, loss = 0.157233, precision = 1.0 (69.085 sec) +INFO:tensorflow:global_step/sec: 1.43725 +INFO:tensorflow:step = 64101, loss = 0.149267, precision = 1.0 (69.577 sec) +Saved checkpoint after 164 epoch(s) to data/resnet20/checkpoints/00164... +INFO:tensorflow:global_step/sec: 1.40414 +INFO:tensorflow:step = 64201, loss = 0.135762, precision = 1.0 (71.218 sec) +INFO:tensorflow:global_step/sec: 1.41343 +INFO:tensorflow:step = 64301, loss = 0.135647, precision = 1.0 (70.750 sec) +INFO:tensorflow:global_step/sec: 1.39856 +INFO:tensorflow:step = 64401, loss = 0.14675, precision = 0.992188 (71.502 sec) +INFO:tensorflow:global_step/sec: 1.3966 +INFO:tensorflow:step = 64501, loss = 0.151823, precision = 0.984375 (71.603 sec) +Saved checkpoint after 165 epoch(s) to data/resnet20/checkpoints/00165... +INFO:tensorflow:global_step/sec: 1.32821 +INFO:tensorflow:step = 64601, loss = 0.144089, precision = 0.992188 (75.289 sec) +INFO:tensorflow:global_step/sec: 1.41406 +INFO:tensorflow:step = 64701, loss = 0.139752, precision = 1.0 (70.718 sec) +INFO:tensorflow:global_step/sec: 1.42192 +INFO:tensorflow:step = 64801, loss = 0.173842, precision = 0.984375 (70.327 sec) +INFO:tensorflow:global_step/sec: 1.43362 +INFO:tensorflow:step = 64901, loss = 0.141282, precision = 1.0 (69.754 sec) +Saved checkpoint after 166 epoch(s) to data/resnet20/checkpoints/00166... +INFO:tensorflow:global_step/sec: 1.44749 +INFO:tensorflow:step = 65001, loss = 0.164778, precision = 0.984375 (69.085 sec) +INFO:tensorflow:global_step/sec: 1.43475 +INFO:tensorflow:step = 65101, loss = 0.156399, precision = 0.992188 (69.698 sec) +INFO:tensorflow:global_step/sec: 1.37664 +INFO:tensorflow:step = 65201, loss = 0.138509, precision = 1.0 (72.640 sec) +Saved checkpoint after 167 epoch(s) to data/resnet20/checkpoints/00167... +INFO:tensorflow:global_step/sec: 1.40518 +INFO:tensorflow:step = 65301, loss = 0.156978, precision = 0.984375 (71.165 sec) +INFO:tensorflow:global_step/sec: 1.42104 +INFO:tensorflow:step = 65401, loss = 0.171634, precision = 0.984375 (70.371 sec) +INFO:tensorflow:global_step/sec: 1.40439 +INFO:tensorflow:step = 65501, loss = 0.13107, precision = 1.0 (71.206 sec) +INFO:tensorflow:global_step/sec: 1.4269 +INFO:tensorflow:step = 65601, loss = 0.191536, precision = 0.976562 (70.082 sec) +Saved checkpoint after 168 epoch(s) to data/resnet20/checkpoints/00168... +INFO:tensorflow:global_step/sec: 1.42246 +INFO:tensorflow:step = 65701, loss = 0.163747, precision = 0.976562 (70.301 sec) +INFO:tensorflow:global_step/sec: 1.39857 +INFO:tensorflow:step = 65801, loss = 0.144858, precision = 1.0 (71.501 sec) +INFO:tensorflow:global_step/sec: 1.3686 +INFO:tensorflow:step = 65901, loss = 0.14529, precision = 0.992188 (73.068 sec) +INFO:tensorflow:global_step/sec: 1.38149 +INFO:tensorflow:step = 66001, loss = 0.182847, precision = 0.976562 (72.385 sec) +Saved checkpoint after 169 epoch(s) to data/resnet20/checkpoints/00169... +INFO:tensorflow:global_step/sec: 1.36133 +INFO:tensorflow:step = 66101, loss = 0.138401, precision = 0.992188 (73.458 sec) +INFO:tensorflow:global_step/sec: 1.38306 +INFO:tensorflow:step = 66201, loss = 0.156002, precision = 0.984375 (72.303 sec) +INFO:tensorflow:global_step/sec: 1.398 +INFO:tensorflow:step = 66301, loss = 0.161193, precision = 0.976562 (71.531 sec) +INFO:tensorflow:global_step/sec: 1.35675 +INFO:tensorflow:step = 66401, loss = 0.147542, precision = 0.992188 (73.706 sec) +Saved checkpoint after 170 epoch(s) to data/resnet20/checkpoints/00170... +INFO:tensorflow:global_step/sec: 1.361 +INFO:tensorflow:step = 66501, loss = 0.140798, precision = 1.0 (73.475 sec) +INFO:tensorflow:global_step/sec: 1.41997 +INFO:tensorflow:step = 66601, loss = 0.145893, precision = 1.0 (70.424 sec) +INFO:tensorflow:global_step/sec: 1.39565 +INFO:tensorflow:step = 66701, loss = 0.139005, precision = 1.0 (71.651 sec) +INFO:tensorflow:global_step/sec: 1.41069 +INFO:tensorflow:step = 66801, loss = 0.145771, precision = 1.0 (70.887 sec) +Saved checkpoint after 171 epoch(s) to data/resnet20/checkpoints/00171... +INFO:tensorflow:global_step/sec: 1.40097 +INFO:tensorflow:step = 66901, loss = 0.150868, precision = 0.992188 (71.380 sec) +INFO:tensorflow:global_step/sec: 1.37465 +INFO:tensorflow:step = 67001, loss = 0.148011, precision = 0.992188 (72.745 sec) +INFO:tensorflow:global_step/sec: 1.3686 +INFO:tensorflow:step = 67101, loss = 0.146831, precision = 0.992188 (73.067 sec) +INFO:tensorflow:global_step/sec: 1.36565 +INFO:tensorflow:step = 67201, loss = 0.149221, precision = 0.992188 (73.225 sec) +Saved checkpoint after 172 epoch(s) to data/resnet20/checkpoints/00172... +INFO:tensorflow:global_step/sec: 1.40109 +INFO:tensorflow:step = 67301, loss = 0.137713, precision = 1.0 (71.373 sec) +INFO:tensorflow:global_step/sec: 1.38337 +INFO:tensorflow:step = 67401, loss = 0.133903, precision = 1.0 (72.287 sec) +INFO:tensorflow:global_step/sec: 1.30401 +INFO:tensorflow:step = 67501, loss = 0.143248, precision = 0.992188 (76.687 sec) +INFO:tensorflow:global_step/sec: 1.27044 +INFO:tensorflow:step = 67601, loss = 0.138825, precision = 1.0 (78.712 sec) +Saved checkpoint after 173 epoch(s) to data/resnet20/checkpoints/00173... +INFO:tensorflow:global_step/sec: 1.31386 +INFO:tensorflow:step = 67701, loss = 0.161405, precision = 0.984375 (76.112 sec) +INFO:tensorflow:global_step/sec: 1.33364 +INFO:tensorflow:step = 67801, loss = 0.143258, precision = 0.992188 (74.983 sec) +INFO:tensorflow:global_step/sec: 1.43036 +INFO:tensorflow:step = 67901, loss = 0.158052, precision = 0.984375 (69.913 sec) +INFO:tensorflow:global_step/sec: 1.36682 +INFO:tensorflow:step = 68001, loss = 0.143085, precision = 0.992188 (73.162 sec) +Saved checkpoint after 174 epoch(s) to data/resnet20/checkpoints/00174... +INFO:tensorflow:global_step/sec: 1.39524 +INFO:tensorflow:step = 68101, loss = 0.14396, precision = 1.0 (71.673 sec) +INFO:tensorflow:global_step/sec: 1.4099 +INFO:tensorflow:step = 68201, loss = 0.144224, precision = 0.984375 (70.927 sec) +INFO:tensorflow:global_step/sec: 1.42601 +INFO:tensorflow:step = 68301, loss = 0.139429, precision = 0.992188 (70.126 sec) +INFO:tensorflow:global_step/sec: 1.4233 +INFO:tensorflow:step = 68401, loss = 0.158292, precision = 0.984375 (70.259 sec) +Saved checkpoint after 175 epoch(s) to data/resnet20/checkpoints/00175... +INFO:tensorflow:global_step/sec: 1.42952 +INFO:tensorflow:step = 68501, loss = 0.147374, precision = 1.0 (69.954 sec) +INFO:tensorflow:global_step/sec: 1.43876 +INFO:tensorflow:step = 68601, loss = 0.13175, precision = 1.0 (69.504 sec) +INFO:tensorflow:global_step/sec: 1.45361 +INFO:tensorflow:step = 68701, loss = 0.160602, precision = 0.984375 (68.794 sec) +INFO:tensorflow:global_step/sec: 1.45183 +INFO:tensorflow:step = 68801, loss = 0.133471, precision = 0.992188 (68.878 sec) +Saved checkpoint after 176 epoch(s) to data/resnet20/checkpoints/00176... +INFO:tensorflow:global_step/sec: 1.45321 +INFO:tensorflow:step = 68901, loss = 0.13774, precision = 0.992188 (68.814 sec) +INFO:tensorflow:global_step/sec: 1.45934 +INFO:tensorflow:step = 69001, loss = 0.143611, precision = 0.992188 (68.524 sec) +INFO:tensorflow:global_step/sec: 1.43218 +INFO:tensorflow:step = 69101, loss = 0.137205, precision = 0.992188 (69.824 sec) +INFO:tensorflow:global_step/sec: 1.37443 +INFO:tensorflow:step = 69201, loss = 0.138539, precision = 0.992188 (72.757 sec) +Saved checkpoint after 177 epoch(s) to data/resnet20/checkpoints/00177... +INFO:tensorflow:global_step/sec: 1.39543 +INFO:tensorflow:step = 69301, loss = 0.148138, precision = 0.984375 (71.663 sec) +INFO:tensorflow:global_step/sec: 1.40344 +INFO:tensorflow:step = 69401, loss = 0.16281, precision = 0.976562 (71.253 sec) +INFO:tensorflow:global_step/sec: 1.37933 +INFO:tensorflow:step = 69501, loss = 0.136094, precision = 0.992188 (72.499 sec) +Saved checkpoint after 178 epoch(s) to data/resnet20/checkpoints/00178... +INFO:tensorflow:global_step/sec: 1.38236 +INFO:tensorflow:step = 69601, loss = 0.126335, precision = 1.0 (72.340 sec) +INFO:tensorflow:global_step/sec: 1.38645 +INFO:tensorflow:step = 69701, loss = 0.127896, precision = 1.0 (72.127 sec) +INFO:tensorflow:global_step/sec: 1.37446 +INFO:tensorflow:step = 69801, loss = 0.16279, precision = 0.984375 (72.756 sec) +INFO:tensorflow:global_step/sec: 1.34354 +INFO:tensorflow:step = 69901, loss = 0.126041, precision = 1.0 (74.430 sec) +Saved checkpoint after 179 epoch(s) to data/resnet20/checkpoints/00179... +INFO:tensorflow:global_step/sec: 1.35248 +INFO:tensorflow:step = 70001, loss = 0.133052, precision = 1.0 (73.938 sec) +INFO:tensorflow:global_step/sec: 1.37812 +INFO:tensorflow:step = 70101, loss = 0.136093, precision = 1.0 (72.563 sec) +INFO:tensorflow:global_step/sec: 1.37372 +INFO:tensorflow:step = 70201, loss = 0.152913, precision = 0.984375 (72.795 sec) +INFO:tensorflow:global_step/sec: 1.36618 +INFO:tensorflow:step = 70301, loss = 0.141387, precision = 0.992188 (73.197 sec) +Saved checkpoint after 180 epoch(s) to data/resnet20/checkpoints/00180... +INFO:tensorflow:global_step/sec: 1.3553 +INFO:tensorflow:step = 70401, loss = 0.140909, precision = 0.992188 (73.785 sec) +INFO:tensorflow:global_step/sec: 1.36336 +INFO:tensorflow:step = 70501, loss = 0.127538, precision = 1.0 (73.348 sec) +INFO:tensorflow:global_step/sec: 1.39546 +INFO:tensorflow:step = 70601, loss = 0.128074, precision = 1.0 (71.661 sec) +INFO:tensorflow:global_step/sec: 1.39208 +INFO:tensorflow:step = 70701, loss = 0.123106, precision = 1.0 (71.835 sec) +Saved checkpoint after 181 epoch(s) to data/resnet20/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet56_train.log b/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet56_train.log new file mode 100644 index 0000000..d3315ba --- /dev/null +++ b/tensorflow/CIFAR10/logs/16vCPUs_gc/resnet56_train.log @@ -0,0 +1,1821 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/850.99k params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_3/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_4/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_5/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_6/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_7/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_8/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_3/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_4/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_5/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_6/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_7/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_8/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_3/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_4/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_5/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_6/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_7/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_8/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/32.12b flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + +======================End of Report========================== +2017-07-31 22:24:18.687509: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 16 visible devices +2017-07-31 22:24:18.696724: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x42de850 executing computations on platform Host. Devices: +2017-07-31 22:24:18.696818: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +INFO:tensorflow:step = 1, loss = 3.31907, precision = 0.132812 +INFO:tensorflow:global_step/sec: 0.530769 +INFO:tensorflow:step = 101, loss = 2.71318, precision = 0.296875 (188.407 sec) +INFO:tensorflow:global_step/sec: 0.569373 +INFO:tensorflow:step = 201, loss = 2.61078, precision = 0.398438 (175.632 sec) +INFO:tensorflow:global_step/sec: 0.569269 +INFO:tensorflow:step = 301, loss = 2.37576, precision = 0.429688 (175.664 sec) +total_params: 850986 +Saved checkpoint after 1 epoch(s) to data/resnet56/checkpoints/00001... +INFO:tensorflow:global_step/sec: 0.564645 +INFO:tensorflow:step = 401, loss = 2.9322, precision = 0.296875 (177.102 sec) +INFO:tensorflow:global_step/sec: 0.574526 +INFO:tensorflow:step = 501, loss = 2.44202, precision = 0.453125 (174.056 sec) +INFO:tensorflow:global_step/sec: 0.576845 +INFO:tensorflow:step = 601, loss = 1.94834, precision = 0.59375 (173.357 sec) +INFO:tensorflow:global_step/sec: 0.570925 +INFO:tensorflow:step = 701, loss = 1.7663, precision = 0.617188 (175.154 sec) +Saved checkpoint after 2 epoch(s) to data/resnet56/checkpoints/00002... +INFO:tensorflow:global_step/sec: 0.567671 +INFO:tensorflow:step = 801, loss = 1.76382, precision = 0.59375 (176.158 sec) +INFO:tensorflow:global_step/sec: 0.574476 +INFO:tensorflow:step = 901, loss = 1.65698, precision = 0.664062 (174.072 sec) +INFO:tensorflow:global_step/sec: 0.573843 +INFO:tensorflow:step = 1001, loss = 1.54349, precision = 0.71875 (174.264 sec) +INFO:tensorflow:global_step/sec: 0.579561 +INFO:tensorflow:step = 1101, loss = 1.47647, precision = 0.6875 (172.544 sec) +Saved checkpoint after 3 epoch(s) to data/resnet56/checkpoints/00003... +INFO:tensorflow:global_step/sec: 0.58063 +INFO:tensorflow:step = 1201, loss = 1.42117, precision = 0.671875 (172.227 sec) +INFO:tensorflow:global_step/sec: 0.580233 +INFO:tensorflow:step = 1301, loss = 1.41527, precision = 0.671875 (172.345 sec) +INFO:tensorflow:global_step/sec: 0.582418 +INFO:tensorflow:step = 1401, loss = 1.22126, precision = 0.726562 (171.698 sec) +INFO:tensorflow:global_step/sec: 0.581333 +INFO:tensorflow:step = 1501, loss = 1.10693, precision = 0.828125 (172.019 sec) +Saved checkpoint after 4 epoch(s) to data/resnet56/checkpoints/00004... +INFO:tensorflow:global_step/sec: 0.583909 +INFO:tensorflow:step = 1601, loss = 1.29412, precision = 0.71875 (171.260 sec) +INFO:tensorflow:global_step/sec: 0.582987 +INFO:tensorflow:step = 1701, loss = 1.15287, precision = 0.742188 (171.530 sec) +INFO:tensorflow:global_step/sec: 0.579258 +INFO:tensorflow:step = 1801, loss = 1.03184, precision = 0.773438 (172.635 sec) +INFO:tensorflow:global_step/sec: 0.584488 +INFO:tensorflow:step = 1901, loss = 1.10054, precision = 0.710938 (171.090 sec) +Saved checkpoint after 5 epoch(s) to data/resnet56/checkpoints/00005... +INFO:tensorflow:global_step/sec: 0.573381 +INFO:tensorflow:step = 2001, loss = 1.06115, precision = 0.765625 (174.404 sec) +INFO:tensorflow:global_step/sec: 0.577812 +INFO:tensorflow:step = 2101, loss = 0.961449, precision = 0.773438 (173.066 sec) +INFO:tensorflow:global_step/sec: 0.573176 +INFO:tensorflow:step = 2201, loss = 1.01412, precision = 0.75 (174.466 sec) +INFO:tensorflow:global_step/sec: 0.566296 +INFO:tensorflow:step = 2301, loss = 0.858767, precision = 0.835938 (176.586 sec) +Saved checkpoint after 6 epoch(s) to data/resnet56/checkpoints/00006... +INFO:tensorflow:global_step/sec: 0.563515 +INFO:tensorflow:step = 2401, loss = 1.06753, precision = 0.742188 (177.458 sec) +INFO:tensorflow:global_step/sec: 0.579616 +INFO:tensorflow:step = 2501, loss = 0.89646, precision = 0.773438 (172.528 sec) +INFO:tensorflow:global_step/sec: 0.580427 +INFO:tensorflow:step = 2601, loss = 0.810582, precision = 0.828125 (172.287 sec) +INFO:tensorflow:global_step/sec: 0.580584 +INFO:tensorflow:step = 2701, loss = 0.86814, precision = 0.796875 (172.240 sec) +Saved checkpoint after 7 epoch(s) to data/resnet56/checkpoints/00007... +INFO:tensorflow:global_step/sec: 0.577683 +INFO:tensorflow:step = 2801, loss = 0.85567, precision = 0.796875 (173.105 sec) +INFO:tensorflow:global_step/sec: 0.579637 +INFO:tensorflow:step = 2901, loss = 0.777482, precision = 0.796875 (172.522 sec) +INFO:tensorflow:global_step/sec: 0.585109 +INFO:tensorflow:step = 3001, loss = 0.868359, precision = 0.78125 (170.908 sec) +INFO:tensorflow:global_step/sec: 0.579584 +INFO:tensorflow:step = 3101, loss = 0.757011, precision = 0.851562 (172.538 sec) +Saved checkpoint after 8 epoch(s) to data/resnet56/checkpoints/00008... +INFO:tensorflow:global_step/sec: 0.575722 +INFO:tensorflow:step = 3201, loss = 0.715379, precision = 0.851562 (173.695 sec) +INFO:tensorflow:global_step/sec: 0.58288 +INFO:tensorflow:step = 3301, loss = 0.815688, precision = 0.859375 (171.562 sec) +INFO:tensorflow:global_step/sec: 0.577009 +INFO:tensorflow:step = 3401, loss = 0.779924, precision = 0.84375 (173.308 sec) +INFO:tensorflow:global_step/sec: 0.57949 +INFO:tensorflow:step = 3501, loss = 0.827195, precision = 0.820312 (172.565 sec) +Saved checkpoint after 9 epoch(s) to data/resnet56/checkpoints/00009... +INFO:tensorflow:global_step/sec: 0.580746 +INFO:tensorflow:step = 3601, loss = 0.781871, precision = 0.828125 (172.192 sec) +INFO:tensorflow:global_step/sec: 0.585323 +INFO:tensorflow:step = 3701, loss = 0.911752, precision = 0.796875 (170.846 sec) +INFO:tensorflow:global_step/sec: 0.586362 +INFO:tensorflow:step = 3801, loss = 0.664689, precision = 0.84375 (170.543 sec) +INFO:tensorflow:global_step/sec: 0.582255 +INFO:tensorflow:step = 3901, loss = 0.822281, precision = 0.78125 (171.746 sec) +Saved checkpoint after 10 epoch(s) to data/resnet56/checkpoints/00010... +INFO:tensorflow:global_step/sec: 0.581337 +INFO:tensorflow:step = 4001, loss = 0.851345, precision = 0.828125 (172.017 sec) +INFO:tensorflow:global_step/sec: 0.57591 +INFO:tensorflow:step = 4101, loss = 0.718069, precision = 0.8125 (173.638 sec) +INFO:tensorflow:global_step/sec: 0.572929 +INFO:tensorflow:step = 4201, loss = 0.756897, precision = 0.8125 (174.542 sec) +Saved checkpoint after 11 epoch(s) to data/resnet56/checkpoints/00011... +INFO:tensorflow:global_step/sec: 0.572663 +INFO:tensorflow:step = 4301, loss = 0.74126, precision = 0.851562 (174.623 sec) +INFO:tensorflow:global_step/sec: 0.577554 +INFO:tensorflow:step = 4401, loss = 0.783009, precision = 0.84375 (173.144 sec) +INFO:tensorflow:global_step/sec: 0.573182 +INFO:tensorflow:step = 4501, loss = 0.734511, precision = 0.820312 (174.465 sec) +INFO:tensorflow:global_step/sec: 0.569088 +INFO:tensorflow:step = 4601, loss = 0.808334, precision = 0.828125 (175.720 sec) +Saved checkpoint after 12 epoch(s) to data/resnet56/checkpoints/00012... +INFO:tensorflow:global_step/sec: 0.571876 +INFO:tensorflow:step = 4701, loss = 0.743991, precision = 0.851562 (174.863 sec) +INFO:tensorflow:global_step/sec: 0.55902 +INFO:tensorflow:step = 4801, loss = 0.704985, precision = 0.851562 (178.884 sec) +INFO:tensorflow:global_step/sec: 0.556775 +INFO:tensorflow:step = 4901, loss = 0.766414, precision = 0.820312 (179.606 sec) +INFO:tensorflow:global_step/sec: 0.554601 +INFO:tensorflow:step = 5001, loss = 0.921725, precision = 0.75 (180.310 sec) +Saved checkpoint after 13 epoch(s) to data/resnet56/checkpoints/00013... +INFO:tensorflow:global_step/sec: 0.558071 +INFO:tensorflow:step = 5101, loss = 0.754501, precision = 0.851562 (179.189 sec) +INFO:tensorflow:global_step/sec: 0.564739 +INFO:tensorflow:step = 5201, loss = 0.78029, precision = 0.820312 (177.073 sec) +INFO:tensorflow:global_step/sec: 0.563343 +INFO:tensorflow:step = 5301, loss = 0.682077, precision = 0.875 (177.512 sec) +INFO:tensorflow:global_step/sec: 0.567024 +INFO:tensorflow:step = 5401, loss = 0.713833, precision = 0.875 (176.359 sec) +Saved checkpoint after 14 epoch(s) to data/resnet56/checkpoints/00014... +INFO:tensorflow:global_step/sec: 0.567358 +INFO:tensorflow:step = 5501, loss = 0.726426, precision = 0.835938 (176.256 sec) +INFO:tensorflow:global_step/sec: 0.562871 +INFO:tensorflow:step = 5601, loss = 0.778458, precision = 0.84375 (177.661 sec) +INFO:tensorflow:global_step/sec: 0.579453 +INFO:tensorflow:step = 5701, loss = 0.754298, precision = 0.84375 (172.576 sec) +INFO:tensorflow:global_step/sec: 0.56777 +INFO:tensorflow:step = 5801, loss = 0.653165, precision = 0.84375 (176.128 sec) +Saved checkpoint after 15 epoch(s) to data/resnet56/checkpoints/00015... +INFO:tensorflow:global_step/sec: 0.574822 +INFO:tensorflow:step = 5901, loss = 0.725361, precision = 0.8125 (173.967 sec) +INFO:tensorflow:global_step/sec: 0.570322 +INFO:tensorflow:step = 6001, loss = 0.592661, precision = 0.890625 (175.339 sec) +INFO:tensorflow:global_step/sec: 0.566928 +INFO:tensorflow:step = 6101, loss = 0.805628, precision = 0.804688 (176.389 sec) +INFO:tensorflow:global_step/sec: 0.55842 +INFO:tensorflow:step = 6201, loss = 0.91011, precision = 0.804688 (179.077 sec) +Saved checkpoint after 16 epoch(s) to data/resnet56/checkpoints/00016... +INFO:tensorflow:global_step/sec: 0.552825 +INFO:tensorflow:step = 6301, loss = 0.647994, precision = 0.875 (180.889 sec) +INFO:tensorflow:global_step/sec: 0.562713 +INFO:tensorflow:step = 6401, loss = 0.616861, precision = 0.882812 (177.710 sec) +INFO:tensorflow:global_step/sec: 0.564089 +INFO:tensorflow:step = 6501, loss = 0.647336, precision = 0.859375 (177.277 sec) +INFO:tensorflow:global_step/sec: 0.571242 +INFO:tensorflow:step = 6601, loss = 0.712285, precision = 0.859375 (175.057 sec) +Saved checkpoint after 17 epoch(s) to data/resnet56/checkpoints/00017... +INFO:tensorflow:global_step/sec: 0.556896 +INFO:tensorflow:step = 6701, loss = 0.607412, precision = 0.84375 (179.567 sec) +INFO:tensorflow:global_step/sec: 0.567835 +INFO:tensorflow:step = 6801, loss = 0.777177, precision = 0.820312 (176.107 sec) +INFO:tensorflow:global_step/sec: 0.571133 +INFO:tensorflow:step = 6901, loss = 0.743748, precision = 0.835938 (175.091 sec) +INFO:tensorflow:global_step/sec: 0.568453 +INFO:tensorflow:step = 7001, loss = 0.792057, precision = 0.820312 (175.916 sec) +Saved checkpoint after 18 epoch(s) to data/resnet56/checkpoints/00018... +INFO:tensorflow:global_step/sec: 0.562597 +INFO:tensorflow:step = 7101, loss = 0.679244, precision = 0.835938 (177.747 sec) +INFO:tensorflow:global_step/sec: 0.590538 +INFO:tensorflow:step = 7201, loss = 0.75687, precision = 0.84375 (169.337 sec) +INFO:tensorflow:global_step/sec: 0.569476 +INFO:tensorflow:step = 7301, loss = 0.66898, precision = 0.875 (175.600 sec) +INFO:tensorflow:global_step/sec: 0.516688 +INFO:tensorflow:step = 7401, loss = 0.749244, precision = 0.8125 (193.540 sec) +Saved checkpoint after 19 epoch(s) to data/resnet56/checkpoints/00019... +INFO:tensorflow:global_step/sec: 0.521164 +INFO:tensorflow:step = 7501, loss = 0.634275, precision = 0.867188 (191.878 sec) +INFO:tensorflow:global_step/sec: 0.534678 +INFO:tensorflow:step = 7601, loss = 0.700683, precision = 0.835938 (187.028 sec) +INFO:tensorflow:global_step/sec: 0.526474 +INFO:tensorflow:step = 7701, loss = 0.750688, precision = 0.828125 (189.943 sec) +INFO:tensorflow:global_step/sec: 0.531316 +INFO:tensorflow:step = 7801, loss = 0.71321, precision = 0.859375 (188.212 sec) +Saved checkpoint after 20 epoch(s) to data/resnet56/checkpoints/00020... +INFO:tensorflow:global_step/sec: 0.547269 +INFO:tensorflow:step = 7901, loss = 0.812995, precision = 0.804688 (182.726 sec) +INFO:tensorflow:global_step/sec: 0.596974 +INFO:tensorflow:step = 8001, loss = 0.701087, precision = 0.867188 (167.511 sec) +INFO:tensorflow:global_step/sec: 0.592745 +INFO:tensorflow:step = 8101, loss = 0.732023, precision = 0.828125 (168.707 sec) +INFO:tensorflow:global_step/sec: 0.594431 +INFO:tensorflow:step = 8201, loss = 0.599271, precision = 0.898438 (168.228 sec) +Saved checkpoint after 21 epoch(s) to data/resnet56/checkpoints/00021... +INFO:tensorflow:global_step/sec: 0.596658 +INFO:tensorflow:step = 8301, loss = 0.698006, precision = 0.84375 (167.600 sec) +INFO:tensorflow:global_step/sec: 0.60386 +INFO:tensorflow:step = 8401, loss = 0.694621, precision = 0.84375 (165.601 sec) +INFO:tensorflow:global_step/sec: 0.589607 +INFO:tensorflow:step = 8501, loss = 0.723915, precision = 0.820312 (169.605 sec) +INFO:tensorflow:global_step/sec: 0.603408 +INFO:tensorflow:step = 8601, loss = 0.71754, precision = 0.8125 (165.725 sec) +Saved checkpoint after 22 epoch(s) to data/resnet56/checkpoints/00022... +INFO:tensorflow:global_step/sec: 0.597882 +INFO:tensorflow:step = 8701, loss = 0.64584, precision = 0.882812 (167.257 sec) +INFO:tensorflow:global_step/sec: 0.610022 +INFO:tensorflow:step = 8801, loss = 0.665508, precision = 0.851562 (163.928 sec) +INFO:tensorflow:global_step/sec: 0.606698 +INFO:tensorflow:step = 8901, loss = 0.707651, precision = 0.867188 (164.827 sec) +Saved checkpoint after 23 epoch(s) to data/resnet56/checkpoints/00023... +INFO:tensorflow:global_step/sec: 0.606877 +INFO:tensorflow:step = 9001, loss = 0.661206, precision = 0.835938 (164.778 sec) +INFO:tensorflow:global_step/sec: 0.607377 +INFO:tensorflow:step = 9101, loss = 0.802053, precision = 0.8125 (164.642 sec) +INFO:tensorflow:global_step/sec: 0.605212 +INFO:tensorflow:step = 9201, loss = 0.645255, precision = 0.875 (165.231 sec) +INFO:tensorflow:global_step/sec: 0.604192 +INFO:tensorflow:step = 9301, loss = 0.535454, precision = 0.914062 (165.510 sec) +Saved checkpoint after 24 epoch(s) to data/resnet56/checkpoints/00024... +INFO:tensorflow:global_step/sec: 0.598841 +INFO:tensorflow:step = 9401, loss = 0.482337, precision = 0.953125 (166.989 sec) +INFO:tensorflow:global_step/sec: 0.607122 +INFO:tensorflow:step = 9501, loss = 0.530964, precision = 0.914062 (164.711 sec) +INFO:tensorflow:global_step/sec: 0.597343 +INFO:tensorflow:step = 9601, loss = 0.738684, precision = 0.84375 (167.408 sec) +INFO:tensorflow:global_step/sec: 0.605727 +INFO:tensorflow:step = 9701, loss = 0.692339, precision = 0.835938 (165.091 sec) +Saved checkpoint after 25 epoch(s) to data/resnet56/checkpoints/00025... +INFO:tensorflow:global_step/sec: 0.599707 +INFO:tensorflow:step = 9801, loss = 0.635321, precision = 0.882812 (166.748 sec) +INFO:tensorflow:global_step/sec: 0.602946 +INFO:tensorflow:step = 9901, loss = 0.599055, precision = 0.859375 (165.852 sec) +INFO:tensorflow:global_step/sec: 0.601405 +INFO:tensorflow:step = 10001, loss = 0.624404, precision = 0.90625 (166.277 sec) +INFO:tensorflow:global_step/sec: 0.602806 +INFO:tensorflow:step = 10101, loss = 0.731318, precision = 0.820312 (165.891 sec) +Saved checkpoint after 26 epoch(s) to data/resnet56/checkpoints/00026... +INFO:tensorflow:global_step/sec: 0.598351 +INFO:tensorflow:step = 10201, loss = 0.798569, precision = 0.835938 (167.126 sec) +INFO:tensorflow:global_step/sec: 0.600117 +INFO:tensorflow:step = 10301, loss = 0.660739, precision = 0.867188 (166.635 sec) +INFO:tensorflow:global_step/sec: 0.603834 +INFO:tensorflow:step = 10401, loss = 0.716807, precision = 0.875 (165.608 sec) +INFO:tensorflow:global_step/sec: 0.601241 +INFO:tensorflow:step = 10501, loss = 0.695957, precision = 0.859375 (166.323 sec) +Saved checkpoint after 27 epoch(s) to data/resnet56/checkpoints/00027... +INFO:tensorflow:global_step/sec: 0.599513 +INFO:tensorflow:step = 10601, loss = 0.590784, precision = 0.898438 (166.802 sec) +INFO:tensorflow:global_step/sec: 0.599935 +INFO:tensorflow:step = 10701, loss = 0.690745, precision = 0.851562 (166.685 sec) +INFO:tensorflow:global_step/sec: 0.599764 +INFO:tensorflow:step = 10801, loss = 0.590003, precision = 0.90625 (166.732 sec) +INFO:tensorflow:global_step/sec: 0.603224 +INFO:tensorflow:step = 10901, loss = 0.668631, precision = 0.859375 (165.776 sec) +Saved checkpoint after 28 epoch(s) to data/resnet56/checkpoints/00028... +INFO:tensorflow:global_step/sec: 0.600225 +INFO:tensorflow:step = 11001, loss = 0.556051, precision = 0.90625 (166.604 sec) +INFO:tensorflow:global_step/sec: 0.606559 +INFO:tensorflow:step = 11101, loss = 0.674825, precision = 0.851562 (164.864 sec) +INFO:tensorflow:global_step/sec: 0.597069 +INFO:tensorflow:step = 11201, loss = 0.685912, precision = 0.835938 (167.485 sec) +INFO:tensorflow:global_step/sec: 0.603135 +INFO:tensorflow:step = 11301, loss = 0.713503, precision = 0.867188 (165.800 sec) +Saved checkpoint after 29 epoch(s) to data/resnet56/checkpoints/00029... +INFO:tensorflow:global_step/sec: 0.596073 +INFO:tensorflow:step = 11401, loss = 0.869556, precision = 0.828125 (167.765 sec) +INFO:tensorflow:global_step/sec: 0.590444 +INFO:tensorflow:step = 11501, loss = 0.616344, precision = 0.898438 (169.364 sec) +INFO:tensorflow:global_step/sec: 0.589402 +INFO:tensorflow:step = 11601, loss = 0.693035, precision = 0.875 (169.663 sec) +INFO:tensorflow:global_step/sec: 0.590494 +INFO:tensorflow:step = 11701, loss = 0.610986, precision = 0.875 (169.350 sec) +Saved checkpoint after 30 epoch(s) to data/resnet56/checkpoints/00030... +INFO:tensorflow:global_step/sec: 0.590833 +INFO:tensorflow:step = 11801, loss = 0.682602, precision = 0.867188 (169.253 sec) +INFO:tensorflow:global_step/sec: 0.594086 +INFO:tensorflow:step = 11901, loss = 0.624719, precision = 0.875 (168.326 sec) +INFO:tensorflow:global_step/sec: 0.596499 +INFO:tensorflow:step = 12001, loss = 0.797066, precision = 0.820312 (167.645 sec) +INFO:tensorflow:global_step/sec: 0.597682 +INFO:tensorflow:step = 12101, loss = 0.680652, precision = 0.867188 (167.313 sec) +Saved checkpoint after 31 epoch(s) to data/resnet56/checkpoints/00031... +INFO:tensorflow:global_step/sec: 0.596652 +INFO:tensorflow:step = 12201, loss = 0.66228, precision = 0.898438 (167.602 sec) +INFO:tensorflow:global_step/sec: 0.602227 +INFO:tensorflow:step = 12301, loss = 0.688136, precision = 0.867188 (166.050 sec) +INFO:tensorflow:global_step/sec: 0.598537 +INFO:tensorflow:step = 12401, loss = 0.633119, precision = 0.859375 (167.074 sec) +INFO:tensorflow:global_step/sec: 0.595288 +INFO:tensorflow:step = 12501, loss = 0.618299, precision = 0.898438 (167.986 sec) +Saved checkpoint after 32 epoch(s) to data/resnet56/checkpoints/00032... +INFO:tensorflow:global_step/sec: 0.595724 +INFO:tensorflow:step = 12601, loss = 0.547064, precision = 0.914062 (167.863 sec) +INFO:tensorflow:global_step/sec: 0.601194 +INFO:tensorflow:step = 12701, loss = 0.740268, precision = 0.835938 (166.336 sec) +INFO:tensorflow:global_step/sec: 0.590211 +INFO:tensorflow:step = 12801, loss = 0.630302, precision = 0.882812 (169.431 sec) +INFO:tensorflow:global_step/sec: 0.587016 +INFO:tensorflow:step = 12901, loss = 0.529215, precision = 0.90625 (170.353 sec) +Saved checkpoint after 33 epoch(s) to data/resnet56/checkpoints/00033... +INFO:tensorflow:global_step/sec: 0.584733 +INFO:tensorflow:step = 13001, loss = 0.57376, precision = 0.921875 (171.018 sec) +INFO:tensorflow:global_step/sec: 0.588034 +INFO:tensorflow:step = 13101, loss = 0.696035, precision = 0.859375 (170.058 sec) +INFO:tensorflow:global_step/sec: 0.588681 +INFO:tensorflow:step = 13201, loss = 0.650243, precision = 0.875 (169.871 sec) +Saved checkpoint after 34 epoch(s) to data/resnet56/checkpoints/00034... +INFO:tensorflow:global_step/sec: 0.587478 +INFO:tensorflow:step = 13301, loss = 0.628242, precision = 0.898438 (170.219 sec) +INFO:tensorflow:global_step/sec: 0.58508 +INFO:tensorflow:step = 13401, loss = 0.616022, precision = 0.867188 (170.917 sec) +INFO:tensorflow:global_step/sec: 0.579153 +INFO:tensorflow:step = 13501, loss = 0.543864, precision = 0.90625 (172.666 sec) +INFO:tensorflow:global_step/sec: 0.583296 +INFO:tensorflow:step = 13601, loss = 0.630498, precision = 0.859375 (171.440 sec) +Saved checkpoint after 35 epoch(s) to data/resnet56/checkpoints/00035... +INFO:tensorflow:global_step/sec: 0.58555 +INFO:tensorflow:step = 13701, loss = 0.628297, precision = 0.875 (170.780 sec) +INFO:tensorflow:global_step/sec: 0.597948 +INFO:tensorflow:step = 13801, loss = 0.611591, precision = 0.90625 (167.238 sec) +INFO:tensorflow:global_step/sec: 0.593022 +INFO:tensorflow:step = 13901, loss = 0.62931, precision = 0.882812 (168.628 sec) +INFO:tensorflow:global_step/sec: 0.59267 +INFO:tensorflow:step = 14001, loss = 0.647498, precision = 0.859375 (168.728 sec) +Saved checkpoint after 36 epoch(s) to data/resnet56/checkpoints/00036... +INFO:tensorflow:global_step/sec: 0.589465 +INFO:tensorflow:step = 14101, loss = 0.650093, precision = 0.890625 (169.646 sec) +INFO:tensorflow:global_step/sec: 0.592715 +INFO:tensorflow:step = 14201, loss = 0.563055, precision = 0.90625 (168.715 sec) +INFO:tensorflow:global_step/sec: 0.585793 +INFO:tensorflow:step = 14301, loss = 0.670441, precision = 0.859375 (170.709 sec) +INFO:tensorflow:global_step/sec: 0.576244 +INFO:tensorflow:step = 14401, loss = 0.743712, precision = 0.875 (173.537 sec) +Saved checkpoint after 37 epoch(s) to data/resnet56/checkpoints/00037... +INFO:tensorflow:global_step/sec: 0.5816 +INFO:tensorflow:step = 14501, loss = 0.541101, precision = 0.914062 (171.939 sec) +INFO:tensorflow:global_step/sec: 0.578736 +INFO:tensorflow:step = 14601, loss = 0.724753, precision = 0.828125 (172.790 sec) +INFO:tensorflow:global_step/sec: 0.570824 +INFO:tensorflow:step = 14701, loss = 0.694067, precision = 0.867188 (175.186 sec) +INFO:tensorflow:global_step/sec: 0.578984 +INFO:tensorflow:step = 14801, loss = 0.723541, precision = 0.835938 (172.716 sec) +Saved checkpoint after 38 epoch(s) to data/resnet56/checkpoints/00038... +INFO:tensorflow:global_step/sec: 0.560278 +INFO:tensorflow:step = 14901, loss = 0.571327, precision = 0.921875 (178.482 sec) +INFO:tensorflow:global_step/sec: 0.562458 +INFO:tensorflow:step = 15001, loss = 0.556574, precision = 0.929688 (177.791 sec) +INFO:tensorflow:global_step/sec: 0.559224 +INFO:tensorflow:step = 15101, loss = 0.719658, precision = 0.859375 (178.819 sec) +INFO:tensorflow:global_step/sec: 0.56299 +INFO:tensorflow:step = 15201, loss = 0.703974, precision = 0.867188 (177.623 sec) +Saved checkpoint after 39 epoch(s) to data/resnet56/checkpoints/00039... +INFO:tensorflow:global_step/sec: 0.562806 +INFO:tensorflow:step = 15301, loss = 0.732881, precision = 0.835938 (177.681 sec) +INFO:tensorflow:global_step/sec: 0.56811 +INFO:tensorflow:step = 15401, loss = 0.703127, precision = 0.859375 (176.022 sec) +INFO:tensorflow:global_step/sec: 0.574895 +INFO:tensorflow:step = 15501, loss = 0.618802, precision = 0.898438 (173.945 sec) +INFO:tensorflow:global_step/sec: 0.584368 +INFO:tensorflow:step = 15601, loss = 0.610017, precision = 0.882812 (171.125 sec) +Saved checkpoint after 40 epoch(s) to data/resnet56/checkpoints/00040... +INFO:tensorflow:global_step/sec: 0.588634 +INFO:tensorflow:step = 15701, loss = 0.618214, precision = 0.882812 (169.885 sec) +INFO:tensorflow:global_step/sec: 0.591102 +INFO:tensorflow:step = 15801, loss = 0.624341, precision = 0.875 (169.176 sec) +INFO:tensorflow:global_step/sec: 0.594851 +INFO:tensorflow:step = 15901, loss = 0.795362, precision = 0.84375 (168.110 sec) +INFO:tensorflow:global_step/sec: 0.596564 +INFO:tensorflow:step = 16001, loss = 0.651764, precision = 0.859375 (167.626 sec) +Saved checkpoint after 41 epoch(s) to data/resnet56/checkpoints/00041... +INFO:tensorflow:global_step/sec: 0.597316 +INFO:tensorflow:step = 16101, loss = 0.54099, precision = 0.921875 (167.416 sec) +INFO:tensorflow:global_step/sec: 0.608692 +INFO:tensorflow:step = 16201, loss = 0.639187, precision = 0.875 (164.287 sec) +INFO:tensorflow:global_step/sec: 0.597538 +INFO:tensorflow:step = 16301, loss = 0.681028, precision = 0.867188 (167.353 sec) +INFO:tensorflow:global_step/sec: 0.596391 +INFO:tensorflow:step = 16401, loss = 0.608893, precision = 0.890625 (167.675 sec) +Saved checkpoint after 42 epoch(s) to data/resnet56/checkpoints/00042... +INFO:tensorflow:global_step/sec: 0.587458 +INFO:tensorflow:step = 16501, loss = 0.661088, precision = 0.859375 (170.225 sec) +INFO:tensorflow:global_step/sec: 0.586169 +INFO:tensorflow:step = 16601, loss = 0.797873, precision = 0.84375 (170.599 sec) +INFO:tensorflow:global_step/sec: 0.558228 +INFO:tensorflow:step = 16701, loss = 0.699536, precision = 0.867188 (179.138 sec) +INFO:tensorflow:global_step/sec: 0.575806 +INFO:tensorflow:step = 16801, loss = 0.580673, precision = 0.898438 (173.670 sec) +Saved checkpoint after 43 epoch(s) to data/resnet56/checkpoints/00043... +INFO:tensorflow:global_step/sec: 0.571094 +INFO:tensorflow:step = 16901, loss = 0.64699, precision = 0.859375 (175.103 sec) +INFO:tensorflow:global_step/sec: 0.578164 +INFO:tensorflow:step = 17001, loss = 0.592338, precision = 0.914062 (172.961 sec) +INFO:tensorflow:global_step/sec: 0.578896 +INFO:tensorflow:step = 17101, loss = 0.738545, precision = 0.859375 (172.743 sec) +INFO:tensorflow:global_step/sec: 0.582741 +INFO:tensorflow:step = 17201, loss = 0.60849, precision = 0.898438 (171.603 sec) +Saved checkpoint after 44 epoch(s) to data/resnet56/checkpoints/00044... +INFO:tensorflow:global_step/sec: 0.572004 +INFO:tensorflow:step = 17301, loss = 0.679876, precision = 0.875 (174.824 sec) +INFO:tensorflow:global_step/sec: 0.578579 +INFO:tensorflow:step = 17401, loss = 0.540346, precision = 0.914062 (172.837 sec) +INFO:tensorflow:global_step/sec: 0.583015 +INFO:tensorflow:step = 17501, loss = 0.670786, precision = 0.851562 (171.522 sec) +Saved checkpoint after 45 epoch(s) to data/resnet56/checkpoints/00045... +INFO:tensorflow:global_step/sec: 0.573801 +INFO:tensorflow:step = 17601, loss = 0.624177, precision = 0.882812 (174.276 sec) +INFO:tensorflow:global_step/sec: 0.57479 +INFO:tensorflow:step = 17701, loss = 0.677975, precision = 0.882812 (173.977 sec) +INFO:tensorflow:global_step/sec: 0.584428 +INFO:tensorflow:step = 17801, loss = 0.706573, precision = 0.859375 (171.107 sec) +INFO:tensorflow:global_step/sec: 0.583507 +INFO:tensorflow:step = 17901, loss = 0.612851, precision = 0.867188 (171.377 sec) +Saved checkpoint after 46 epoch(s) to data/resnet56/checkpoints/00046... +INFO:tensorflow:global_step/sec: 0.579492 +INFO:tensorflow:step = 18001, loss = 0.665954, precision = 0.859375 (172.565 sec) +INFO:tensorflow:global_step/sec: 0.58777 +INFO:tensorflow:step = 18101, loss = 0.803368, precision = 0.84375 (170.135 sec) +INFO:tensorflow:global_step/sec: 0.58213 +INFO:tensorflow:step = 18201, loss = 0.616897, precision = 0.875 (171.783 sec) +INFO:tensorflow:global_step/sec: 0.5962 +INFO:tensorflow:step = 18301, loss = 0.663905, precision = 0.851562 (167.729 sec) +Saved checkpoint after 47 epoch(s) to data/resnet56/checkpoints/00047... +INFO:tensorflow:global_step/sec: 0.590608 +INFO:tensorflow:step = 18401, loss = 0.640531, precision = 0.898438 (169.317 sec) +INFO:tensorflow:global_step/sec: 0.602916 +INFO:tensorflow:step = 18501, loss = 0.601887, precision = 0.890625 (165.860 sec) +INFO:tensorflow:global_step/sec: 0.601609 +INFO:tensorflow:step = 18601, loss = 0.858925, precision = 0.796875 (166.221 sec) +INFO:tensorflow:global_step/sec: 0.602695 +INFO:tensorflow:step = 18701, loss = 0.642856, precision = 0.875 (165.921 sec) +Saved checkpoint after 48 epoch(s) to data/resnet56/checkpoints/00048... +INFO:tensorflow:global_step/sec: 0.590442 +INFO:tensorflow:step = 18801, loss = 0.698009, precision = 0.867188 (169.365 sec) +INFO:tensorflow:global_step/sec: 0.592852 +INFO:tensorflow:step = 18901, loss = 0.687045, precision = 0.84375 (168.676 sec) +INFO:tensorflow:global_step/sec: 0.597818 +INFO:tensorflow:step = 19001, loss = 0.593744, precision = 0.898438 (167.275 sec) +INFO:tensorflow:global_step/sec: 0.601161 +INFO:tensorflow:step = 19101, loss = 0.571065, precision = 0.898438 (166.345 sec) +Saved checkpoint after 49 epoch(s) to data/resnet56/checkpoints/00049... +INFO:tensorflow:global_step/sec: 0.599524 +INFO:tensorflow:step = 19201, loss = 0.560972, precision = 0.921875 (166.799 sec) +INFO:tensorflow:global_step/sec: 0.59606 +INFO:tensorflow:step = 19301, loss = 0.582549, precision = 0.914062 (167.768 sec) +INFO:tensorflow:global_step/sec: 0.599987 +INFO:tensorflow:step = 19401, loss = 0.693915, precision = 0.851562 (166.670 sec) +INFO:tensorflow:global_step/sec: 0.597775 +INFO:tensorflow:step = 19501, loss = 0.554292, precision = 0.898438 (167.287 sec) +Saved checkpoint after 50 epoch(s) to data/resnet56/checkpoints/00050... +INFO:tensorflow:global_step/sec: 0.59639 +INFO:tensorflow:step = 19601, loss = 0.552641, precision = 0.921875 (167.676 sec) +INFO:tensorflow:global_step/sec: 0.611888 +INFO:tensorflow:step = 19701, loss = 0.595318, precision = 0.90625 (163.429 sec) +INFO:tensorflow:global_step/sec: 0.598842 +INFO:tensorflow:step = 19801, loss = 0.699215, precision = 0.882812 (166.989 sec) +INFO:tensorflow:global_step/sec: 0.603384 +INFO:tensorflow:step = 19901, loss = 0.618921, precision = 0.875 (165.732 sec) +Saved checkpoint after 51 epoch(s) to data/resnet56/checkpoints/00051... +INFO:tensorflow:global_step/sec: 0.60325 +INFO:tensorflow:step = 20001, loss = 0.54109, precision = 0.90625 (165.769 sec) +INFO:tensorflow:global_step/sec: 0.610375 +INFO:tensorflow:step = 20101, loss = 0.583032, precision = 0.890625 (163.834 sec) +INFO:tensorflow:global_step/sec: 0.608315 +INFO:tensorflow:step = 20201, loss = 0.703517, precision = 0.875 (164.388 sec) +INFO:tensorflow:global_step/sec: 0.606114 +INFO:tensorflow:step = 20301, loss = 0.625557, precision = 0.898438 (164.985 sec) +Saved checkpoint after 52 epoch(s) to data/resnet56/checkpoints/00052... +INFO:tensorflow:global_step/sec: 0.584808 +INFO:tensorflow:step = 20401, loss = 0.669792, precision = 0.875 (170.996 sec) +INFO:tensorflow:global_step/sec: 0.59298 +INFO:tensorflow:step = 20501, loss = 0.718094, precision = 0.851562 (168.640 sec) +INFO:tensorflow:global_step/sec: 0.594889 +INFO:tensorflow:step = 20601, loss = 0.766363, precision = 0.835938 (168.099 sec) +INFO:tensorflow:global_step/sec: 0.60166 +INFO:tensorflow:step = 20701, loss = 0.622692, precision = 0.890625 (166.207 sec) +Saved checkpoint after 53 epoch(s) to data/resnet56/checkpoints/00053... +INFO:tensorflow:global_step/sec: 0.594551 +INFO:tensorflow:step = 20801, loss = 0.665376, precision = 0.851562 (168.194 sec) +INFO:tensorflow:global_step/sec: 0.597586 +INFO:tensorflow:step = 20901, loss = 0.72687, precision = 0.882812 (167.340 sec) +INFO:tensorflow:global_step/sec: 0.599647 +INFO:tensorflow:step = 21001, loss = 0.752716, precision = 0.867188 (166.765 sec) +INFO:tensorflow:global_step/sec: 0.598021 +INFO:tensorflow:step = 21101, loss = 0.604954, precision = 0.867188 (167.218 sec) +Saved checkpoint after 54 epoch(s) to data/resnet56/checkpoints/00054... +INFO:tensorflow:global_step/sec: 0.598445 +INFO:tensorflow:step = 21201, loss = 0.547008, precision = 0.890625 (167.100 sec) +INFO:tensorflow:global_step/sec: 0.597511 +INFO:tensorflow:step = 21301, loss = 0.75172, precision = 0.828125 (167.361 sec) +INFO:tensorflow:global_step/sec: 0.601559 +INFO:tensorflow:step = 21401, loss = 0.538599, precision = 0.914062 (166.235 sec) +INFO:tensorflow:global_step/sec: 0.601034 +INFO:tensorflow:step = 21501, loss = 0.587225, precision = 0.898438 (166.380 sec) +Saved checkpoint after 55 epoch(s) to data/resnet56/checkpoints/00055... +INFO:tensorflow:global_step/sec: 0.603477 +INFO:tensorflow:step = 21601, loss = 0.648937, precision = 0.84375 (165.707 sec) +INFO:tensorflow:global_step/sec: 0.603373 +INFO:tensorflow:step = 21701, loss = 0.604408, precision = 0.914062 (165.735 sec) +INFO:tensorflow:global_step/sec: 0.59895 +INFO:tensorflow:step = 21801, loss = 0.666254, precision = 0.867188 (166.959 sec) +Saved checkpoint after 56 epoch(s) to data/resnet56/checkpoints/00056... +INFO:tensorflow:global_step/sec: 0.599016 +INFO:tensorflow:step = 21901, loss = 0.62882, precision = 0.867188 (166.941 sec) +INFO:tensorflow:global_step/sec: 0.589562 +INFO:tensorflow:step = 22001, loss = 0.62586, precision = 0.898438 (169.617 sec) +INFO:tensorflow:global_step/sec: 0.587572 +INFO:tensorflow:step = 22101, loss = 0.593058, precision = 0.882812 (170.192 sec) +INFO:tensorflow:global_step/sec: 0.584412 +INFO:tensorflow:step = 22201, loss = 0.712361, precision = 0.875 (171.112 sec) +Saved checkpoint after 57 epoch(s) to data/resnet56/checkpoints/00057... +INFO:tensorflow:global_step/sec: 0.577552 +INFO:tensorflow:step = 22301, loss = 0.697588, precision = 0.84375 (173.145 sec) +INFO:tensorflow:global_step/sec: 0.579292 +INFO:tensorflow:step = 22401, loss = 0.538158, precision = 0.90625 (172.624 sec) +INFO:tensorflow:global_step/sec: 0.588001 +INFO:tensorflow:step = 22501, loss = 0.603544, precision = 0.882812 (170.068 sec) +INFO:tensorflow:global_step/sec: 0.593227 +INFO:tensorflow:step = 22601, loss = 0.609458, precision = 0.914062 (168.569 sec) +Saved checkpoint after 58 epoch(s) to data/resnet56/checkpoints/00058... +INFO:tensorflow:global_step/sec: 0.587591 +INFO:tensorflow:step = 22701, loss = 0.559605, precision = 0.9375 (170.186 sec) +INFO:tensorflow:global_step/sec: 0.592312 +INFO:tensorflow:step = 22801, loss = 0.596878, precision = 0.914062 (168.830 sec) +INFO:tensorflow:global_step/sec: 0.589614 +INFO:tensorflow:step = 22901, loss = 0.693123, precision = 0.859375 (169.603 sec) +INFO:tensorflow:global_step/sec: 0.589181 +INFO:tensorflow:step = 23001, loss = 0.707997, precision = 0.875 (169.727 sec) +Saved checkpoint after 59 epoch(s) to data/resnet56/checkpoints/00059... +INFO:tensorflow:global_step/sec: 0.586131 +INFO:tensorflow:step = 23101, loss = 0.782239, precision = 0.835938 (170.610 sec) +INFO:tensorflow:global_step/sec: 0.589707 +INFO:tensorflow:step = 23201, loss = 0.604028, precision = 0.90625 (169.576 sec) +INFO:tensorflow:global_step/sec: 0.590512 +INFO:tensorflow:step = 23301, loss = 0.632509, precision = 0.90625 (169.344 sec) +INFO:tensorflow:global_step/sec: 0.584689 +INFO:tensorflow:step = 23401, loss = 0.546696, precision = 0.914062 (171.031 sec) +Saved checkpoint after 60 epoch(s) to data/resnet56/checkpoints/00060... +INFO:tensorflow:global_step/sec: 0.585958 +INFO:tensorflow:step = 23501, loss = 0.628059, precision = 0.898438 (170.661 sec) +INFO:tensorflow:global_step/sec: 0.584215 +INFO:tensorflow:step = 23601, loss = 0.750654, precision = 0.867188 (171.170 sec) +INFO:tensorflow:global_step/sec: 0.597663 +INFO:tensorflow:step = 23701, loss = 0.801872, precision = 0.851562 (167.318 sec) +INFO:tensorflow:global_step/sec: 0.586827 +INFO:tensorflow:step = 23801, loss = 0.658572, precision = 0.867188 (170.408 sec) +Saved checkpoint after 61 epoch(s) to data/resnet56/checkpoints/00061... +INFO:tensorflow:global_step/sec: 0.573167 +INFO:tensorflow:step = 23901, loss = 0.701035, precision = 0.867188 (174.469 sec) +INFO:tensorflow:global_step/sec: 0.587993 +INFO:tensorflow:step = 24001, loss = 0.584648, precision = 0.921875 (170.070 sec) +INFO:tensorflow:global_step/sec: 0.583339 +INFO:tensorflow:step = 24101, loss = 0.684654, precision = 0.859375 (171.427 sec) +INFO:tensorflow:global_step/sec: 0.585666 +INFO:tensorflow:step = 24201, loss = 0.539098, precision = 0.921875 (170.746 sec) +Saved checkpoint after 62 epoch(s) to data/resnet56/checkpoints/00062... +INFO:tensorflow:global_step/sec: 0.577965 +INFO:tensorflow:step = 24301, loss = 0.627715, precision = 0.890625 (173.021 sec) +INFO:tensorflow:global_step/sec: 0.586335 +INFO:tensorflow:step = 24401, loss = 0.666257, precision = 0.882812 (170.551 sec) +INFO:tensorflow:global_step/sec: 0.584279 +INFO:tensorflow:step = 24501, loss = 0.57933, precision = 0.914062 (171.151 sec) +INFO:tensorflow:global_step/sec: 0.590666 +INFO:tensorflow:step = 24601, loss = 0.52254, precision = 0.921875 (169.300 sec) +Saved checkpoint after 63 epoch(s) to data/resnet56/checkpoints/00063... +INFO:tensorflow:global_step/sec: 0.591415 +INFO:tensorflow:step = 24701, loss = 0.60671, precision = 0.914062 (169.086 sec) +INFO:tensorflow:global_step/sec: 0.596714 +INFO:tensorflow:step = 24801, loss = 0.610645, precision = 0.890625 (167.585 sec) +INFO:tensorflow:global_step/sec: 0.595996 +INFO:tensorflow:step = 24901, loss = 0.68577, precision = 0.867188 (167.786 sec) +INFO:tensorflow:global_step/sec: 0.593395 +INFO:tensorflow:step = 25001, loss = 0.717171, precision = 0.851562 (168.522 sec) +Saved checkpoint after 64 epoch(s) to data/resnet56/checkpoints/00064... +INFO:tensorflow:global_step/sec: 0.582929 +INFO:tensorflow:step = 25101, loss = 0.576707, precision = 0.90625 (171.548 sec) +INFO:tensorflow:global_step/sec: 0.586255 +INFO:tensorflow:step = 25201, loss = 0.669223, precision = 0.867188 (170.574 sec) +INFO:tensorflow:global_step/sec: 0.583058 +INFO:tensorflow:step = 25301, loss = 0.662254, precision = 0.867188 (171.509 sec) +INFO:tensorflow:global_step/sec: 0.577189 +INFO:tensorflow:step = 25401, loss = 0.618378, precision = 0.890625 (173.254 sec) +Saved checkpoint after 65 epoch(s) to data/resnet56/checkpoints/00065... +INFO:tensorflow:global_step/sec: 0.572679 +INFO:tensorflow:step = 25501, loss = 0.55132, precision = 0.898438 (174.618 sec) +INFO:tensorflow:global_step/sec: 0.570678 +INFO:tensorflow:step = 25601, loss = 0.590378, precision = 0.90625 (175.230 sec) +INFO:tensorflow:global_step/sec: 0.554395 +INFO:tensorflow:step = 25701, loss = 0.827887, precision = 0.835938 (180.377 sec) +INFO:tensorflow:global_step/sec: 0.506741 +INFO:tensorflow:step = 25801, loss = 0.725113, precision = 0.867188 (197.340 sec) +Saved checkpoint after 66 epoch(s) to data/resnet56/checkpoints/00066... +INFO:tensorflow:global_step/sec: 0.584016 +INFO:tensorflow:step = 25901, loss = 0.67117, precision = 0.875 (171.228 sec) +INFO:tensorflow:global_step/sec: 0.58458 +INFO:tensorflow:step = 26001, loss = 0.678611, precision = 0.875 (171.063 sec) +INFO:tensorflow:global_step/sec: 0.589363 +INFO:tensorflow:step = 26101, loss = 0.668273, precision = 0.867188 (169.675 sec) +Saved checkpoint after 67 epoch(s) to data/resnet56/checkpoints/00067... +INFO:tensorflow:global_step/sec: 0.582069 +INFO:tensorflow:step = 26201, loss = 0.598299, precision = 0.898438 (171.801 sec) +INFO:tensorflow:global_step/sec: 0.592312 +INFO:tensorflow:step = 26301, loss = 0.584773, precision = 0.929688 (168.830 sec) +INFO:tensorflow:global_step/sec: 0.59902 +INFO:tensorflow:step = 26401, loss = 0.755666, precision = 0.875 (166.939 sec) +INFO:tensorflow:global_step/sec: 0.594556 +INFO:tensorflow:step = 26501, loss = 0.628396, precision = 0.875 (168.193 sec) +Saved checkpoint after 68 epoch(s) to data/resnet56/checkpoints/00068... +INFO:tensorflow:global_step/sec: 0.591356 +INFO:tensorflow:step = 26601, loss = 0.696414, precision = 0.851562 (169.103 sec) +INFO:tensorflow:global_step/sec: 0.593455 +INFO:tensorflow:step = 26701, loss = 0.472707, precision = 0.96875 (168.505 sec) +INFO:tensorflow:global_step/sec: 0.599203 +INFO:tensorflow:step = 26801, loss = 0.675453, precision = 0.859375 (166.888 sec) +INFO:tensorflow:global_step/sec: 0.591457 +INFO:tensorflow:step = 26901, loss = 0.612018, precision = 0.90625 (169.074 sec) +Saved checkpoint after 69 epoch(s) to data/resnet56/checkpoints/00069... +INFO:tensorflow:global_step/sec: 0.590833 +INFO:tensorflow:step = 27001, loss = 0.570926, precision = 0.914062 (169.253 sec) +INFO:tensorflow:global_step/sec: 0.589403 +INFO:tensorflow:step = 27101, loss = 0.614928, precision = 0.859375 (169.663 sec) +INFO:tensorflow:global_step/sec: 0.595223 +INFO:tensorflow:step = 27201, loss = 0.620028, precision = 0.90625 (168.004 sec) +INFO:tensorflow:global_step/sec: 0.589998 +INFO:tensorflow:step = 27301, loss = 0.665557, precision = 0.90625 (169.492 sec) +Saved checkpoint after 70 epoch(s) to data/resnet56/checkpoints/00070... +INFO:tensorflow:global_step/sec: 0.583751 +INFO:tensorflow:step = 27401, loss = 0.61072, precision = 0.90625 (171.306 sec) +INFO:tensorflow:global_step/sec: 0.585677 +INFO:tensorflow:step = 27501, loss = 0.65009, precision = 0.882812 (170.743 sec) +INFO:tensorflow:global_step/sec: 0.578037 +INFO:tensorflow:step = 27601, loss = 0.548116, precision = 0.914062 (172.999 sec) +INFO:tensorflow:global_step/sec: 0.587746 +INFO:tensorflow:step = 27701, loss = 0.629321, precision = 0.890625 (170.141 sec) +Saved checkpoint after 71 epoch(s) to data/resnet56/checkpoints/00071... +INFO:tensorflow:global_step/sec: 0.57831 +INFO:tensorflow:step = 27801, loss = 0.663627, precision = 0.882812 (172.918 sec) +INFO:tensorflow:global_step/sec: 0.579359 +INFO:tensorflow:step = 27901, loss = 0.621712, precision = 0.882812 (172.605 sec) +INFO:tensorflow:global_step/sec: 0.574241 +INFO:tensorflow:step = 28001, loss = 0.533858, precision = 0.914062 (174.143 sec) +INFO:tensorflow:global_step/sec: 0.57343 +INFO:tensorflow:step = 28101, loss = 0.581661, precision = 0.921875 (174.389 sec) +Saved checkpoint after 72 epoch(s) to data/resnet56/checkpoints/00072... +INFO:tensorflow:global_step/sec: 0.580557 +INFO:tensorflow:step = 28201, loss = 0.611429, precision = 0.914062 (172.248 sec) +INFO:tensorflow:global_step/sec: 0.59137 +INFO:tensorflow:step = 28301, loss = 0.556513, precision = 0.914062 (169.099 sec) +INFO:tensorflow:global_step/sec: 0.576962 +INFO:tensorflow:step = 28401, loss = 0.527308, precision = 0.921875 (173.321 sec) +INFO:tensorflow:global_step/sec: 0.592806 +INFO:tensorflow:step = 28501, loss = 0.564569, precision = 0.921875 (168.689 sec) +Saved checkpoint after 73 epoch(s) to data/resnet56/checkpoints/00073... +INFO:tensorflow:global_step/sec: 0.592346 +INFO:tensorflow:step = 28601, loss = 0.647063, precision = 0.882812 (168.820 sec) +INFO:tensorflow:global_step/sec: 0.589739 +INFO:tensorflow:step = 28701, loss = 0.54337, precision = 0.914062 (169.566 sec) +INFO:tensorflow:global_step/sec: 0.592894 +INFO:tensorflow:step = 28801, loss = 0.654495, precision = 0.882812 (168.664 sec) +INFO:tensorflow:global_step/sec: 0.59194 +INFO:tensorflow:step = 28901, loss = 0.754956, precision = 0.867188 (168.936 sec) +Saved checkpoint after 74 epoch(s) to data/resnet56/checkpoints/00074... +INFO:tensorflow:global_step/sec: 0.58154 +INFO:tensorflow:step = 29001, loss = 0.717217, precision = 0.851562 (171.957 sec) +INFO:tensorflow:global_step/sec: 0.582044 +INFO:tensorflow:step = 29101, loss = 0.601396, precision = 0.882812 (171.808 sec) +INFO:tensorflow:global_step/sec: 0.580513 +INFO:tensorflow:step = 29201, loss = 0.704087, precision = 0.851562 (172.261 sec) +INFO:tensorflow:global_step/sec: 0.58037 +INFO:tensorflow:step = 29301, loss = 0.611242, precision = 0.898438 (172.304 sec) +Saved checkpoint after 75 epoch(s) to data/resnet56/checkpoints/00075... +INFO:tensorflow:global_step/sec: 0.5776 +INFO:tensorflow:step = 29401, loss = 0.683077, precision = 0.867188 (173.130 sec) +INFO:tensorflow:global_step/sec: 0.58745 +INFO:tensorflow:step = 29501, loss = 0.55648, precision = 0.921875 (170.228 sec) +INFO:tensorflow:global_step/sec: 0.585721 +INFO:tensorflow:step = 29601, loss = 0.64024, precision = 0.867188 (170.730 sec) +INFO:tensorflow:global_step/sec: 0.584114 +INFO:tensorflow:step = 29701, loss = 0.484238, precision = 0.9375 (171.199 sec) +Saved checkpoint after 76 epoch(s) to data/resnet56/checkpoints/00076... +INFO:tensorflow:global_step/sec: 0.591604 +INFO:tensorflow:step = 29801, loss = 0.596899, precision = 0.929688 (169.032 sec) +INFO:tensorflow:global_step/sec: 0.590409 +INFO:tensorflow:step = 29901, loss = 0.7265, precision = 0.867188 (169.374 sec) +INFO:tensorflow:global_step/sec: 0.594802 +INFO:tensorflow:step = 30001, loss = 0.561095, precision = 0.90625 (168.123 sec) +INFO:tensorflow:global_step/sec: 0.581024 +INFO:tensorflow:step = 30101, loss = 0.666671, precision = 0.890625 (172.110 sec) +Saved checkpoint after 77 epoch(s) to data/resnet56/checkpoints/00077... +INFO:tensorflow:global_step/sec: 0.589036 +INFO:tensorflow:step = 30201, loss = 0.786367, precision = 0.867188 (169.769 sec) +INFO:tensorflow:global_step/sec: 0.58664 +INFO:tensorflow:step = 30301, loss = 0.543961, precision = 0.929688 (170.462 sec) +INFO:tensorflow:global_step/sec: 0.592771 +INFO:tensorflow:step = 30401, loss = 0.572756, precision = 0.90625 (168.699 sec) +Saved checkpoint after 78 epoch(s) to data/resnet56/checkpoints/00078... +INFO:tensorflow:global_step/sec: 0.588527 +INFO:tensorflow:step = 30501, loss = 0.564636, precision = 0.929688 (169.916 sec) +INFO:tensorflow:global_step/sec: 0.585924 +INFO:tensorflow:step = 30601, loss = 0.620129, precision = 0.882812 (170.670 sec) +INFO:tensorflow:global_step/sec: 0.586092 +INFO:tensorflow:step = 30701, loss = 0.559417, precision = 0.914062 (170.622 sec) +INFO:tensorflow:global_step/sec: 0.585835 +INFO:tensorflow:step = 30801, loss = 0.625592, precision = 0.882812 (170.697 sec) +Saved checkpoint after 79 epoch(s) to data/resnet56/checkpoints/00079... +INFO:tensorflow:global_step/sec: 0.588873 +INFO:tensorflow:step = 30901, loss = 0.765835, precision = 0.851562 (169.816 sec) +INFO:tensorflow:global_step/sec: 0.58793 +INFO:tensorflow:step = 31001, loss = 0.674475, precision = 0.851562 (170.088 sec) +INFO:tensorflow:global_step/sec: 0.580094 +INFO:tensorflow:step = 31101, loss = 0.554832, precision = 0.882812 (172.386 sec) +INFO:tensorflow:global_step/sec: 0.58345 +INFO:tensorflow:step = 31201, loss = 0.662127, precision = 0.867188 (171.394 sec) +Saved checkpoint after 80 epoch(s) to data/resnet56/checkpoints/00080... +INFO:tensorflow:global_step/sec: 0.589006 +INFO:tensorflow:step = 31301, loss = 0.540322, precision = 0.90625 (169.778 sec) +INFO:tensorflow:global_step/sec: 0.596826 +INFO:tensorflow:step = 31401, loss = 0.553436, precision = 0.898438 (167.553 sec) +INFO:tensorflow:global_step/sec: 0.588464 +INFO:tensorflow:step = 31501, loss = 0.759205, precision = 0.828125 (169.934 sec) +INFO:tensorflow:global_step/sec: 0.590395 +INFO:tensorflow:step = 31601, loss = 0.678297, precision = 0.835938 (169.378 sec) +Saved checkpoint after 81 epoch(s) to data/resnet56/checkpoints/00081... +INFO:tensorflow:global_step/sec: 0.595639 +INFO:tensorflow:step = 31701, loss = 0.679281, precision = 0.84375 (167.887 sec) +INFO:tensorflow:global_step/sec: 0.591057 +INFO:tensorflow:step = 31801, loss = 0.704014, precision = 0.84375 (169.188 sec) +INFO:tensorflow:global_step/sec: 0.595575 +INFO:tensorflow:step = 31901, loss = 0.734908, precision = 0.867188 (167.905 sec) +INFO:tensorflow:global_step/sec: 0.594686 +INFO:tensorflow:step = 32001, loss = 0.673219, precision = 0.882812 (168.156 sec) +Saved checkpoint after 82 epoch(s) to data/resnet56/checkpoints/00082... +INFO:tensorflow:global_step/sec: 0.597116 +INFO:tensorflow:step = 32101, loss = 0.661542, precision = 0.875 (167.472 sec) +INFO:tensorflow:global_step/sec: 0.592132 +INFO:tensorflow:step = 32201, loss = 0.933072, precision = 0.796875 (168.881 sec) +INFO:tensorflow:global_step/sec: 0.584195 +INFO:tensorflow:step = 32301, loss = 0.733055, precision = 0.875 (171.176 sec) +INFO:tensorflow:global_step/sec: 0.576655 +INFO:tensorflow:step = 32401, loss = 0.60651, precision = 0.90625 (173.414 sec) +Saved checkpoint after 83 epoch(s) to data/resnet56/checkpoints/00083... +INFO:tensorflow:global_step/sec: 0.578129 +INFO:tensorflow:step = 32501, loss = 0.6608, precision = 0.875 (172.972 sec) +INFO:tensorflow:global_step/sec: 0.573392 +INFO:tensorflow:step = 32601, loss = 0.68421, precision = 0.828125 (174.400 sec) +INFO:tensorflow:global_step/sec: 0.580766 +INFO:tensorflow:step = 32701, loss = 0.632847, precision = 0.882812 (172.186 sec) +INFO:tensorflow:global_step/sec: 0.591764 +INFO:tensorflow:step = 32801, loss = 0.506486, precision = 0.921875 (168.986 sec) +Saved checkpoint after 84 epoch(s) to data/resnet56/checkpoints/00084... +INFO:tensorflow:global_step/sec: 0.568908 +INFO:tensorflow:step = 32901, loss = 0.532096, precision = 0.914062 (175.776 sec) +INFO:tensorflow:global_step/sec: 0.572133 +INFO:tensorflow:step = 33001, loss = 0.681341, precision = 0.851562 (174.784 sec) +INFO:tensorflow:global_step/sec: 0.56695 +INFO:tensorflow:step = 33101, loss = 0.650557, precision = 0.890625 (176.383 sec) +INFO:tensorflow:global_step/sec: 0.564063 +INFO:tensorflow:step = 33201, loss = 0.631156, precision = 0.890625 (177.285 sec) +Saved checkpoint after 85 epoch(s) to data/resnet56/checkpoints/00085... +INFO:tensorflow:global_step/sec: 0.581693 +INFO:tensorflow:step = 33301, loss = 0.568451, precision = 0.914062 (171.912 sec) +INFO:tensorflow:global_step/sec: 0.582302 +INFO:tensorflow:step = 33401, loss = 0.680601, precision = 0.867188 (171.732 sec) +INFO:tensorflow:global_step/sec: 0.586822 +INFO:tensorflow:step = 33501, loss = 0.638175, precision = 0.882812 (170.409 sec) +INFO:tensorflow:global_step/sec: 0.584313 +INFO:tensorflow:step = 33601, loss = 0.576034, precision = 0.898438 (171.141 sec) +Saved checkpoint after 86 epoch(s) to data/resnet56/checkpoints/00086... +INFO:tensorflow:global_step/sec: 0.577079 +INFO:tensorflow:step = 33701, loss = 0.665158, precision = 0.867188 (173.287 sec) +INFO:tensorflow:global_step/sec: 0.586154 +INFO:tensorflow:step = 33801, loss = 0.566992, precision = 0.929688 (170.604 sec) +INFO:tensorflow:global_step/sec: 0.577297 +INFO:tensorflow:step = 33901, loss = 0.567496, precision = 0.898438 (173.221 sec) +INFO:tensorflow:global_step/sec: 0.570536 +INFO:tensorflow:step = 34001, loss = 0.640981, precision = 0.898438 (175.274 sec) +Saved checkpoint after 87 epoch(s) to data/resnet56/checkpoints/00087... +INFO:tensorflow:global_step/sec: 0.55635 +INFO:tensorflow:step = 34101, loss = 0.708838, precision = 0.84375 (179.743 sec) +INFO:tensorflow:global_step/sec: 0.559663 +INFO:tensorflow:step = 34201, loss = 0.654042, precision = 0.882812 (178.679 sec) +INFO:tensorflow:global_step/sec: 0.563367 +INFO:tensorflow:step = 34301, loss = 0.76532, precision = 0.859375 (177.504 sec) +INFO:tensorflow:global_step/sec: 0.573014 +INFO:tensorflow:step = 34401, loss = 0.742773, precision = 0.859375 (174.516 sec) +Saved checkpoint after 88 epoch(s) to data/resnet56/checkpoints/00088... +INFO:tensorflow:global_step/sec: 0.568796 +INFO:tensorflow:step = 34501, loss = 0.611157, precision = 0.859375 (175.810 sec) +INFO:tensorflow:global_step/sec: 0.572211 +INFO:tensorflow:step = 34601, loss = 0.54334, precision = 0.921875 (174.761 sec) +INFO:tensorflow:global_step/sec: 0.575156 +INFO:tensorflow:step = 34701, loss = 0.634267, precision = 0.882812 (173.866 sec) +Saved checkpoint after 89 epoch(s) to data/resnet56/checkpoints/00089... +INFO:tensorflow:global_step/sec: 0.568932 +INFO:tensorflow:step = 34801, loss = 0.733873, precision = 0.820312 (175.768 sec) +INFO:tensorflow:global_step/sec: 0.575602 +INFO:tensorflow:step = 34901, loss = 0.739144, precision = 0.859375 (173.731 sec) +INFO:tensorflow:global_step/sec: 0.564576 +INFO:tensorflow:step = 35001, loss = 0.592213, precision = 0.90625 (177.124 sec) +INFO:tensorflow:global_step/sec: 0.56424 +INFO:tensorflow:step = 35101, loss = 0.505313, precision = 0.945312 (177.230 sec) +Saved checkpoint after 90 epoch(s) to data/resnet56/checkpoints/00090... +INFO:tensorflow:global_step/sec: 0.564404 +INFO:tensorflow:step = 35201, loss = 0.708558, precision = 0.882812 (177.178 sec) +INFO:tensorflow:global_step/sec: 0.562538 +INFO:tensorflow:step = 35301, loss = 0.584262, precision = 0.898438 (177.766 sec) +INFO:tensorflow:global_step/sec: 0.575381 +INFO:tensorflow:step = 35401, loss = 0.615185, precision = 0.867188 (173.798 sec) +INFO:tensorflow:global_step/sec: 0.564293 +INFO:tensorflow:step = 35501, loss = 0.514908, precision = 0.945312 (177.213 sec) +Saved checkpoint after 91 epoch(s) to data/resnet56/checkpoints/00091... +INFO:tensorflow:global_step/sec: 0.567152 +INFO:tensorflow:step = 35601, loss = 0.476121, precision = 0.945312 (176.320 sec) +INFO:tensorflow:global_step/sec: 0.569378 +INFO:tensorflow:step = 35701, loss = 0.510128, precision = 0.9375 (175.630 sec) +INFO:tensorflow:global_step/sec: 0.573168 +INFO:tensorflow:step = 35801, loss = 0.517168, precision = 0.929688 (174.469 sec) +INFO:tensorflow:global_step/sec: 0.572952 +INFO:tensorflow:step = 35901, loss = 0.379709, precision = 0.976562 (174.535 sec) +Saved checkpoint after 92 epoch(s) to data/resnet56/checkpoints/00092... +INFO:tensorflow:global_step/sec: 0.558944 +INFO:tensorflow:step = 36001, loss = 0.421816, precision = 0.953125 (178.909 sec) +INFO:tensorflow:global_step/sec: 0.567705 +INFO:tensorflow:step = 36101, loss = 0.414741, precision = 0.9375 (176.148 sec) +INFO:tensorflow:global_step/sec: 0.569731 +INFO:tensorflow:step = 36201, loss = 0.400598, precision = 0.96875 (175.521 sec) +INFO:tensorflow:global_step/sec: 0.571667 +INFO:tensorflow:step = 36301, loss = 0.441414, precision = 0.953125 (174.927 sec) +Saved checkpoint after 93 epoch(s) to data/resnet56/checkpoints/00093... +INFO:tensorflow:global_step/sec: 0.569504 +INFO:tensorflow:step = 36401, loss = 0.416873, precision = 0.945312 (175.592 sec) +INFO:tensorflow:global_step/sec: 0.542781 +INFO:tensorflow:step = 36501, loss = 0.442295, precision = 0.9375 (184.236 sec) +INFO:tensorflow:global_step/sec: 0.557355 +INFO:tensorflow:step = 36601, loss = 0.471042, precision = 0.953125 (179.419 sec) +INFO:tensorflow:global_step/sec: 0.559097 +INFO:tensorflow:step = 36701, loss = 0.403698, precision = 0.96875 (178.860 sec) +Saved checkpoint after 94 epoch(s) to data/resnet56/checkpoints/00094... +INFO:tensorflow:global_step/sec: 0.57045 +INFO:tensorflow:step = 36801, loss = 0.405776, precision = 0.953125 (175.300 sec) +INFO:tensorflow:global_step/sec: 0.576455 +INFO:tensorflow:step = 36901, loss = 0.383595, precision = 0.953125 (173.474 sec) +INFO:tensorflow:global_step/sec: 0.574101 +INFO:tensorflow:step = 37001, loss = 0.388507, precision = 0.96875 (174.186 sec) +INFO:tensorflow:global_step/sec: 0.573522 +INFO:tensorflow:step = 37101, loss = 0.400136, precision = 0.960938 (174.361 sec) +Saved checkpoint after 95 epoch(s) to data/resnet56/checkpoints/00095... +INFO:tensorflow:global_step/sec: 0.562522 +INFO:tensorflow:step = 37201, loss = 0.401013, precision = 0.945312 (177.771 sec) +INFO:tensorflow:global_step/sec: 0.563717 +INFO:tensorflow:step = 37301, loss = 0.35753, precision = 0.984375 (177.394 sec) +INFO:tensorflow:global_step/sec: 0.572277 +INFO:tensorflow:step = 37401, loss = 0.374704, precision = 0.96875 (174.741 sec) +INFO:tensorflow:global_step/sec: 0.563693 +INFO:tensorflow:step = 37501, loss = 0.38425, precision = 0.945312 (177.402 sec) +Saved checkpoint after 96 epoch(s) to data/resnet56/checkpoints/00096... +INFO:tensorflow:global_step/sec: 0.557017 +INFO:tensorflow:step = 37601, loss = 0.357272, precision = 0.96875 (179.528 sec) +INFO:tensorflow:global_step/sec: 0.560623 +INFO:tensorflow:step = 37701, loss = 0.31228, precision = 0.984375 (178.373 sec) +INFO:tensorflow:global_step/sec: 0.560726 +INFO:tensorflow:step = 37801, loss = 0.34326, precision = 0.976562 (178.340 sec) +INFO:tensorflow:global_step/sec: 0.567935 +INFO:tensorflow:step = 37901, loss = 0.36564, precision = 0.960938 (176.076 sec) +Saved checkpoint after 97 epoch(s) to data/resnet56/checkpoints/00097... +INFO:tensorflow:global_step/sec: 0.568834 +INFO:tensorflow:step = 38001, loss = 0.346709, precision = 0.960938 (175.798 sec) +INFO:tensorflow:global_step/sec: 0.56712 +INFO:tensorflow:step = 38101, loss = 0.33496, precision = 0.976562 (176.329 sec) +INFO:tensorflow:global_step/sec: 0.566694 +INFO:tensorflow:step = 38201, loss = 0.352021, precision = 0.984375 (176.462 sec) +INFO:tensorflow:global_step/sec: 0.568803 +INFO:tensorflow:step = 38301, loss = 0.373126, precision = 0.960938 (175.808 sec) +Saved checkpoint after 98 epoch(s) to data/resnet56/checkpoints/00098... +INFO:tensorflow:global_step/sec: 0.566918 +INFO:tensorflow:step = 38401, loss = 0.343342, precision = 0.953125 (176.393 sec) +INFO:tensorflow:global_step/sec: 0.56889 +INFO:tensorflow:step = 38501, loss = 0.309137, precision = 0.960938 (175.781 sec) +INFO:tensorflow:global_step/sec: 0.571112 +INFO:tensorflow:step = 38601, loss = 0.311538, precision = 0.96875 (175.097 sec) +INFO:tensorflow:global_step/sec: 0.568998 +INFO:tensorflow:step = 38701, loss = 0.285044, precision = 0.992188 (175.747 sec) +Saved checkpoint after 99 epoch(s) to data/resnet56/checkpoints/00099... +INFO:tensorflow:global_step/sec: 0.566807 +INFO:tensorflow:step = 38801, loss = 0.335581, precision = 0.960938 (176.427 sec) +INFO:tensorflow:global_step/sec: 0.571652 +INFO:tensorflow:step = 38901, loss = 0.353553, precision = 0.953125 (174.932 sec) +INFO:tensorflow:global_step/sec: 0.561582 +INFO:tensorflow:step = 39001, loss = 0.319167, precision = 0.953125 (178.068 sec) +Saved checkpoint after 100 epoch(s) to data/resnet56/checkpoints/00100... +INFO:tensorflow:global_step/sec: 0.545553 +INFO:tensorflow:step = 39101, loss = 0.263716, precision = 0.992188 (183.300 sec) +INFO:tensorflow:global_step/sec: 0.551286 +INFO:tensorflow:step = 39201, loss = 0.273399, precision = 0.976562 (181.394 sec) +INFO:tensorflow:global_step/sec: 0.559238 +INFO:tensorflow:step = 39301, loss = 0.301308, precision = 0.96875 (178.815 sec) +INFO:tensorflow:global_step/sec: 0.566043 +INFO:tensorflow:step = 39401, loss = 0.405967, precision = 0.929688 (176.665 sec) +Saved checkpoint after 101 epoch(s) to data/resnet56/checkpoints/00101... +INFO:tensorflow:global_step/sec: 0.564919 +INFO:tensorflow:step = 39501, loss = 0.278493, precision = 0.984375 (177.017 sec) +INFO:tensorflow:global_step/sec: 0.561498 +INFO:tensorflow:step = 39601, loss = 0.278441, precision = 0.976562 (178.095 sec) +INFO:tensorflow:global_step/sec: 0.567901 +INFO:tensorflow:step = 39701, loss = 0.276202, precision = 0.992188 (176.087 sec) +INFO:tensorflow:global_step/sec: 0.566036 +INFO:tensorflow:step = 39801, loss = 0.262369, precision = 0.992188 (176.667 sec) +Saved checkpoint after 102 epoch(s) to data/resnet56/checkpoints/00102... +INFO:tensorflow:global_step/sec: 0.561027 +INFO:tensorflow:step = 39901, loss = 0.322247, precision = 0.960938 (178.245 sec) +INFO:tensorflow:global_step/sec: 0.56788 +INFO:tensorflow:step = 40001, loss = 0.257054, precision = 0.984375 (176.093 sec) +INFO:tensorflow:global_step/sec: 0.57102 +INFO:tensorflow:step = 40101, loss = 0.279352, precision = 0.96875 (175.125 sec) +INFO:tensorflow:global_step/sec: 0.572568 +INFO:tensorflow:step = 40201, loss = 0.279738, precision = 0.96875 (174.652 sec) +Saved checkpoint after 103 epoch(s) to data/resnet56/checkpoints/00103... +INFO:tensorflow:global_step/sec: 0.567581 +INFO:tensorflow:step = 40301, loss = 0.258253, precision = 0.984375 (176.186 sec) +INFO:tensorflow:global_step/sec: 0.57044 +INFO:tensorflow:step = 40401, loss = 0.272103, precision = 0.984375 (175.303 sec) +INFO:tensorflow:global_step/sec: 0.56972 +INFO:tensorflow:step = 40501, loss = 0.276319, precision = 0.984375 (175.525 sec) +INFO:tensorflow:global_step/sec: 0.572897 +INFO:tensorflow:step = 40601, loss = 0.319767, precision = 0.96875 (174.552 sec) +Saved checkpoint after 104 epoch(s) to data/resnet56/checkpoints/00104... +INFO:tensorflow:global_step/sec: 0.567972 +INFO:tensorflow:step = 40701, loss = 0.271698, precision = 0.984375 (176.065 sec) +INFO:tensorflow:global_step/sec: 0.572153 +INFO:tensorflow:step = 40801, loss = 0.31935, precision = 0.960938 (174.778 sec) +INFO:tensorflow:global_step/sec: 0.571781 +INFO:tensorflow:step = 40901, loss = 0.291166, precision = 0.960938 (174.892 sec) +INFO:tensorflow:global_step/sec: 0.578286 +INFO:tensorflow:step = 41001, loss = 0.240726, precision = 0.992188 (172.925 sec) +Saved checkpoint after 105 epoch(s) to data/resnet56/checkpoints/00105... +INFO:tensorflow:global_step/sec: 0.56567 +INFO:tensorflow:step = 41101, loss = 0.35541, precision = 0.953125 (176.782 sec) +INFO:tensorflow:global_step/sec: 0.574793 +INFO:tensorflow:step = 41201, loss = 0.258917, precision = 0.976562 (173.975 sec) +INFO:tensorflow:global_step/sec: 0.569281 +INFO:tensorflow:step = 41301, loss = 0.262586, precision = 0.96875 (175.660 sec) +INFO:tensorflow:global_step/sec: 0.565329 +INFO:tensorflow:step = 41401, loss = 0.277859, precision = 0.976562 (176.888 sec) +Saved checkpoint after 106 epoch(s) to data/resnet56/checkpoints/00106... +INFO:tensorflow:global_step/sec: 0.568079 +INFO:tensorflow:step = 41501, loss = 0.255684, precision = 0.976562 (176.032 sec) +INFO:tensorflow:global_step/sec: 0.571252 +INFO:tensorflow:step = 41601, loss = 0.293098, precision = 0.953125 (175.054 sec) +INFO:tensorflow:global_step/sec: 0.571003 +INFO:tensorflow:step = 41701, loss = 0.29196, precision = 0.953125 (175.130 sec) +INFO:tensorflow:global_step/sec: 0.571673 +INFO:tensorflow:step = 41801, loss = 0.242733, precision = 0.976562 (174.925 sec) +Saved checkpoint after 107 epoch(s) to data/resnet56/checkpoints/00107... +INFO:tensorflow:global_step/sec: 0.571791 +INFO:tensorflow:step = 41901, loss = 0.233669, precision = 0.992188 (174.889 sec) +INFO:tensorflow:global_step/sec: 0.574251 +INFO:tensorflow:step = 42001, loss = 0.246032, precision = 0.984375 (174.140 sec) +INFO:tensorflow:global_step/sec: 0.571865 +INFO:tensorflow:step = 42101, loss = 0.247626, precision = 0.984375 (174.866 sec) +INFO:tensorflow:global_step/sec: 0.574816 +INFO:tensorflow:step = 42201, loss = 0.293953, precision = 0.960938 (173.969 sec) +Saved checkpoint after 108 epoch(s) to data/resnet56/checkpoints/00108... +INFO:tensorflow:global_step/sec: 0.569071 +INFO:tensorflow:step = 42301, loss = 0.21979, precision = 0.992188 (175.725 sec) +INFO:tensorflow:global_step/sec: 0.57027 +INFO:tensorflow:step = 42401, loss = 0.265493, precision = 0.96875 (175.355 sec) +INFO:tensorflow:global_step/sec: 0.578258 +INFO:tensorflow:step = 42501, loss = 0.264722, precision = 0.976562 (172.933 sec) +INFO:tensorflow:global_step/sec: 0.57473 +INFO:tensorflow:step = 42601, loss = 0.273082, precision = 0.976562 (173.995 sec) +Saved checkpoint after 109 epoch(s) to data/resnet56/checkpoints/00109... +INFO:tensorflow:global_step/sec: 0.581417 +INFO:tensorflow:step = 42701, loss = 0.239096, precision = 0.976562 (171.994 sec) +INFO:tensorflow:global_step/sec: 0.575213 +INFO:tensorflow:step = 42801, loss = 0.213736, precision = 0.992188 (173.848 sec) +INFO:tensorflow:global_step/sec: 0.57132 +INFO:tensorflow:step = 42901, loss = 0.217209, precision = 0.984375 (175.033 sec) +INFO:tensorflow:global_step/sec: 0.580499 +INFO:tensorflow:step = 43001, loss = 0.234122, precision = 0.984375 (172.265 sec) +Saved checkpoint after 110 epoch(s) to data/resnet56/checkpoints/00110... +INFO:tensorflow:global_step/sec: 0.564774 +INFO:tensorflow:step = 43101, loss = 0.239555, precision = 0.976562 (177.062 sec) +INFO:tensorflow:global_step/sec: 0.57 +INFO:tensorflow:step = 43201, loss = 0.240754, precision = 0.976562 (175.439 sec) +INFO:tensorflow:global_step/sec: 0.573837 +INFO:tensorflow:step = 43301, loss = 0.292822, precision = 0.960938 (174.265 sec) +Saved checkpoint after 111 epoch(s) to data/resnet56/checkpoints/00111... +INFO:tensorflow:global_step/sec: 0.587549 +INFO:tensorflow:step = 43401, loss = 0.231461, precision = 0.992188 (170.199 sec) +INFO:tensorflow:global_step/sec: 0.587271 +INFO:tensorflow:step = 43501, loss = 0.247365, precision = 0.976562 (170.279 sec) +INFO:tensorflow:global_step/sec: 0.590415 +INFO:tensorflow:step = 43601, loss = 0.258429, precision = 0.976562 (169.372 sec) +INFO:tensorflow:global_step/sec: 0.57192 +INFO:tensorflow:step = 43701, loss = 0.222886, precision = 0.96875 (174.850 sec) +Saved checkpoint after 112 epoch(s) to data/resnet56/checkpoints/00112... +INFO:tensorflow:global_step/sec: 0.582627 +INFO:tensorflow:step = 43801, loss = 0.239699, precision = 0.960938 (171.636 sec) +INFO:tensorflow:global_step/sec: 0.587825 +INFO:tensorflow:step = 43901, loss = 0.254171, precision = 0.984375 (170.118 sec) +INFO:tensorflow:global_step/sec: 0.583549 +INFO:tensorflow:step = 44001, loss = 0.220368, precision = 0.984375 (171.365 sec) +INFO:tensorflow:global_step/sec: 0.594538 +INFO:tensorflow:step = 44101, loss = 0.212371, precision = 0.992188 (168.198 sec) +Saved checkpoint after 113 epoch(s) to data/resnet56/checkpoints/00113... +INFO:tensorflow:global_step/sec: 0.582653 +INFO:tensorflow:step = 44201, loss = 0.251742, precision = 0.976562 (171.629 sec) +INFO:tensorflow:global_step/sec: 0.571009 +INFO:tensorflow:step = 44301, loss = 0.204114, precision = 0.992188 (175.128 sec) +INFO:tensorflow:global_step/sec: 0.574692 +INFO:tensorflow:step = 44401, loss = 0.250436, precision = 0.976562 (174.006 sec) +INFO:tensorflow:global_step/sec: 0.573529 +INFO:tensorflow:step = 44501, loss = 0.202712, precision = 0.992188 (174.359 sec) +Saved checkpoint after 114 epoch(s) to data/resnet56/checkpoints/00114... +INFO:tensorflow:global_step/sec: 0.581823 +INFO:tensorflow:step = 44601, loss = 0.23802, precision = 0.976562 (171.874 sec) +INFO:tensorflow:global_step/sec: 0.588967 +INFO:tensorflow:step = 44701, loss = 0.213028, precision = 0.976562 (169.789 sec) +INFO:tensorflow:global_step/sec: 0.596578 +INFO:tensorflow:step = 44801, loss = 0.266734, precision = 0.953125 (167.623 sec) +INFO:tensorflow:global_step/sec: 0.57063 +INFO:tensorflow:step = 44901, loss = 0.267874, precision = 0.960938 (175.245 sec) +Saved checkpoint after 115 epoch(s) to data/resnet56/checkpoints/00115... +INFO:tensorflow:global_step/sec: 0.572096 +INFO:tensorflow:step = 45001, loss = 0.2702, precision = 0.960938 (174.796 sec) +INFO:tensorflow:global_step/sec: 0.586535 +INFO:tensorflow:step = 45101, loss = 0.241201, precision = 0.984375 (170.493 sec) +INFO:tensorflow:global_step/sec: 0.588578 +INFO:tensorflow:step = 45201, loss = 0.191495, precision = 0.992188 (169.901 sec) +INFO:tensorflow:global_step/sec: 0.575764 +INFO:tensorflow:step = 45301, loss = 0.185488, precision = 0.992188 (173.682 sec) +Saved checkpoint after 116 epoch(s) to data/resnet56/checkpoints/00116... +INFO:tensorflow:global_step/sec: 0.57391 +INFO:tensorflow:step = 45401, loss = 0.216516, precision = 0.984375 (174.243 sec) +INFO:tensorflow:global_step/sec: 0.580897 +INFO:tensorflow:step = 45501, loss = 0.215947, precision = 0.992188 (172.147 sec) +INFO:tensorflow:global_step/sec: 0.582874 +INFO:tensorflow:step = 45601, loss = 0.260555, precision = 0.984375 (171.564 sec) +INFO:tensorflow:global_step/sec: 0.582382 +INFO:tensorflow:step = 45701, loss = 0.190573, precision = 0.984375 (171.709 sec) +Saved checkpoint after 117 epoch(s) to data/resnet56/checkpoints/00117... +INFO:tensorflow:global_step/sec: 0.582234 +INFO:tensorflow:step = 45801, loss = 0.206574, precision = 0.984375 (171.753 sec) +INFO:tensorflow:global_step/sec: 0.580124 +INFO:tensorflow:step = 45901, loss = 0.18443, precision = 0.992188 (172.377 sec) +INFO:tensorflow:global_step/sec: 0.581684 +INFO:tensorflow:step = 46001, loss = 0.272526, precision = 0.96875 (171.915 sec) +INFO:tensorflow:global_step/sec: 0.581574 +INFO:tensorflow:step = 46101, loss = 0.192142, precision = 0.984375 (171.947 sec) +Saved checkpoint after 118 epoch(s) to data/resnet56/checkpoints/00118... +INFO:tensorflow:global_step/sec: 0.58626 +INFO:tensorflow:step = 46201, loss = 0.181529, precision = 0.992188 (170.573 sec) +INFO:tensorflow:global_step/sec: 0.585427 +INFO:tensorflow:step = 46301, loss = 0.219964, precision = 0.984375 (170.815 sec) +INFO:tensorflow:global_step/sec: 0.588814 +INFO:tensorflow:step = 46401, loss = 0.208549, precision = 0.984375 (169.833 sec) +INFO:tensorflow:global_step/sec: 0.589067 +INFO:tensorflow:step = 46501, loss = 0.206409, precision = 0.984375 (169.760 sec) +Saved checkpoint after 119 epoch(s) to data/resnet56/checkpoints/00119... +INFO:tensorflow:global_step/sec: 0.57646 +INFO:tensorflow:step = 46601, loss = 0.237661, precision = 0.960938 (173.473 sec) +INFO:tensorflow:global_step/sec: 0.577302 +INFO:tensorflow:step = 46701, loss = 0.242484, precision = 0.960938 (173.220 sec) +INFO:tensorflow:global_step/sec: 0.579794 +INFO:tensorflow:step = 46801, loss = 0.21267, precision = 0.984375 (172.475 sec) +INFO:tensorflow:global_step/sec: 0.59119 +INFO:tensorflow:step = 46901, loss = 0.257385, precision = 0.960938 (169.150 sec) +Saved checkpoint after 120 epoch(s) to data/resnet56/checkpoints/00120... +INFO:tensorflow:global_step/sec: 0.585302 +INFO:tensorflow:step = 47001, loss = 0.175895, precision = 1.0 (170.852 sec) +INFO:tensorflow:global_step/sec: 0.579798 +INFO:tensorflow:step = 47101, loss = 0.243752, precision = 0.960938 (172.474 sec) +INFO:tensorflow:global_step/sec: 0.581224 +INFO:tensorflow:step = 47201, loss = 0.188069, precision = 0.992188 (172.051 sec) +INFO:tensorflow:global_step/sec: 0.584556 +INFO:tensorflow:step = 47301, loss = 0.196209, precision = 0.984375 (171.070 sec) +Saved checkpoint after 121 epoch(s) to data/resnet56/checkpoints/00121... +INFO:tensorflow:global_step/sec: 0.580752 +INFO:tensorflow:step = 47401, loss = 0.194763, precision = 0.992188 (172.191 sec) +INFO:tensorflow:global_step/sec: 0.591138 +INFO:tensorflow:step = 47501, loss = 0.25794, precision = 0.953125 (169.165 sec) +INFO:tensorflow:global_step/sec: 0.591874 +INFO:tensorflow:step = 47601, loss = 0.237763, precision = 0.945312 (168.955 sec) +INFO:tensorflow:global_step/sec: 0.587821 +INFO:tensorflow:step = 47701, loss = 0.202191, precision = 0.96875 (170.120 sec) +Saved checkpoint after 122 epoch(s) to data/resnet56/checkpoints/00122... +INFO:tensorflow:global_step/sec: 0.590759 +INFO:tensorflow:step = 47801, loss = 0.228542, precision = 0.96875 (169.274 sec) +INFO:tensorflow:global_step/sec: 0.587732 +INFO:tensorflow:step = 47901, loss = 0.202337, precision = 0.96875 (170.146 sec) +INFO:tensorflow:global_step/sec: 0.583002 +INFO:tensorflow:step = 48001, loss = 0.191713, precision = 0.992188 (171.526 sec) +Saved checkpoint after 123 epoch(s) to data/resnet56/checkpoints/00123... +INFO:tensorflow:global_step/sec: 0.588755 +INFO:tensorflow:step = 48101, loss = 0.184696, precision = 0.984375 (169.850 sec) +INFO:tensorflow:global_step/sec: 0.581195 +INFO:tensorflow:step = 48201, loss = 0.189334, precision = 0.976562 (172.059 sec) +INFO:tensorflow:global_step/sec: 0.588389 +INFO:tensorflow:step = 48301, loss = 0.242735, precision = 0.953125 (169.956 sec) +INFO:tensorflow:global_step/sec: 0.585905 +INFO:tensorflow:step = 48401, loss = 0.260462, precision = 0.96875 (170.676 sec) +Saved checkpoint after 124 epoch(s) to data/resnet56/checkpoints/00124... +INFO:tensorflow:global_step/sec: 0.578712 +INFO:tensorflow:step = 48501, loss = 0.243193, precision = 0.960938 (172.798 sec) +INFO:tensorflow:global_step/sec: 0.591034 +INFO:tensorflow:step = 48601, loss = 0.192729, precision = 0.984375 (169.195 sec) +INFO:tensorflow:global_step/sec: 0.582866 +INFO:tensorflow:step = 48701, loss = 0.211006, precision = 0.976562 (171.566 sec) +INFO:tensorflow:global_step/sec: 0.588679 +INFO:tensorflow:step = 48801, loss = 0.206234, precision = 0.984375 (169.872 sec) +Saved checkpoint after 125 epoch(s) to data/resnet56/checkpoints/00125... +INFO:tensorflow:global_step/sec: 0.574832 +INFO:tensorflow:step = 48901, loss = 0.201195, precision = 0.96875 (173.964 sec) +INFO:tensorflow:global_step/sec: 0.581786 +INFO:tensorflow:step = 49001, loss = 0.191467, precision = 0.984375 (171.884 sec) +INFO:tensorflow:global_step/sec: 0.586151 +INFO:tensorflow:step = 49101, loss = 0.204432, precision = 0.976562 (170.605 sec) +INFO:tensorflow:global_step/sec: 0.580852 +INFO:tensorflow:step = 49201, loss = 0.221556, precision = 0.960938 (172.161 sec) +Saved checkpoint after 126 epoch(s) to data/resnet56/checkpoints/00126... +INFO:tensorflow:global_step/sec: 0.583126 +INFO:tensorflow:step = 49301, loss = 0.222962, precision = 0.960938 (171.490 sec) +INFO:tensorflow:global_step/sec: 0.585991 +INFO:tensorflow:step = 49401, loss = 0.200298, precision = 0.984375 (170.651 sec) +INFO:tensorflow:global_step/sec: 0.588865 +INFO:tensorflow:step = 49501, loss = 0.219637, precision = 0.96875 (169.818 sec) +INFO:tensorflow:global_step/sec: 0.591113 +INFO:tensorflow:step = 49601, loss = 0.173043, precision = 0.992188 (169.172 sec) +Saved checkpoint after 127 epoch(s) to data/resnet56/checkpoints/00127... +INFO:tensorflow:global_step/sec: 0.587376 +INFO:tensorflow:step = 49701, loss = 0.181859, precision = 0.984375 (170.249 sec) +INFO:tensorflow:global_step/sec: 0.589503 +INFO:tensorflow:step = 49801, loss = 0.265884, precision = 0.960938 (169.634 sec) +INFO:tensorflow:global_step/sec: 0.586738 +INFO:tensorflow:step = 49901, loss = 0.188659, precision = 0.984375 (170.434 sec) +INFO:tensorflow:global_step/sec: 0.587893 +INFO:tensorflow:step = 50001, loss = 0.192078, precision = 0.984375 (170.099 sec) +Saved checkpoint after 128 epoch(s) to data/resnet56/checkpoints/00128... +INFO:tensorflow:global_step/sec: 0.585677 +INFO:tensorflow:step = 50101, loss = 0.175365, precision = 0.984375 (170.743 sec) +INFO:tensorflow:global_step/sec: 0.583874 +INFO:tensorflow:step = 50201, loss = 0.258648, precision = 0.953125 (171.270 sec) +INFO:tensorflow:global_step/sec: 0.582921 +INFO:tensorflow:step = 50301, loss = 0.185095, precision = 0.992188 (171.550 sec) +INFO:tensorflow:global_step/sec: 0.592305 +INFO:tensorflow:step = 50401, loss = 0.184663, precision = 0.984375 (168.832 sec) +Saved checkpoint after 129 epoch(s) to data/resnet56/checkpoints/00129... +INFO:tensorflow:global_step/sec: 0.581381 +INFO:tensorflow:step = 50501, loss = 0.159441, precision = 0.992188 (172.004 sec) +INFO:tensorflow:global_step/sec: 0.592867 +INFO:tensorflow:step = 50601, loss = 0.192726, precision = 0.984375 (168.672 sec) +INFO:tensorflow:global_step/sec: 0.593185 +INFO:tensorflow:step = 50701, loss = 0.169674, precision = 0.984375 (168.581 sec) +INFO:tensorflow:global_step/sec: 0.588017 +INFO:tensorflow:step = 50801, loss = 0.245099, precision = 0.953125 (170.063 sec) +Saved checkpoint after 130 epoch(s) to data/resnet56/checkpoints/00130... +INFO:tensorflow:global_step/sec: 0.584455 +INFO:tensorflow:step = 50901, loss = 0.264156, precision = 0.960938 (171.100 sec) +INFO:tensorflow:global_step/sec: 0.583028 +INFO:tensorflow:step = 51001, loss = 0.230278, precision = 0.96875 (171.518 sec) +INFO:tensorflow:global_step/sec: 0.582601 +INFO:tensorflow:step = 51101, loss = 0.230415, precision = 0.976562 (171.644 sec) +INFO:tensorflow:global_step/sec: 0.58435 +INFO:tensorflow:step = 51201, loss = 0.196827, precision = 0.984375 (171.130 sec) +Saved checkpoint after 131 epoch(s) to data/resnet56/checkpoints/00131... +INFO:tensorflow:global_step/sec: 0.584037 +INFO:tensorflow:step = 51301, loss = 0.242953, precision = 0.960938 (171.222 sec) +INFO:tensorflow:global_step/sec: 0.584513 +INFO:tensorflow:step = 51401, loss = 0.213419, precision = 0.984375 (171.083 sec) +INFO:tensorflow:global_step/sec: 0.580771 +INFO:tensorflow:step = 51501, loss = 0.215599, precision = 0.984375 (172.185 sec) +INFO:tensorflow:global_step/sec: 0.585397 +INFO:tensorflow:step = 51601, loss = 0.200292, precision = 0.984375 (170.824 sec) +Saved checkpoint after 132 epoch(s) to data/resnet56/checkpoints/00132... +INFO:tensorflow:global_step/sec: 0.585407 +INFO:tensorflow:step = 51701, loss = 0.220088, precision = 0.960938 (170.821 sec) +INFO:tensorflow:global_step/sec: 0.590619 +INFO:tensorflow:step = 51801, loss = 0.165528, precision = 0.992188 (169.314 sec) +INFO:tensorflow:global_step/sec: 0.583539 +INFO:tensorflow:step = 51901, loss = 0.181754, precision = 1.0 (171.368 sec) +INFO:tensorflow:global_step/sec: 0.590656 +INFO:tensorflow:step = 52001, loss = 0.191553, precision = 0.992188 (169.304 sec) +Saved checkpoint after 133 epoch(s) to data/resnet56/checkpoints/00133... +INFO:tensorflow:global_step/sec: 0.586018 +INFO:tensorflow:step = 52101, loss = 0.186677, precision = 0.976562 (170.643 sec) +INFO:tensorflow:global_step/sec: 0.590749 +INFO:tensorflow:step = 52201, loss = 0.165029, precision = 0.992188 (169.277 sec) +INFO:tensorflow:global_step/sec: 0.589651 +INFO:tensorflow:step = 52301, loss = 0.223036, precision = 0.960938 (169.592 sec) +Saved checkpoint after 134 epoch(s) to data/resnet56/checkpoints/00134... +INFO:tensorflow:global_step/sec: 0.578648 +INFO:tensorflow:step = 52401, loss = 0.218295, precision = 0.976562 (172.817 sec) +INFO:tensorflow:global_step/sec: 0.586757 +INFO:tensorflow:step = 52501, loss = 0.176365, precision = 0.984375 (170.428 sec) +INFO:tensorflow:global_step/sec: 0.576382 +INFO:tensorflow:step = 52601, loss = 0.166002, precision = 1.0 (173.496 sec) +INFO:tensorflow:global_step/sec: 0.586703 +INFO:tensorflow:step = 52701, loss = 0.178096, precision = 0.96875 (170.444 sec) +Saved checkpoint after 135 epoch(s) to data/resnet56/checkpoints/00135... +INFO:tensorflow:global_step/sec: 0.585325 +INFO:tensorflow:step = 52801, loss = 0.210649, precision = 0.976562 (170.845 sec) +INFO:tensorflow:global_step/sec: 0.589102 +INFO:tensorflow:step = 52901, loss = 0.216308, precision = 0.960938 (169.750 sec) +INFO:tensorflow:global_step/sec: 0.591965 +INFO:tensorflow:step = 53001, loss = 0.189755, precision = 0.992188 (168.929 sec) +INFO:tensorflow:global_step/sec: 0.587281 +INFO:tensorflow:step = 53101, loss = 0.199523, precision = 0.976562 (170.276 sec) +Saved checkpoint after 136 epoch(s) to data/resnet56/checkpoints/00136... +INFO:tensorflow:global_step/sec: 0.588184 +INFO:tensorflow:step = 53201, loss = 0.160935, precision = 1.0 (170.015 sec) +INFO:tensorflow:global_step/sec: 0.590845 +INFO:tensorflow:step = 53301, loss = 0.169499, precision = 0.984375 (169.249 sec) +INFO:tensorflow:global_step/sec: 0.59358 +INFO:tensorflow:step = 53401, loss = 0.168158, precision = 0.984375 (168.469 sec) +INFO:tensorflow:global_step/sec: 0.582969 +INFO:tensorflow:step = 53501, loss = 0.178729, precision = 0.992188 (171.536 sec) +Saved checkpoint after 137 epoch(s) to data/resnet56/checkpoints/00137... +INFO:tensorflow:global_step/sec: 0.586444 +INFO:tensorflow:step = 53601, loss = 0.172125, precision = 0.984375 (170.519 sec) +INFO:tensorflow:global_step/sec: 0.59251 +INFO:tensorflow:step = 53701, loss = 0.155056, precision = 0.992188 (168.773 sec) +INFO:tensorflow:global_step/sec: 0.598875 +INFO:tensorflow:step = 53801, loss = 0.145668, precision = 1.0 (166.980 sec) +INFO:tensorflow:global_step/sec: 0.597788 +INFO:tensorflow:step = 53901, loss = 0.147835, precision = 0.992188 (167.283 sec) +Saved checkpoint after 138 epoch(s) to data/resnet56/checkpoints/00138... +INFO:tensorflow:global_step/sec: 0.593637 +INFO:tensorflow:step = 54001, loss = 0.14933, precision = 1.0 (168.453 sec) +INFO:tensorflow:global_step/sec: 0.591219 +INFO:tensorflow:step = 54101, loss = 0.142893, precision = 1.0 (169.142 sec) +INFO:tensorflow:global_step/sec: 0.579639 +INFO:tensorflow:step = 54201, loss = 0.138457, precision = 1.0 (172.521 sec) +INFO:tensorflow:global_step/sec: 0.586906 +INFO:tensorflow:step = 54301, loss = 0.140837, precision = 1.0 (170.385 sec) +Saved checkpoint after 139 epoch(s) to data/resnet56/checkpoints/00139... +INFO:tensorflow:global_step/sec: 0.58532 +INFO:tensorflow:step = 54401, loss = 0.175631, precision = 0.984375 (170.847 sec) +INFO:tensorflow:global_step/sec: 0.589309 +INFO:tensorflow:step = 54501, loss = 0.150923, precision = 1.0 (169.690 sec) +INFO:tensorflow:global_step/sec: 0.591971 +INFO:tensorflow:step = 54601, loss = 0.143696, precision = 1.0 (168.927 sec) +INFO:tensorflow:global_step/sec: 0.591333 +INFO:tensorflow:step = 54701, loss = 0.144847, precision = 1.0 (169.109 sec) +Saved checkpoint after 140 epoch(s) to data/resnet56/checkpoints/00140... +INFO:tensorflow:global_step/sec: 0.592633 +INFO:tensorflow:step = 54801, loss = 0.14888, precision = 1.0 (168.739 sec) +INFO:tensorflow:global_step/sec: 0.588108 +INFO:tensorflow:step = 54901, loss = 0.147651, precision = 1.0 (170.037 sec) +INFO:tensorflow:global_step/sec: 0.59679 +INFO:tensorflow:step = 55001, loss = 0.146253, precision = 0.992188 (167.563 sec) +INFO:tensorflow:global_step/sec: 0.591192 +INFO:tensorflow:step = 55101, loss = 0.145153, precision = 1.0 (169.150 sec) +Saved checkpoint after 141 epoch(s) to data/resnet56/checkpoints/00141... +INFO:tensorflow:global_step/sec: 0.58542 +INFO:tensorflow:step = 55201, loss = 0.144712, precision = 1.0 (170.818 sec) +INFO:tensorflow:global_step/sec: 0.592922 +INFO:tensorflow:step = 55301, loss = 0.141329, precision = 1.0 (168.656 sec) +INFO:tensorflow:global_step/sec: 0.59244 +INFO:tensorflow:step = 55401, loss = 0.149071, precision = 0.992188 (168.793 sec) +INFO:tensorflow:global_step/sec: 0.589203 +INFO:tensorflow:step = 55501, loss = 0.14455, precision = 1.0 (169.721 sec) +Saved checkpoint after 142 epoch(s) to data/resnet56/checkpoints/00142... +INFO:tensorflow:global_step/sec: 0.58672 +INFO:tensorflow:step = 55601, loss = 0.142311, precision = 0.992188 (170.439 sec) +INFO:tensorflow:global_step/sec: 0.591844 +INFO:tensorflow:step = 55701, loss = 0.14693, precision = 0.992188 (168.963 sec) +INFO:tensorflow:global_step/sec: 0.585556 +INFO:tensorflow:step = 55801, loss = 0.142985, precision = 1.0 (170.778 sec) +INFO:tensorflow:global_step/sec: 0.590701 +INFO:tensorflow:step = 55901, loss = 0.146497, precision = 1.0 (169.290 sec) +Saved checkpoint after 143 epoch(s) to data/resnet56/checkpoints/00143... +INFO:tensorflow:global_step/sec: 0.581976 +INFO:tensorflow:step = 56001, loss = 0.139286, precision = 1.0 (171.829 sec) +INFO:tensorflow:global_step/sec: 0.575256 +INFO:tensorflow:step = 56101, loss = 0.139397, precision = 1.0 (173.836 sec) +INFO:tensorflow:global_step/sec: 0.58803 +INFO:tensorflow:step = 56201, loss = 0.15371, precision = 0.992188 (170.059 sec) +INFO:tensorflow:global_step/sec: 0.574005 +INFO:tensorflow:step = 56301, loss = 0.135965, precision = 1.0 (174.215 sec) +Saved checkpoint after 144 epoch(s) to data/resnet56/checkpoints/00144... +INFO:tensorflow:global_step/sec: 0.578682 +INFO:tensorflow:step = 56401, loss = 0.144661, precision = 1.0 (172.806 sec) +INFO:tensorflow:global_step/sec: 0.583636 +INFO:tensorflow:step = 56501, loss = 0.142776, precision = 1.0 (171.340 sec) +INFO:tensorflow:global_step/sec: 0.584259 +INFO:tensorflow:step = 56601, loss = 0.140047, precision = 1.0 (171.157 sec) +Saved checkpoint after 145 epoch(s) to data/resnet56/checkpoints/00145... +INFO:tensorflow:global_step/sec: 0.574852 +INFO:tensorflow:step = 56701, loss = 0.151603, precision = 0.992188 (173.958 sec) +INFO:tensorflow:global_step/sec: 0.573696 +INFO:tensorflow:step = 56801, loss = 0.152703, precision = 0.992188 (174.308 sec) +INFO:tensorflow:global_step/sec: 0.580912 +INFO:tensorflow:step = 56901, loss = 0.139237, precision = 1.0 (172.143 sec) +INFO:tensorflow:global_step/sec: 0.581002 +INFO:tensorflow:step = 57001, loss = 0.149827, precision = 0.992188 (172.116 sec) +Saved checkpoint after 146 epoch(s) to data/resnet56/checkpoints/00146... +INFO:tensorflow:global_step/sec: 0.572852 +INFO:tensorflow:step = 57101, loss = 0.141541, precision = 0.992188 (174.565 sec) +INFO:tensorflow:global_step/sec: 0.574663 +INFO:tensorflow:step = 57201, loss = 0.137334, precision = 1.0 (174.015 sec) +INFO:tensorflow:global_step/sec: 0.571949 +INFO:tensorflow:step = 57301, loss = 0.147868, precision = 0.992188 (174.841 sec) +INFO:tensorflow:global_step/sec: 0.580592 +INFO:tensorflow:step = 57401, loss = 0.14356, precision = 1.0 (172.238 sec) +Saved checkpoint after 147 epoch(s) to data/resnet56/checkpoints/00147... +INFO:tensorflow:global_step/sec: 0.577238 +INFO:tensorflow:step = 57501, loss = 0.14559, precision = 0.992188 (173.239 sec) +INFO:tensorflow:global_step/sec: 0.585458 +INFO:tensorflow:step = 57601, loss = 0.138077, precision = 1.0 (170.806 sec) +INFO:tensorflow:global_step/sec: 0.58161 +INFO:tensorflow:step = 57701, loss = 0.149236, precision = 0.992188 (171.936 sec) +INFO:tensorflow:global_step/sec: 0.587498 +INFO:tensorflow:step = 57801, loss = 0.136139, precision = 1.0 (170.213 sec) +Saved checkpoint after 148 epoch(s) to data/resnet56/checkpoints/00148... +INFO:tensorflow:global_step/sec: 0.574053 +INFO:tensorflow:step = 57901, loss = 0.135151, precision = 1.0 (174.200 sec) +INFO:tensorflow:global_step/sec: 0.582754 +INFO:tensorflow:step = 58001, loss = 0.148228, precision = 0.992188 (171.599 sec) +INFO:tensorflow:global_step/sec: 0.585416 +INFO:tensorflow:step = 58101, loss = 0.141207, precision = 1.0 (170.819 sec) +INFO:tensorflow:global_step/sec: 0.587248 +INFO:tensorflow:step = 58201, loss = 0.146379, precision = 0.992188 (170.286 sec) +Saved checkpoint after 149 epoch(s) to data/resnet56/checkpoints/00149... +INFO:tensorflow:global_step/sec: 0.580705 +INFO:tensorflow:step = 58301, loss = 0.140327, precision = 1.0 (172.205 sec) +INFO:tensorflow:global_step/sec: 0.580788 +INFO:tensorflow:step = 58401, loss = 0.140841, precision = 0.992188 (172.180 sec) +INFO:tensorflow:global_step/sec: 0.587907 +INFO:tensorflow:step = 58501, loss = 0.140107, precision = 1.0 (170.095 sec) +INFO:tensorflow:global_step/sec: 0.584277 +INFO:tensorflow:step = 58601, loss = 0.146685, precision = 1.0 (171.152 sec) +Saved checkpoint after 150 epoch(s) to data/resnet56/checkpoints/00150... +INFO:tensorflow:global_step/sec: 0.583047 +INFO:tensorflow:step = 58701, loss = 0.141078, precision = 1.0 (171.513 sec) +INFO:tensorflow:global_step/sec: 0.584214 +INFO:tensorflow:step = 58801, loss = 0.134361, precision = 1.0 (171.170 sec) +INFO:tensorflow:global_step/sec: 0.586775 +INFO:tensorflow:step = 58901, loss = 0.135409, precision = 1.0 (170.423 sec) +INFO:tensorflow:global_step/sec: 0.585902 +INFO:tensorflow:step = 59001, loss = 0.132206, precision = 1.0 (170.677 sec) +Saved checkpoint after 151 epoch(s) to data/resnet56/checkpoints/00151... +INFO:tensorflow:global_step/sec: 0.581346 +INFO:tensorflow:step = 59101, loss = 0.130536, precision = 1.0 (172.015 sec) +INFO:tensorflow:global_step/sec: 0.589133 +INFO:tensorflow:step = 59201, loss = 0.134107, precision = 1.0 (169.741 sec) +INFO:tensorflow:global_step/sec: 0.586001 +INFO:tensorflow:step = 59301, loss = 0.132804, precision = 1.0 (170.648 sec) +INFO:tensorflow:global_step/sec: 0.579276 +INFO:tensorflow:step = 59401, loss = 0.135404, precision = 1.0 (172.629 sec) +Saved checkpoint after 152 epoch(s) to data/resnet56/checkpoints/00152... +INFO:tensorflow:global_step/sec: 0.581374 +INFO:tensorflow:step = 59501, loss = 0.131767, precision = 1.0 (172.006 sec) +INFO:tensorflow:global_step/sec: 0.585445 +INFO:tensorflow:step = 59601, loss = 0.13854, precision = 1.0 (170.810 sec) +INFO:tensorflow:global_step/sec: 0.589751 +INFO:tensorflow:step = 59701, loss = 0.132071, precision = 1.0 (169.563 sec) +INFO:tensorflow:global_step/sec: 0.589053 +INFO:tensorflow:step = 59801, loss = 0.142666, precision = 0.992188 (169.764 sec) +Saved checkpoint after 153 epoch(s) to data/resnet56/checkpoints/00153... +INFO:tensorflow:global_step/sec: 0.587255 +INFO:tensorflow:step = 59901, loss = 0.133152, precision = 1.0 (170.284 sec) +INFO:tensorflow:global_step/sec: 0.589552 +INFO:tensorflow:step = 60001, loss = 0.134008, precision = 1.0 (169.620 sec) +INFO:tensorflow:global_step/sec: 0.588773 +INFO:tensorflow:step = 60101, loss = 0.131677, precision = 1.0 (169.845 sec) +INFO:tensorflow:global_step/sec: 0.590883 +INFO:tensorflow:step = 60201, loss = 0.141366, precision = 1.0 (169.238 sec) +Saved checkpoint after 154 epoch(s) to data/resnet56/checkpoints/00154... +INFO:tensorflow:global_step/sec: 0.585986 +INFO:tensorflow:step = 60301, loss = 0.135128, precision = 1.0 (170.653 sec) +INFO:tensorflow:global_step/sec: 0.587139 +INFO:tensorflow:step = 60401, loss = 0.137808, precision = 0.992188 (170.317 sec) +INFO:tensorflow:global_step/sec: 0.582491 +INFO:tensorflow:step = 60501, loss = 0.132143, precision = 1.0 (171.676 sec) +INFO:tensorflow:global_step/sec: 0.592442 +INFO:tensorflow:step = 60601, loss = 0.131418, precision = 1.0 (168.793 sec) +Saved checkpoint after 155 epoch(s) to data/resnet56/checkpoints/00155... +INFO:tensorflow:global_step/sec: 0.586212 +INFO:tensorflow:step = 60701, loss = 0.136583, precision = 1.0 (170.587 sec) +INFO:tensorflow:global_step/sec: 0.58894 +INFO:tensorflow:step = 60801, loss = 0.131756, precision = 1.0 (169.796 sec) +INFO:tensorflow:global_step/sec: 0.591938 +INFO:tensorflow:step = 60901, loss = 0.127777, precision = 1.0 (168.937 sec) +Saved checkpoint after 156 epoch(s) to data/resnet56/checkpoints/00156... +INFO:tensorflow:global_step/sec: 0.586824 +INFO:tensorflow:step = 61001, loss = 0.135133, precision = 1.0 (170.409 sec) +INFO:tensorflow:global_step/sec: 0.593225 +INFO:tensorflow:step = 61101, loss = 0.138741, precision = 0.992188 (168.570 sec) +INFO:tensorflow:global_step/sec: 0.593319 +INFO:tensorflow:step = 61201, loss = 0.143516, precision = 0.992188 (168.544 sec) +INFO:tensorflow:global_step/sec: 0.593144 +INFO:tensorflow:step = 61301, loss = 0.132166, precision = 1.0 (168.593 sec) +Saved checkpoint after 157 epoch(s) to data/resnet56/checkpoints/00157... +INFO:tensorflow:global_step/sec: 0.587464 +INFO:tensorflow:step = 61401, loss = 0.12987, precision = 1.0 (170.223 sec) +INFO:tensorflow:global_step/sec: 0.593409 +INFO:tensorflow:step = 61501, loss = 0.140301, precision = 0.992188 (168.518 sec) +INFO:tensorflow:global_step/sec: 0.593489 +INFO:tensorflow:step = 61601, loss = 0.129205, precision = 1.0 (168.495 sec) +INFO:tensorflow:global_step/sec: 0.591708 +INFO:tensorflow:step = 61701, loss = 0.130847, precision = 1.0 (169.002 sec) +Saved checkpoint after 158 epoch(s) to data/resnet56/checkpoints/00158... +INFO:tensorflow:global_step/sec: 0.593808 +INFO:tensorflow:step = 61801, loss = 0.127597, precision = 1.0 (168.405 sec) +INFO:tensorflow:global_step/sec: 0.590655 +INFO:tensorflow:step = 61901, loss = 0.129762, precision = 1.0 (169.304 sec) +INFO:tensorflow:global_step/sec: 0.596333 +INFO:tensorflow:step = 62001, loss = 0.130047, precision = 1.0 (167.691 sec) +INFO:tensorflow:global_step/sec: 0.590256 +INFO:tensorflow:step = 62101, loss = 0.129228, precision = 1.0 (169.418 sec) +Saved checkpoint after 159 epoch(s) to data/resnet56/checkpoints/00159... +INFO:tensorflow:global_step/sec: 0.598902 +INFO:tensorflow:step = 62201, loss = 0.129451, precision = 1.0 (166.972 sec) +INFO:tensorflow:global_step/sec: 0.596558 +INFO:tensorflow:step = 62301, loss = 0.12876, precision = 1.0 (167.628 sec) +INFO:tensorflow:global_step/sec: 0.595559 +INFO:tensorflow:step = 62401, loss = 0.129242, precision = 1.0 (167.909 sec) +INFO:tensorflow:global_step/sec: 0.598782 +INFO:tensorflow:step = 62501, loss = 0.135131, precision = 0.992188 (167.006 sec) +Saved checkpoint after 160 epoch(s) to data/resnet56/checkpoints/00160... +INFO:tensorflow:global_step/sec: 0.590809 +INFO:tensorflow:step = 62601, loss = 0.127143, precision = 1.0 (169.259 sec) +INFO:tensorflow:global_step/sec: 0.598318 +INFO:tensorflow:step = 62701, loss = 0.13865, precision = 0.992188 (167.135 sec) +INFO:tensorflow:global_step/sec: 0.598157 +INFO:tensorflow:step = 62801, loss = 0.129429, precision = 1.0 (167.180 sec) +INFO:tensorflow:global_step/sec: 0.594336 +INFO:tensorflow:step = 62901, loss = 0.128193, precision = 1.0 (168.255 sec) +Saved checkpoint after 161 epoch(s) to data/resnet56/checkpoints/00161... +INFO:tensorflow:global_step/sec: 0.591052 +INFO:tensorflow:step = 63001, loss = 0.12781, precision = 1.0 (169.190 sec) +INFO:tensorflow:global_step/sec: 0.598775 +INFO:tensorflow:step = 63101, loss = 0.125024, precision = 1.0 (167.008 sec) +INFO:tensorflow:global_step/sec: 0.594293 +INFO:tensorflow:step = 63201, loss = 0.131214, precision = 1.0 (168.267 sec) +INFO:tensorflow:global_step/sec: 0.594677 +INFO:tensorflow:step = 63301, loss = 0.146762, precision = 0.992188 (168.158 sec) +Saved checkpoint after 162 epoch(s) to data/resnet56/checkpoints/00162... +INFO:tensorflow:global_step/sec: 0.591007 +INFO:tensorflow:step = 63401, loss = 0.128512, precision = 1.0 (169.203 sec) +INFO:tensorflow:global_step/sec: 0.596731 +INFO:tensorflow:step = 63501, loss = 0.130439, precision = 1.0 (167.580 sec) +INFO:tensorflow:global_step/sec: 0.589815 +INFO:tensorflow:step = 63601, loss = 0.131738, precision = 1.0 (169.545 sec) +INFO:tensorflow:global_step/sec: 0.592663 +INFO:tensorflow:step = 63701, loss = 0.125655, precision = 1.0 (168.730 sec) +Saved checkpoint after 163 epoch(s) to data/resnet56/checkpoints/00163... +INFO:tensorflow:global_step/sec: 0.589421 +INFO:tensorflow:step = 63801, loss = 0.12567, precision = 1.0 (169.658 sec) +INFO:tensorflow:global_step/sec: 0.588753 +INFO:tensorflow:step = 63901, loss = 0.128471, precision = 1.0 (169.851 sec) +INFO:tensorflow:global_step/sec: 0.5854 +INFO:tensorflow:step = 64001, loss = 0.128883, precision = 1.0 (170.823 sec) +INFO:tensorflow:global_step/sec: 0.587167 +INFO:tensorflow:step = 64101, loss = 0.129105, precision = 1.0 (170.309 sec) +Saved checkpoint after 164 epoch(s) to data/resnet56/checkpoints/00164... +INFO:tensorflow:global_step/sec: 0.584344 +INFO:tensorflow:step = 64201, loss = 0.131156, precision = 1.0 (171.132 sec) +INFO:tensorflow:global_step/sec: 0.586159 +INFO:tensorflow:step = 64301, loss = 0.125541, precision = 1.0 (170.602 sec) +INFO:tensorflow:global_step/sec: 0.588715 +INFO:tensorflow:step = 64401, loss = 0.124871, precision = 1.0 (169.861 sec) +INFO:tensorflow:global_step/sec: 0.58885 +INFO:tensorflow:step = 64501, loss = 0.124115, precision = 1.0 (169.822 sec) +Saved checkpoint after 165 epoch(s) to data/resnet56/checkpoints/00165... +INFO:tensorflow:global_step/sec: 0.58101 +INFO:tensorflow:step = 64601, loss = 0.130588, precision = 1.0 (172.114 sec) +INFO:tensorflow:global_step/sec: 0.585469 +INFO:tensorflow:step = 64701, loss = 0.123443, precision = 1.0 (170.803 sec) +INFO:tensorflow:global_step/sec: 0.586446 +INFO:tensorflow:step = 64801, loss = 0.125783, precision = 1.0 (170.519 sec) +INFO:tensorflow:global_step/sec: 0.585423 +INFO:tensorflow:step = 64901, loss = 0.142395, precision = 0.992188 (170.817 sec) +Saved checkpoint after 166 epoch(s) to data/resnet56/checkpoints/00166... +INFO:tensorflow:global_step/sec: 0.586141 +INFO:tensorflow:step = 65001, loss = 0.122639, precision = 1.0 (170.608 sec) +INFO:tensorflow:global_step/sec: 0.584046 +INFO:tensorflow:step = 65101, loss = 0.124181, precision = 1.0 (171.219 sec) +INFO:tensorflow:global_step/sec: 0.587967 +INFO:tensorflow:step = 65201, loss = 0.12352, precision = 1.0 (170.078 sec) +Saved checkpoint after 167 epoch(s) to data/resnet56/checkpoints/00167... +INFO:tensorflow:global_step/sec: 0.586098 +INFO:tensorflow:step = 65301, loss = 0.125851, precision = 1.0 (170.620 sec) +INFO:tensorflow:global_step/sec: 0.588188 +INFO:tensorflow:step = 65401, loss = 0.125074, precision = 1.0 (170.014 sec) +INFO:tensorflow:global_step/sec: 0.59198 +INFO:tensorflow:step = 65501, loss = 0.124356, precision = 1.0 (168.925 sec) +INFO:tensorflow:global_step/sec: 0.593501 +INFO:tensorflow:step = 65601, loss = 0.150994, precision = 0.992188 (168.492 sec) +Saved checkpoint after 168 epoch(s) to data/resnet56/checkpoints/00168... +INFO:tensorflow:global_step/sec: 0.590954 +INFO:tensorflow:step = 65701, loss = 0.122475, precision = 1.0 (169.218 sec) +INFO:tensorflow:global_step/sec: 0.579357 +INFO:tensorflow:step = 65801, loss = 0.122748, precision = 1.0 (172.605 sec) +INFO:tensorflow:global_step/sec: 0.584896 +INFO:tensorflow:step = 65901, loss = 0.121829, precision = 1.0 (170.971 sec) +INFO:tensorflow:global_step/sec: 0.586391 +INFO:tensorflow:step = 66001, loss = 0.127703, precision = 1.0 (170.535 sec) +Saved checkpoint after 169 epoch(s) to data/resnet56/checkpoints/00169... +INFO:tensorflow:global_step/sec: 0.590149 +INFO:tensorflow:step = 66101, loss = 0.132194, precision = 0.992188 (169.449 sec) +INFO:tensorflow:global_step/sec: 0.596208 +INFO:tensorflow:step = 66201, loss = 0.129607, precision = 1.0 (167.727 sec) +INFO:tensorflow:global_step/sec: 0.592114 +INFO:tensorflow:step = 66301, loss = 0.121117, precision = 1.0 (168.886 sec) +INFO:tensorflow:global_step/sec: 0.591168 +INFO:tensorflow:step = 66401, loss = 0.127876, precision = 1.0 (169.156 sec) +Saved checkpoint after 170 epoch(s) to data/resnet56/checkpoints/00170... +INFO:tensorflow:global_step/sec: 0.585798 +INFO:tensorflow:step = 66501, loss = 0.122513, precision = 1.0 (170.707 sec) +INFO:tensorflow:global_step/sec: 0.595461 +INFO:tensorflow:step = 66601, loss = 0.122439, precision = 1.0 (167.937 sec) +INFO:tensorflow:global_step/sec: 0.593255 +INFO:tensorflow:step = 66701, loss = 0.127352, precision = 1.0 (168.561 sec) +INFO:tensorflow:global_step/sec: 0.595396 +INFO:tensorflow:step = 66801, loss = 0.122031, precision = 1.0 (167.956 sec) +Saved checkpoint after 171 epoch(s) to data/resnet56/checkpoints/00171... +INFO:tensorflow:global_step/sec: 0.592039 +INFO:tensorflow:step = 66901, loss = 0.120657, precision = 1.0 (168.908 sec) +INFO:tensorflow:global_step/sec: 0.589879 +INFO:tensorflow:step = 67001, loss = 0.121015, precision = 1.0 (169.526 sec) +INFO:tensorflow:global_step/sec: 0.597332 +INFO:tensorflow:step = 67101, loss = 0.123604, precision = 1.0 (167.411 sec) +INFO:tensorflow:global_step/sec: 0.595331 +INFO:tensorflow:step = 67201, loss = 0.122606, precision = 1.0 (167.973 sec) +Saved checkpoint after 172 epoch(s) to data/resnet56/checkpoints/00172... +INFO:tensorflow:global_step/sec: 0.595484 +INFO:tensorflow:step = 67301, loss = 0.121392, precision = 1.0 (167.931 sec) +INFO:tensorflow:global_step/sec: 0.59823 +INFO:tensorflow:step = 67401, loss = 0.123827, precision = 1.0 (167.160 sec) +INFO:tensorflow:global_step/sec: 0.59741 +INFO:tensorflow:step = 67501, loss = 0.122073, precision = 1.0 (167.389 sec) +INFO:tensorflow:global_step/sec: 0.594298 +INFO:tensorflow:step = 67601, loss = 0.132874, precision = 0.992188 (168.266 sec) +Saved checkpoint after 173 epoch(s) to data/resnet56/checkpoints/00173... +INFO:tensorflow:global_step/sec: 0.596911 +INFO:tensorflow:step = 67701, loss = 0.123427, precision = 1.0 (167.529 sec) +INFO:tensorflow:global_step/sec: 0.5922 +INFO:tensorflow:step = 67801, loss = 0.12294, precision = 1.0 (168.862 sec) +INFO:tensorflow:global_step/sec: 0.590859 +INFO:tensorflow:step = 67901, loss = 0.122604, precision = 1.0 (169.245 sec) +INFO:tensorflow:global_step/sec: 0.593479 +INFO:tensorflow:step = 68001, loss = 0.122426, precision = 1.0 (168.498 sec) +Saved checkpoint after 174 epoch(s) to data/resnet56/checkpoints/00174... +INFO:tensorflow:global_step/sec: 0.588567 +INFO:tensorflow:step = 68101, loss = 0.121063, precision = 1.0 (169.904 sec) +INFO:tensorflow:global_step/sec: 0.587747 +INFO:tensorflow:step = 68201, loss = 0.118646, precision = 1.0 (170.141 sec) +INFO:tensorflow:global_step/sec: 0.590958 +INFO:tensorflow:step = 68301, loss = 0.121767, precision = 1.0 (169.217 sec) +INFO:tensorflow:global_step/sec: 0.587256 +INFO:tensorflow:step = 68401, loss = 0.120933, precision = 1.0 (170.284 sec) +Saved checkpoint after 175 epoch(s) to data/resnet56/checkpoints/00175... +INFO:tensorflow:global_step/sec: 0.584247 +INFO:tensorflow:step = 68501, loss = 0.148222, precision = 0.992188 (171.161 sec) +INFO:tensorflow:global_step/sec: 0.585239 +INFO:tensorflow:step = 68601, loss = 0.120236, precision = 1.0 (170.870 sec) +INFO:tensorflow:global_step/sec: 0.597298 +INFO:tensorflow:step = 68701, loss = 0.122339, precision = 1.0 (167.421 sec) +INFO:tensorflow:global_step/sec: 0.586953 +INFO:tensorflow:step = 68801, loss = 0.123364, precision = 1.0 (170.371 sec) +Saved checkpoint after 176 epoch(s) to data/resnet56/checkpoints/00176... +INFO:tensorflow:global_step/sec: 0.59318 +INFO:tensorflow:step = 68901, loss = 0.119608, precision = 1.0 (168.583 sec) +INFO:tensorflow:global_step/sec: 0.591412 +INFO:tensorflow:step = 69001, loss = 0.120803, precision = 1.0 (169.086 sec) +INFO:tensorflow:global_step/sec: 0.599396 +INFO:tensorflow:step = 69101, loss = 0.120599, precision = 1.0 (166.835 sec) +INFO:tensorflow:global_step/sec: 0.588603 +INFO:tensorflow:step = 69201, loss = 0.138202, precision = 0.992188 (169.894 sec) +Saved checkpoint after 177 epoch(s) to data/resnet56/checkpoints/00177... +INFO:tensorflow:global_step/sec: 0.578105 +INFO:tensorflow:step = 69301, loss = 0.122722, precision = 1.0 (172.979 sec) +INFO:tensorflow:global_step/sec: 0.573496 +INFO:tensorflow:step = 69401, loss = 0.118368, precision = 1.0 (174.369 sec) +INFO:tensorflow:global_step/sec: 0.574733 +INFO:tensorflow:step = 69501, loss = 0.118571, precision = 1.0 (173.994 sec) +Saved checkpoint after 178 epoch(s) to data/resnet56/checkpoints/00178... +INFO:tensorflow:global_step/sec: 0.572655 +INFO:tensorflow:step = 69601, loss = 0.118751, precision = 1.0 (174.625 sec) +INFO:tensorflow:global_step/sec: 0.572581 +INFO:tensorflow:step = 69701, loss = 0.125554, precision = 1.0 (174.648 sec) +INFO:tensorflow:global_step/sec: 0.583363 +INFO:tensorflow:step = 69801, loss = 0.117577, precision = 1.0 (171.420 sec) +INFO:tensorflow:global_step/sec: 0.583653 +INFO:tensorflow:step = 69901, loss = 0.118768, precision = 1.0 (171.335 sec) +Saved checkpoint after 179 epoch(s) to data/resnet56/checkpoints/00179... +INFO:tensorflow:global_step/sec: 0.588548 +INFO:tensorflow:step = 70001, loss = 0.11834, precision = 1.0 (169.910 sec) +INFO:tensorflow:global_step/sec: 0.590881 +INFO:tensorflow:step = 70101, loss = 0.128075, precision = 1.0 (169.239 sec) +INFO:tensorflow:global_step/sec: 0.597342 +INFO:tensorflow:step = 70201, loss = 0.118014, precision = 1.0 (167.408 sec) +INFO:tensorflow:global_step/sec: 0.598737 +INFO:tensorflow:step = 70301, loss = 0.118766, precision = 1.0 (167.018 sec) +Saved checkpoint after 180 epoch(s) to data/resnet56/checkpoints/00180... +INFO:tensorflow:global_step/sec: 0.589837 +INFO:tensorflow:step = 70401, loss = 0.120127, precision = 1.0 (169.538 sec) +INFO:tensorflow:global_step/sec: 0.584336 +INFO:tensorflow:step = 70501, loss = 0.116933, precision = 1.0 (171.134 sec) +INFO:tensorflow:global_step/sec: 0.580785 +INFO:tensorflow:step = 70601, loss = 0.119157, precision = 1.0 (172.181 sec) +INFO:tensorflow:global_step/sec: 0.581011 +INFO:tensorflow:step = 70701, loss = 0.119451, precision = 1.0 (172.114 sec) +Saved checkpoint after 181 epoch(s) to data/resnet56/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1k80_ec2/resnet164_b_train.log b/tensorflow/CIFAR10/logs/1k80_ec2/resnet164_b_train.log new file mode 100644 index 0000000..10c9f5c --- /dev/null +++ b/tensorflow/CIFAR10/logs/1k80_ec2/resnet164_b_train.log @@ -0,0 +1,2222 @@ +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/1.69m params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (256x10, 2.56k/2.56k params) + logit/biases (10, 10/10 params) + unit_1_0/common_bn_relu/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (1x1x16x16, 256/256 params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub3/bn3/beta (16, 16/16 params) + unit_1_0/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_0/sub_add/project/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_1/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_1/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub3/bn3/beta (16, 16/16 params) + unit_1_1/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_10/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_10/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_10/sub2/bn2/beta (16, 16/16 params) + unit_1_10/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/sub3/bn3/beta (16, 16/16 params) + unit_1_10/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_11/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_11/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_11/sub2/bn2/beta (16, 16/16 params) + unit_1_11/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/sub3/bn3/beta (16, 16/16 params) + unit_1_11/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_12/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_12/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_12/sub2/bn2/beta (16, 16/16 params) + unit_1_12/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/sub3/bn3/beta (16, 16/16 params) + unit_1_12/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_13/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_13/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_13/sub2/bn2/beta (16, 16/16 params) + unit_1_13/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/sub3/bn3/beta (16, 16/16 params) + unit_1_13/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_14/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_14/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_14/sub2/bn2/beta (16, 16/16 params) + unit_1_14/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/sub3/bn3/beta (16, 16/16 params) + unit_1_14/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_15/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_15/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_15/sub2/bn2/beta (16, 16/16 params) + unit_1_15/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/sub3/bn3/beta (16, 16/16 params) + unit_1_15/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_16/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_16/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_16/sub2/bn2/beta (16, 16/16 params) + unit_1_16/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/sub3/bn3/beta (16, 16/16 params) + unit_1_16/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_17/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_17/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_17/sub2/bn2/beta (16, 16/16 params) + unit_1_17/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/sub3/bn3/beta (16, 16/16 params) + unit_1_17/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_2/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_2/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub3/bn3/beta (16, 16/16 params) + unit_1_2/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_3/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_3/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub3/bn3/beta (16, 16/16 params) + unit_1_3/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_4/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_4/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub3/bn3/beta (16, 16/16 params) + unit_1_4/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_5/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_5/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub3/bn3/beta (16, 16/16 params) + unit_1_5/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_6/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_6/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub3/bn3/beta (16, 16/16 params) + unit_1_6/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_7/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_7/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub3/bn3/beta (16, 16/16 params) + unit_1_7/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_8/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_8/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub3/bn3/beta (16, 16/16 params) + unit_1_8/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_9/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_9/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_9/sub2/bn2/beta (16, 16/16 params) + unit_1_9/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/sub3/bn3/beta (16, 16/16 params) + unit_1_9/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_2_0/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_2_0/sub1/conv1/DW (1x1x64x32, 2.05k/2.05k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_0/sub3/bn3/beta (32, 32/32 params) + unit_2_0/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_0/sub_add/project/DW (1x1x64x128, 8.19k/8.19k params) + unit_2_1/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_1/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub3/bn3/beta (32, 32/32 params) + unit_2_1/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_10/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_10/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_10/sub2/bn2/beta (32, 32/32 params) + unit_2_10/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/sub3/bn3/beta (32, 32/32 params) + unit_2_10/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_11/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_11/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_11/sub2/bn2/beta (32, 32/32 params) + unit_2_11/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/sub3/bn3/beta (32, 32/32 params) + unit_2_11/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_12/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_12/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_12/sub2/bn2/beta (32, 32/32 params) + unit_2_12/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/sub3/bn3/beta (32, 32/32 params) + unit_2_12/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_13/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_13/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_13/sub2/bn2/beta (32, 32/32 params) + unit_2_13/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/sub3/bn3/beta (32, 32/32 params) + unit_2_13/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_14/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_14/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_14/sub2/bn2/beta (32, 32/32 params) + unit_2_14/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/sub3/bn3/beta (32, 32/32 params) + unit_2_14/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_15/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_15/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_15/sub2/bn2/beta (32, 32/32 params) + unit_2_15/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/sub3/bn3/beta (32, 32/32 params) + unit_2_15/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_16/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_16/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_16/sub2/bn2/beta (32, 32/32 params) + unit_2_16/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/sub3/bn3/beta (32, 32/32 params) + unit_2_16/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_17/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_17/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_17/sub2/bn2/beta (32, 32/32 params) + unit_2_17/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/sub3/bn3/beta (32, 32/32 params) + unit_2_17/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_2/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_2/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub3/bn3/beta (32, 32/32 params) + unit_2_2/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_3/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_3/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub3/bn3/beta (32, 32/32 params) + unit_2_3/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_4/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_4/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub3/bn3/beta (32, 32/32 params) + unit_2_4/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_5/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_5/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub3/bn3/beta (32, 32/32 params) + unit_2_5/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_6/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_6/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub3/bn3/beta (32, 32/32 params) + unit_2_6/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_7/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_7/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub3/bn3/beta (32, 32/32 params) + unit_2_7/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_8/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_8/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub3/bn3/beta (32, 32/32 params) + unit_2_8/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_9/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_9/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_9/sub2/bn2/beta (32, 32/32 params) + unit_2_9/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/sub3/bn3/beta (32, 32/32 params) + unit_2_9/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_3_0/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_3_0/sub1/conv1/DW (1x1x128x64, 8.19k/8.19k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_0/sub3/bn3/beta (64, 64/64 params) + unit_3_0/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_0/sub_add/project/DW (1x1x128x256, 32.77k/32.77k params) + unit_3_1/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_1/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub3/bn3/beta (64, 64/64 params) + unit_3_1/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_10/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_10/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_10/sub2/bn2/beta (64, 64/64 params) + unit_3_10/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/sub3/bn3/beta (64, 64/64 params) + unit_3_10/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_11/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_11/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_11/sub2/bn2/beta (64, 64/64 params) + unit_3_11/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/sub3/bn3/beta (64, 64/64 params) + unit_3_11/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_12/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_12/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_12/sub2/bn2/beta (64, 64/64 params) + unit_3_12/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/sub3/bn3/beta (64, 64/64 params) + unit_3_12/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_13/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_13/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_13/sub2/bn2/beta (64, 64/64 params) + unit_3_13/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/sub3/bn3/beta (64, 64/64 params) + unit_3_13/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_14/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_14/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_14/sub2/bn2/beta (64, 64/64 params) + unit_3_14/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/sub3/bn3/beta (64, 64/64 params) + unit_3_14/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_15/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_15/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_15/sub2/bn2/beta (64, 64/64 params) + unit_3_15/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/sub3/bn3/beta (64, 64/64 params) + unit_3_15/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_16/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_16/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_16/sub2/bn2/beta (64, 64/64 params) + unit_3_16/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/sub3/bn3/beta (64, 64/64 params) + unit_3_16/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_17/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_17/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_17/sub2/bn2/beta (64, 64/64 params) + unit_3_17/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/sub3/bn3/beta (64, 64/64 params) + unit_3_17/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_2/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_2/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub3/bn3/beta (64, 64/64 params) + unit_3_2/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_3/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_3/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub3/bn3/beta (64, 64/64 params) + unit_3_3/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_4/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_4/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub3/bn3/beta (64, 64/64 params) + unit_3_4/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_5/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_5/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub3/bn3/beta (64, 64/64 params) + unit_3_5/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_6/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_6/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub3/bn3/beta (64, 64/64 params) + unit_3_6/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_7/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_7/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub3/bn3/beta (64, 64/64 params) + unit_3_7/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_8/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_8/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub3/bn3/beta (64, 64/64 params) + unit_3_8/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_9/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_9/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_9/sub2/bn2/beta (64, 64/64 params) + unit_3_9/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/sub3/bn3/beta (64, 64/64 params) + unit_3_9/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_last/final_bn/beta (256, 256/256 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/62.59b flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub_add/project/Conv2D (536.87m/536.87m flops) + unit_2_0/sub_add/project/Conv2D (536.87m/536.87m flops) + unit_3_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_0/sub_add/project/Conv2D (268.44m/268.44m flops) + unit_1_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_0/sub1/conv1/Conv2D (134.22m/134.22m flops) + unit_3_0/sub1/conv1/Conv2D (134.22m/134.22m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + unit_1_0/sub1/conv1/Conv2D (67.11m/67.11m flops) + logit/xw_plus_b (1.28k/656.64k flops) + logit/xw_plus_b/MatMul (655.36k/655.36k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (655.36k/655.36k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (655.36k/655.36k flops) + +======================End of Report========================== +2017-07-30 07:41:57.006265: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero +2017-07-30 07:41:57.006786: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla K80 +major: 3 minor: 7 memoryClockRate (GHz) 0.8235 +pciBusID 0000:00:1e.0 +Total memory: 11.17GiB +Free memory: 11.11GiB +2017-07-30 07:41:57.006818: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-07-30 07:41:57.006831: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-07-30 07:41:57.006854: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:00:1e.0) +2017-07-30 07:41:59.946054: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 07:41:59.946107: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 4 visible devices +2017-07-30 07:41:59.947413: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x8238f90 executing computations on platform Host. Devices: +2017-07-30 07:41:59.947437: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-07-30 07:41:59.948160: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 07:41:59.948196: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 4 visible devices +2017-07-30 07:41:59.948547: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x95fb9a0 executing computations on platform CUDA. Devices: +2017-07-30 07:41:59.948570: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla K80, Compute Capability 3.7 +2017-07-30 07:42:19.019344: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:247] PoolAllocator: After 1152 get requests, put_count=1100 evicted_count=1000 eviction_rate=0.909091 and unsatisfied allocation rate=1 +2017-07-30 07:42:19.019415: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:259] Raising pool_size_limit_ from 100 to 110 +INFO:tensorflow:step = 1, loss = 8.46424, precision = 0.0859375 +INFO:tensorflow:global_step/sec: 1.34358 +INFO:tensorflow:step = 101, loss = 7.82338, precision = 0.296875 (74.429 sec) +INFO:tensorflow:global_step/sec: 1.37965 +INFO:tensorflow:step = 201, loss = 7.60608, precision = 0.375 (72.482 sec) +INFO:tensorflow:global_step/sec: 1.37908 +INFO:tensorflow:step = 301, loss = 7.48315, precision = 0.4375 (72.512 sec) +total_params: 1691146 +Saved checkpoint after 1 epoch(s) to data/resnet164/checkpoints/00001... +INFO:tensorflow:global_step/sec: 1.33483 +INFO:tensorflow:step = 401, loss = 7.73463, precision = 0.296875 (74.916 sec) +INFO:tensorflow:global_step/sec: 1.3795 +INFO:tensorflow:step = 501, loss = 6.50942, precision = 0.5 (72.490 sec) +INFO:tensorflow:global_step/sec: 1.37821 +INFO:tensorflow:step = 601, loss = 6.1439, precision = 0.546875 (72.558 sec) +INFO:tensorflow:global_step/sec: 1.37714 +INFO:tensorflow:step = 701, loss = 5.59173, precision = 0.5625 (72.614 sec) +Saved checkpoint after 2 epoch(s) to data/resnet164/checkpoints/00002... +INFO:tensorflow:global_step/sec: 1.33812 +INFO:tensorflow:step = 801, loss = 5.01685, precision = 0.59375 (74.732 sec) +INFO:tensorflow:global_step/sec: 1.37733 +INFO:tensorflow:step = 901, loss = 4.41052, precision = 0.734375 (72.604 sec) +INFO:tensorflow:global_step/sec: 1.3776 +INFO:tensorflow:step = 1001, loss = 4.16249, precision = 0.664062 (72.590 sec) +INFO:tensorflow:global_step/sec: 1.37715 +INFO:tensorflow:step = 1101, loss = 3.93496, precision = 0.710938 (72.614 sec) +Saved checkpoint after 3 epoch(s) to data/resnet164/checkpoints/00003... +INFO:tensorflow:global_step/sec: 1.33741 +INFO:tensorflow:step = 1201, loss = 3.43293, precision = 0.742188 (74.771 sec) +INFO:tensorflow:global_step/sec: 1.3768 +INFO:tensorflow:step = 1301, loss = 3.27816, precision = 0.710938 (72.632 sec) +INFO:tensorflow:global_step/sec: 1.37619 +INFO:tensorflow:step = 1401, loss = 3.07797, precision = 0.710938 (72.665 sec) +INFO:tensorflow:global_step/sec: 1.37679 +INFO:tensorflow:step = 1501, loss = 2.83782, precision = 0.765625 (72.633 sec) +Saved checkpoint after 4 epoch(s) to data/resnet164/checkpoints/00004... +INFO:tensorflow:global_step/sec: 1.33718 +INFO:tensorflow:step = 1601, loss = 2.7081, precision = 0.734375 (74.784 sec) +INFO:tensorflow:global_step/sec: 1.37689 +INFO:tensorflow:step = 1701, loss = 2.45321, precision = 0.78125 (72.627 sec) +INFO:tensorflow:global_step/sec: 1.37674 +INFO:tensorflow:step = 1801, loss = 2.22573, precision = 0.78125 (72.635 sec) +INFO:tensorflow:global_step/sec: 1.37706 +INFO:tensorflow:step = 1901, loss = 2.36243, precision = 0.726562 (72.618 sec) +Saved checkpoint after 5 epoch(s) to data/resnet164/checkpoints/00005... +INFO:tensorflow:global_step/sec: 1.33696 +INFO:tensorflow:step = 2001, loss = 1.97359, precision = 0.789062 (74.797 sec) +INFO:tensorflow:global_step/sec: 1.37709 +INFO:tensorflow:step = 2101, loss = 1.87949, precision = 0.765625 (72.617 sec) +INFO:tensorflow:global_step/sec: 1.37643 +INFO:tensorflow:step = 2201, loss = 1.82702, precision = 0.78125 (72.651 sec) +INFO:tensorflow:global_step/sec: 1.37721 +INFO:tensorflow:step = 2301, loss = 1.66494, precision = 0.773438 (72.611 sec) +Saved checkpoint after 6 epoch(s) to data/resnet164/checkpoints/00006... +INFO:tensorflow:global_step/sec: 1.33657 +INFO:tensorflow:step = 2401, loss = 1.77858, precision = 0.757812 (74.818 sec) +INFO:tensorflow:global_step/sec: 1.37661 +INFO:tensorflow:step = 2501, loss = 1.53193, precision = 0.78125 (72.642 sec) +INFO:tensorflow:global_step/sec: 1.377 +INFO:tensorflow:step = 2601, loss = 1.52141, precision = 0.710938 (72.622 sec) +INFO:tensorflow:global_step/sec: 1.37709 +INFO:tensorflow:step = 2701, loss = 1.47355, precision = 0.734375 (72.617 sec) +Saved checkpoint after 7 epoch(s) to data/resnet164/checkpoints/00007... +INFO:tensorflow:global_step/sec: 1.33741 +INFO:tensorflow:step = 2801, loss = 1.42488, precision = 0.75 (74.772 sec) +INFO:tensorflow:global_step/sec: 1.37627 +INFO:tensorflow:step = 2901, loss = 1.16906, precision = 0.835938 (72.660 sec) +INFO:tensorflow:global_step/sec: 1.37696 +INFO:tensorflow:step = 3001, loss = 1.21333, precision = 0.78125 (72.624 sec) +INFO:tensorflow:global_step/sec: 1.37645 +INFO:tensorflow:step = 3101, loss = 1.25673, precision = 0.796875 (72.651 sec) +Saved checkpoint after 8 epoch(s) to data/resnet164/checkpoints/00008... +INFO:tensorflow:global_step/sec: 1.33345 +INFO:tensorflow:step = 3201, loss = 1.12581, precision = 0.796875 (74.993 sec) +INFO:tensorflow:global_step/sec: 1.37682 +INFO:tensorflow:step = 3301, loss = 1.16972, precision = 0.820312 (72.631 sec) +INFO:tensorflow:global_step/sec: 1.37628 +INFO:tensorflow:step = 3401, loss = 1.06053, precision = 0.820312 (72.659 sec) +INFO:tensorflow:global_step/sec: 1.37655 +INFO:tensorflow:step = 3501, loss = 1.02247, precision = 0.851562 (72.646 sec) +Saved checkpoint after 9 epoch(s) to data/resnet164/checkpoints/00009... +INFO:tensorflow:global_step/sec: 1.3371 +INFO:tensorflow:step = 3601, loss = 0.943095, precision = 0.84375 (74.789 sec) +INFO:tensorflow:global_step/sec: 1.37626 +INFO:tensorflow:step = 3701, loss = 1.07037, precision = 0.789062 (72.660 sec) +INFO:tensorflow:global_step/sec: 1.37736 +INFO:tensorflow:step = 3801, loss = 1.07123, precision = 0.773438 (72.603 sec) +INFO:tensorflow:global_step/sec: 1.37783 +INFO:tensorflow:step = 3901, loss = 0.856166, precision = 0.859375 (72.578 sec) +Saved checkpoint after 10 epoch(s) to data/resnet164/checkpoints/00010... +INFO:tensorflow:global_step/sec: 1.33766 +INFO:tensorflow:step = 4001, loss = 0.94175, precision = 0.8125 (74.757 sec) +INFO:tensorflow:global_step/sec: 1.37744 +INFO:tensorflow:step = 4101, loss = 0.88265, precision = 0.828125 (72.598 sec) +INFO:tensorflow:global_step/sec: 1.37722 +INFO:tensorflow:step = 4201, loss = 0.968336, precision = 0.8125 (72.610 sec) +Saved checkpoint after 11 epoch(s) to data/resnet164/checkpoints/00011... +INFO:tensorflow:global_step/sec: 1.33725 +INFO:tensorflow:step = 4301, loss = 0.893978, precision = 0.8125 (74.781 sec) +INFO:tensorflow:global_step/sec: 1.37688 +INFO:tensorflow:step = 4401, loss = 0.787089, precision = 0.867188 (72.628 sec) +INFO:tensorflow:global_step/sec: 1.37704 +INFO:tensorflow:step = 4501, loss = 1.03729, precision = 0.765625 (72.620 sec) +INFO:tensorflow:global_step/sec: 1.37689 +INFO:tensorflow:step = 4601, loss = 0.824919, precision = 0.875 (72.627 sec) +Saved checkpoint after 12 epoch(s) to data/resnet164/checkpoints/00012... +INFO:tensorflow:global_step/sec: 1.34006 +INFO:tensorflow:step = 4701, loss = 0.950717, precision = 0.796875 (74.624 sec) +INFO:tensorflow:global_step/sec: 1.37646 +INFO:tensorflow:step = 4801, loss = 0.712384, precision = 0.867188 (72.650 sec) +INFO:tensorflow:global_step/sec: 1.37723 +INFO:tensorflow:step = 4901, loss = 0.804012, precision = 0.84375 (72.609 sec) +INFO:tensorflow:global_step/sec: 1.37729 +INFO:tensorflow:step = 5001, loss = 0.987956, precision = 0.820312 (72.606 sec) +Saved checkpoint after 13 epoch(s) to data/resnet164/checkpoints/00013... +INFO:tensorflow:global_step/sec: 1.33897 +INFO:tensorflow:step = 5101, loss = 0.706754, precision = 0.867188 (74.684 sec) +INFO:tensorflow:global_step/sec: 1.37682 +INFO:tensorflow:step = 5201, loss = 0.946938, precision = 0.796875 (72.631 sec) +INFO:tensorflow:global_step/sec: 1.37719 +INFO:tensorflow:step = 5301, loss = 0.838443, precision = 0.804688 (72.611 sec) +INFO:tensorflow:global_step/sec: 1.37787 +INFO:tensorflow:step = 5401, loss = 0.88765, precision = 0.84375 (72.576 sec) +Saved checkpoint after 14 epoch(s) to data/resnet164/checkpoints/00014... +INFO:tensorflow:global_step/sec: 1.33922 +INFO:tensorflow:step = 5501, loss = 0.818394, precision = 0.835938 (74.670 sec) +INFO:tensorflow:global_step/sec: 1.37739 +INFO:tensorflow:step = 5601, loss = 0.83783, precision = 0.828125 (72.601 sec) +INFO:tensorflow:global_step/sec: 1.37628 +INFO:tensorflow:step = 5701, loss = 0.844821, precision = 0.84375 (72.659 sec) +INFO:tensorflow:global_step/sec: 1.37768 +INFO:tensorflow:step = 5801, loss = 0.919349, precision = 0.804688 (72.586 sec) +Saved checkpoint after 15 epoch(s) to data/resnet164/checkpoints/00015... +INFO:tensorflow:global_step/sec: 1.33988 +INFO:tensorflow:step = 5901, loss = 0.858986, precision = 0.828125 (74.634 sec) +INFO:tensorflow:global_step/sec: 1.37694 +INFO:tensorflow:step = 6001, loss = 0.617653, precision = 0.890625 (72.625 sec) +INFO:tensorflow:global_step/sec: 1.37738 +INFO:tensorflow:step = 6101, loss = 0.667601, precision = 0.90625 (72.602 sec) +INFO:tensorflow:global_step/sec: 1.37687 +INFO:tensorflow:step = 6201, loss = 0.918304, precision = 0.8125 (72.628 sec) +Saved checkpoint after 16 epoch(s) to data/resnet164/checkpoints/00016... +INFO:tensorflow:global_step/sec: 1.34037 +INFO:tensorflow:step = 6301, loss = 0.746274, precision = 0.851562 (74.607 sec) +INFO:tensorflow:global_step/sec: 1.3771 +INFO:tensorflow:step = 6401, loss = 0.843747, precision = 0.835938 (72.616 sec) +INFO:tensorflow:global_step/sec: 1.37696 +INFO:tensorflow:step = 6501, loss = 0.908957, precision = 0.804688 (72.624 sec) +INFO:tensorflow:global_step/sec: 1.37727 +INFO:tensorflow:step = 6601, loss = 0.951037, precision = 0.78125 (72.607 sec) +Saved checkpoint after 17 epoch(s) to data/resnet164/checkpoints/00017... +INFO:tensorflow:global_step/sec: 1.33957 +INFO:tensorflow:step = 6701, loss = 0.989919, precision = 0.75 (74.651 sec) +INFO:tensorflow:global_step/sec: 1.37763 +INFO:tensorflow:step = 6801, loss = 0.796029, precision = 0.828125 (72.588 sec) +INFO:tensorflow:global_step/sec: 1.37777 +INFO:tensorflow:step = 6901, loss = 0.682412, precision = 0.882812 (72.581 sec) +INFO:tensorflow:global_step/sec: 1.37755 +INFO:tensorflow:step = 7001, loss = 0.624103, precision = 0.90625 (72.593 sec) +Saved checkpoint after 18 epoch(s) to data/resnet164/checkpoints/00018... +INFO:tensorflow:global_step/sec: 1.3351 +INFO:tensorflow:step = 7101, loss = 0.682106, precision = 0.859375 (74.901 sec) +INFO:tensorflow:global_step/sec: 1.37795 +INFO:tensorflow:step = 7201, loss = 0.887067, precision = 0.765625 (72.572 sec) +INFO:tensorflow:global_step/sec: 1.37756 +INFO:tensorflow:step = 7301, loss = 0.615967, precision = 0.882812 (72.592 sec) +INFO:tensorflow:global_step/sec: 1.37759 +INFO:tensorflow:step = 7401, loss = 0.690455, precision = 0.882812 (72.590 sec) +Saved checkpoint after 19 epoch(s) to data/resnet164/checkpoints/00019... +INFO:tensorflow:global_step/sec: 1.33976 +INFO:tensorflow:step = 7501, loss = 0.770583, precision = 0.835938 (74.640 sec) +INFO:tensorflow:global_step/sec: 1.37784 +INFO:tensorflow:step = 7601, loss = 0.715225, precision = 0.882812 (72.577 sec) +INFO:tensorflow:global_step/sec: 1.37758 +INFO:tensorflow:step = 7701, loss = 0.732184, precision = 0.867188 (72.591 sec) +INFO:tensorflow:global_step/sec: 1.37869 +INFO:tensorflow:step = 7801, loss = 0.704029, precision = 0.867188 (72.532 sec) +Saved checkpoint after 20 epoch(s) to data/resnet164/checkpoints/00020... +INFO:tensorflow:global_step/sec: 1.34037 +INFO:tensorflow:step = 7901, loss = 0.797964, precision = 0.820312 (74.606 sec) +INFO:tensorflow:global_step/sec: 1.37702 +INFO:tensorflow:step = 8001, loss = 0.677112, precision = 0.875 (72.621 sec) +INFO:tensorflow:global_step/sec: 1.37774 +INFO:tensorflow:step = 8101, loss = 0.694703, precision = 0.851562 (72.583 sec) +INFO:tensorflow:global_step/sec: 1.37838 +INFO:tensorflow:step = 8201, loss = 0.650832, precision = 0.90625 (72.549 sec) +Saved checkpoint after 21 epoch(s) to data/resnet164/checkpoints/00021... +INFO:tensorflow:global_step/sec: 1.33932 +INFO:tensorflow:step = 8301, loss = 0.834305, precision = 0.828125 (74.665 sec) +INFO:tensorflow:global_step/sec: 1.37732 +INFO:tensorflow:step = 8401, loss = 0.806239, precision = 0.835938 (72.605 sec) +INFO:tensorflow:global_step/sec: 1.37789 +INFO:tensorflow:step = 8501, loss = 0.767266, precision = 0.84375 (72.575 sec) +INFO:tensorflow:global_step/sec: 1.37762 +INFO:tensorflow:step = 8601, loss = 0.843163, precision = 0.804688 (72.589 sec) +Saved checkpoint after 22 epoch(s) to data/resnet164/checkpoints/00022... +INFO:tensorflow:global_step/sec: 1.33982 +INFO:tensorflow:step = 8701, loss = 0.751494, precision = 0.867188 (74.637 sec) +INFO:tensorflow:global_step/sec: 1.37761 +INFO:tensorflow:step = 8801, loss = 0.802086, precision = 0.820312 (72.589 sec) +INFO:tensorflow:global_step/sec: 1.37722 +INFO:tensorflow:step = 8901, loss = 0.723777, precision = 0.851562 (72.610 sec) +Saved checkpoint after 23 epoch(s) to data/resnet164/checkpoints/00023... +INFO:tensorflow:global_step/sec: 1.34071 +INFO:tensorflow:step = 9001, loss = 0.731245, precision = 0.84375 (74.587 sec) +INFO:tensorflow:global_step/sec: 1.3774 +INFO:tensorflow:step = 9101, loss = 0.722927, precision = 0.851562 (72.601 sec) +INFO:tensorflow:global_step/sec: 1.37709 +INFO:tensorflow:step = 9201, loss = 0.61571, precision = 0.875 (72.617 sec) +INFO:tensorflow:global_step/sec: 1.37699 +INFO:tensorflow:step = 9301, loss = 0.610563, precision = 0.890625 (72.622 sec) +Saved checkpoint after 24 epoch(s) to data/resnet164/checkpoints/00024... +INFO:tensorflow:global_step/sec: 1.34035 +INFO:tensorflow:step = 9401, loss = 0.727506, precision = 0.835938 (74.608 sec) +INFO:tensorflow:global_step/sec: 1.37766 +INFO:tensorflow:step = 9501, loss = 0.761504, precision = 0.867188 (72.587 sec) +INFO:tensorflow:global_step/sec: 1.37865 +INFO:tensorflow:step = 9601, loss = 0.622428, precision = 0.890625 (72.535 sec) +INFO:tensorflow:global_step/sec: 1.37719 +INFO:tensorflow:step = 9701, loss = 0.687173, precision = 0.867188 (72.611 sec) +Saved checkpoint after 25 epoch(s) to data/resnet164/checkpoints/00025... +INFO:tensorflow:global_step/sec: 1.33969 +INFO:tensorflow:step = 9801, loss = 0.719173, precision = 0.867188 (74.644 sec) +INFO:tensorflow:global_step/sec: 1.37754 +INFO:tensorflow:step = 9901, loss = 0.564124, precision = 0.929688 (72.593 sec) +INFO:tensorflow:global_step/sec: 1.37764 +INFO:tensorflow:step = 10001, loss = 0.812655, precision = 0.84375 (72.588 sec) +INFO:tensorflow:global_step/sec: 1.37718 +INFO:tensorflow:step = 10101, loss = 0.742682, precision = 0.851562 (72.612 sec) +Saved checkpoint after 26 epoch(s) to data/resnet164/checkpoints/00026... +INFO:tensorflow:global_step/sec: 1.34043 +INFO:tensorflow:step = 10201, loss = 0.627432, precision = 0.90625 (74.603 sec) +INFO:tensorflow:global_step/sec: 1.37766 +INFO:tensorflow:step = 10301, loss = 0.600433, precision = 0.921875 (72.587 sec) +INFO:tensorflow:global_step/sec: 1.37779 +INFO:tensorflow:step = 10401, loss = 0.667424, precision = 0.875 (72.580 sec) +INFO:tensorflow:global_step/sec: 1.37806 +INFO:tensorflow:step = 10501, loss = 0.681582, precision = 0.890625 (72.566 sec) +Saved checkpoint after 27 epoch(s) to data/resnet164/checkpoints/00027... +INFO:tensorflow:global_step/sec: 1.34011 +INFO:tensorflow:step = 10601, loss = 0.73994, precision = 0.859375 (74.621 sec) +INFO:tensorflow:global_step/sec: 1.37824 +INFO:tensorflow:step = 10701, loss = 0.776297, precision = 0.859375 (72.556 sec) +INFO:tensorflow:global_step/sec: 1.37816 +INFO:tensorflow:step = 10801, loss = 0.718323, precision = 0.875 (72.560 sec) +INFO:tensorflow:global_step/sec: 1.37788 +INFO:tensorflow:step = 10901, loss = 0.640572, precision = 0.90625 (72.576 sec) +Saved checkpoint after 28 epoch(s) to data/resnet164/checkpoints/00028... +INFO:tensorflow:global_step/sec: 1.34088 +INFO:tensorflow:step = 11001, loss = 0.591966, precision = 0.929688 (74.578 sec) +INFO:tensorflow:global_step/sec: 1.37812 +INFO:tensorflow:step = 11101, loss = 0.737495, precision = 0.84375 (72.563 sec) +INFO:tensorflow:global_step/sec: 1.37807 +INFO:tensorflow:step = 11201, loss = 0.677721, precision = 0.875 (72.565 sec) +INFO:tensorflow:global_step/sec: 1.37827 +INFO:tensorflow:step = 11301, loss = 0.688239, precision = 0.882812 (72.555 sec) +Saved checkpoint after 29 epoch(s) to data/resnet164/checkpoints/00029... +INFO:tensorflow:global_step/sec: 1.33672 +INFO:tensorflow:step = 11401, loss = 0.706461, precision = 0.867188 (74.810 sec) +INFO:tensorflow:global_step/sec: 1.3781 +INFO:tensorflow:step = 11501, loss = 0.67039, precision = 0.867188 (72.564 sec) +INFO:tensorflow:global_step/sec: 1.37743 +INFO:tensorflow:step = 11601, loss = 0.699218, precision = 0.867188 (72.599 sec) +INFO:tensorflow:global_step/sec: 1.37813 +INFO:tensorflow:step = 11701, loss = 0.750121, precision = 0.851562 (72.562 sec) +Saved checkpoint after 30 epoch(s) to data/resnet164/checkpoints/00030... +INFO:tensorflow:global_step/sec: 1.34078 +INFO:tensorflow:step = 11801, loss = 0.655804, precision = 0.875 (74.584 sec) +INFO:tensorflow:global_step/sec: 1.37824 +INFO:tensorflow:step = 11901, loss = 0.595699, precision = 0.882812 (72.556 sec) +INFO:tensorflow:global_step/sec: 1.37842 +INFO:tensorflow:step = 12001, loss = 0.657342, precision = 0.875 (72.546 sec) +INFO:tensorflow:global_step/sec: 1.37752 +INFO:tensorflow:step = 12101, loss = 0.693741, precision = 0.90625 (72.594 sec) +Saved checkpoint after 31 epoch(s) to data/resnet164/checkpoints/00031... +INFO:tensorflow:global_step/sec: 1.34098 +INFO:tensorflow:step = 12201, loss = 0.614049, precision = 0.890625 (74.573 sec) +INFO:tensorflow:global_step/sec: 1.37887 +INFO:tensorflow:step = 12301, loss = 0.595076, precision = 0.921875 (72.523 sec) +INFO:tensorflow:global_step/sec: 1.37757 +INFO:tensorflow:step = 12401, loss = 0.691496, precision = 0.882812 (72.592 sec) +INFO:tensorflow:global_step/sec: 1.37844 +INFO:tensorflow:step = 12501, loss = 0.681837, precision = 0.859375 (72.546 sec) +Saved checkpoint after 32 epoch(s) to data/resnet164/checkpoints/00032... +INFO:tensorflow:global_step/sec: 1.34117 +INFO:tensorflow:step = 12601, loss = 0.65547, precision = 0.890625 (74.562 sec) +INFO:tensorflow:global_step/sec: 1.3784 +INFO:tensorflow:step = 12701, loss = 0.638318, precision = 0.882812 (72.548 sec) +INFO:tensorflow:global_step/sec: 1.37788 +INFO:tensorflow:step = 12801, loss = 0.643844, precision = 0.898438 (72.575 sec) +INFO:tensorflow:global_step/sec: 1.37801 +INFO:tensorflow:step = 12901, loss = 0.618052, precision = 0.921875 (72.569 sec) +Saved checkpoint after 33 epoch(s) to data/resnet164/checkpoints/00033... +INFO:tensorflow:global_step/sec: 1.34009 +INFO:tensorflow:step = 13001, loss = 0.627401, precision = 0.882812 (74.622 sec) +INFO:tensorflow:global_step/sec: 1.3782 +INFO:tensorflow:step = 13101, loss = 0.647483, precision = 0.882812 (72.558 sec) +INFO:tensorflow:global_step/sec: 1.37769 +INFO:tensorflow:step = 13201, loss = 0.719269, precision = 0.851562 (72.585 sec) +Saved checkpoint after 34 epoch(s) to data/resnet164/checkpoints/00034... +INFO:tensorflow:global_step/sec: 1.33976 +INFO:tensorflow:step = 13301, loss = 0.682402, precision = 0.867188 (74.640 sec) +INFO:tensorflow:global_step/sec: 1.37895 +INFO:tensorflow:step = 13401, loss = 0.749962, precision = 0.835938 (72.519 sec) +INFO:tensorflow:global_step/sec: 1.37818 +INFO:tensorflow:step = 13501, loss = 0.65225, precision = 0.90625 (72.559 sec) +INFO:tensorflow:global_step/sec: 1.37835 +INFO:tensorflow:step = 13601, loss = 0.550811, precision = 0.960938 (72.551 sec) +Saved checkpoint after 35 epoch(s) to data/resnet164/checkpoints/00035... +INFO:tensorflow:global_step/sec: 1.3408 +INFO:tensorflow:step = 13701, loss = 0.745314, precision = 0.859375 (74.582 sec) +INFO:tensorflow:global_step/sec: 1.37828 +INFO:tensorflow:step = 13801, loss = 0.817602, precision = 0.84375 (72.554 sec) +INFO:tensorflow:global_step/sec: 1.37809 +INFO:tensorflow:step = 13901, loss = 0.689746, precision = 0.867188 (72.564 sec) +INFO:tensorflow:global_step/sec: 1.37921 +INFO:tensorflow:step = 14001, loss = 0.80642, precision = 0.859375 (72.505 sec) +Saved checkpoint after 36 epoch(s) to data/resnet164/checkpoints/00036... +INFO:tensorflow:global_step/sec: 1.34202 +INFO:tensorflow:step = 14101, loss = 0.748369, precision = 0.851562 (74.515 sec) +INFO:tensorflow:global_step/sec: 1.37901 +INFO:tensorflow:step = 14201, loss = 0.703682, precision = 0.859375 (72.516 sec) +INFO:tensorflow:global_step/sec: 1.37849 +INFO:tensorflow:step = 14301, loss = 0.639059, precision = 0.890625 (72.543 sec) +INFO:tensorflow:global_step/sec: 1.37836 +INFO:tensorflow:step = 14401, loss = 0.745252, precision = 0.804688 (72.550 sec) +Saved checkpoint after 37 epoch(s) to data/resnet164/checkpoints/00037... +INFO:tensorflow:global_step/sec: 1.34033 +INFO:tensorflow:step = 14501, loss = 0.570316, precision = 0.921875 (74.609 sec) +INFO:tensorflow:global_step/sec: 1.37915 +INFO:tensorflow:step = 14601, loss = 0.787856, precision = 0.851562 (72.508 sec) +INFO:tensorflow:global_step/sec: 1.37838 +INFO:tensorflow:step = 14701, loss = 0.59748, precision = 0.898438 (72.549 sec) +INFO:tensorflow:global_step/sec: 1.37867 +INFO:tensorflow:step = 14801, loss = 0.815916, precision = 0.828125 (72.534 sec) +Saved checkpoint after 38 epoch(s) to data/resnet164/checkpoints/00038... +INFO:tensorflow:global_step/sec: 1.3413 +INFO:tensorflow:step = 14901, loss = 0.622122, precision = 0.90625 (74.555 sec) +INFO:tensorflow:global_step/sec: 1.37926 +INFO:tensorflow:step = 15001, loss = 0.730797, precision = 0.851562 (72.502 sec) +INFO:tensorflow:global_step/sec: 1.37901 +INFO:tensorflow:step = 15101, loss = 0.663222, precision = 0.898438 (72.516 sec) +INFO:tensorflow:global_step/sec: 1.37879 +INFO:tensorflow:step = 15201, loss = 0.741436, precision = 0.867188 (72.527 sec) +Saved checkpoint after 39 epoch(s) to data/resnet164/checkpoints/00039... +INFO:tensorflow:global_step/sec: 1.33812 +INFO:tensorflow:step = 15301, loss = 0.714269, precision = 0.867188 (74.732 sec) +INFO:tensorflow:global_step/sec: 1.37878 +INFO:tensorflow:step = 15401, loss = 0.685099, precision = 0.90625 (72.528 sec) +INFO:tensorflow:global_step/sec: 1.37831 +INFO:tensorflow:step = 15501, loss = 0.802023, precision = 0.859375 (72.553 sec) +INFO:tensorflow:global_step/sec: 1.37858 +INFO:tensorflow:step = 15601, loss = 0.664628, precision = 0.898438 (72.538 sec) +Saved checkpoint after 40 epoch(s) to data/resnet164/checkpoints/00040... +INFO:tensorflow:global_step/sec: 1.34155 +INFO:tensorflow:step = 15701, loss = 0.648857, precision = 0.890625 (74.541 sec) +INFO:tensorflow:global_step/sec: 1.37886 +INFO:tensorflow:step = 15801, loss = 0.698171, precision = 0.867188 (72.523 sec) +INFO:tensorflow:global_step/sec: 1.37839 +INFO:tensorflow:step = 15901, loss = 0.724847, precision = 0.851562 (72.548 sec) +INFO:tensorflow:global_step/sec: 1.37894 +INFO:tensorflow:step = 16001, loss = 0.66815, precision = 0.890625 (72.519 sec) +Saved checkpoint after 41 epoch(s) to data/resnet164/checkpoints/00041... +INFO:tensorflow:global_step/sec: 1.34015 +INFO:tensorflow:step = 16101, loss = 0.700409, precision = 0.875 (74.619 sec) +INFO:tensorflow:global_step/sec: 1.37865 +INFO:tensorflow:step = 16201, loss = 0.582918, precision = 0.914062 (72.534 sec) +INFO:tensorflow:global_step/sec: 1.37885 +INFO:tensorflow:step = 16301, loss = 0.743346, precision = 0.828125 (72.524 sec) +INFO:tensorflow:global_step/sec: 1.37845 +INFO:tensorflow:step = 16401, loss = 0.631356, precision = 0.882812 (72.545 sec) +Saved checkpoint after 42 epoch(s) to data/resnet164/checkpoints/00042... +INFO:tensorflow:global_step/sec: 1.34088 +INFO:tensorflow:step = 16501, loss = 0.657592, precision = 0.875 (74.578 sec) +INFO:tensorflow:global_step/sec: 1.3781 +INFO:tensorflow:step = 16601, loss = 0.647792, precision = 0.90625 (72.563 sec) +INFO:tensorflow:global_step/sec: 1.37808 +INFO:tensorflow:step = 16701, loss = 0.67499, precision = 0.867188 (72.565 sec) +INFO:tensorflow:global_step/sec: 1.37921 +INFO:tensorflow:step = 16801, loss = 0.672154, precision = 0.875 (72.505 sec) +Saved checkpoint after 43 epoch(s) to data/resnet164/checkpoints/00043... +INFO:tensorflow:global_step/sec: 1.34087 +INFO:tensorflow:step = 16901, loss = 0.699771, precision = 0.867188 (74.578 sec) +INFO:tensorflow:global_step/sec: 1.37846 +INFO:tensorflow:step = 17001, loss = 0.556742, precision = 0.929688 (72.544 sec) +INFO:tensorflow:global_step/sec: 1.37834 +INFO:tensorflow:step = 17101, loss = 0.783389, precision = 0.851562 (72.551 sec) +INFO:tensorflow:global_step/sec: 1.37748 +INFO:tensorflow:step = 17201, loss = 0.620527, precision = 0.90625 (72.596 sec) +Saved checkpoint after 44 epoch(s) to data/resnet164/checkpoints/00044... +INFO:tensorflow:global_step/sec: 1.34058 +INFO:tensorflow:step = 17301, loss = 0.775469, precision = 0.851562 (74.595 sec) +INFO:tensorflow:global_step/sec: 1.37842 +INFO:tensorflow:step = 17401, loss = 0.632777, precision = 0.882812 (72.547 sec) +INFO:tensorflow:global_step/sec: 1.37796 +INFO:tensorflow:step = 17501, loss = 0.786241, precision = 0.835938 (72.571 sec) +Saved checkpoint after 45 epoch(s) to data/resnet164/checkpoints/00045... +INFO:tensorflow:global_step/sec: 1.34099 +INFO:tensorflow:step = 17601, loss = 0.6207, precision = 0.90625 (74.572 sec) +INFO:tensorflow:global_step/sec: 1.3781 +INFO:tensorflow:step = 17701, loss = 0.58297, precision = 0.914062 (72.563 sec) +INFO:tensorflow:global_step/sec: 1.37768 +INFO:tensorflow:step = 17801, loss = 0.6471, precision = 0.867188 (72.586 sec) +INFO:tensorflow:global_step/sec: 1.3779 +INFO:tensorflow:step = 17901, loss = 0.629814, precision = 0.890625 (72.574 sec) +Saved checkpoint after 46 epoch(s) to data/resnet164/checkpoints/00046... +INFO:tensorflow:global_step/sec: 1.34128 +INFO:tensorflow:step = 18001, loss = 0.617315, precision = 0.882812 (74.556 sec) +INFO:tensorflow:global_step/sec: 1.3787 +INFO:tensorflow:step = 18101, loss = 0.646881, precision = 0.882812 (72.532 sec) +INFO:tensorflow:global_step/sec: 1.37888 +INFO:tensorflow:step = 18201, loss = 0.622697, precision = 0.898438 (72.522 sec) +INFO:tensorflow:global_step/sec: 1.37882 +INFO:tensorflow:step = 18301, loss = 0.688701, precision = 0.898438 (72.526 sec) +Saved checkpoint after 47 epoch(s) to data/resnet164/checkpoints/00047... +INFO:tensorflow:global_step/sec: 1.34019 +INFO:tensorflow:step = 18401, loss = 0.734685, precision = 0.828125 (74.616 sec) +INFO:tensorflow:global_step/sec: 1.37847 +INFO:tensorflow:step = 18501, loss = 0.773936, precision = 0.835938 (72.544 sec) +INFO:tensorflow:global_step/sec: 1.3788 +INFO:tensorflow:step = 18601, loss = 0.752637, precision = 0.890625 (72.527 sec) +INFO:tensorflow:global_step/sec: 1.3787 +INFO:tensorflow:step = 18701, loss = 0.751309, precision = 0.890625 (72.532 sec) +Saved checkpoint after 48 epoch(s) to data/resnet164/checkpoints/00048... +INFO:tensorflow:global_step/sec: 1.34098 +INFO:tensorflow:step = 18801, loss = 0.67672, precision = 0.867188 (74.572 sec) +INFO:tensorflow:global_step/sec: 1.37859 +INFO:tensorflow:step = 18901, loss = 0.600304, precision = 0.90625 (72.538 sec) +INFO:tensorflow:global_step/sec: 1.37797 +INFO:tensorflow:step = 19001, loss = 0.534198, precision = 0.945312 (72.570 sec) +INFO:tensorflow:global_step/sec: 1.3783 +INFO:tensorflow:step = 19101, loss = 0.817732, precision = 0.867188 (72.553 sec) +Saved checkpoint after 49 epoch(s) to data/resnet164/checkpoints/00049... +INFO:tensorflow:global_step/sec: 1.33865 +INFO:tensorflow:step = 19201, loss = 0.609718, precision = 0.9375 (74.702 sec) +INFO:tensorflow:global_step/sec: 1.3791 +INFO:tensorflow:step = 19301, loss = 0.60526, precision = 0.898438 (72.511 sec) +INFO:tensorflow:global_step/sec: 1.37886 +INFO:tensorflow:step = 19401, loss = 0.570862, precision = 0.882812 (72.524 sec) +INFO:tensorflow:global_step/sec: 1.37877 +INFO:tensorflow:step = 19501, loss = 0.625753, precision = 0.875 (72.528 sec) +Saved checkpoint after 50 epoch(s) to data/resnet164/checkpoints/00050... +INFO:tensorflow:global_step/sec: 1.34141 +INFO:tensorflow:step = 19601, loss = 0.591345, precision = 0.890625 (74.549 sec) +INFO:tensorflow:global_step/sec: 1.37896 +INFO:tensorflow:step = 19701, loss = 0.660099, precision = 0.898438 (72.519 sec) +INFO:tensorflow:global_step/sec: 1.37817 +INFO:tensorflow:step = 19801, loss = 0.679568, precision = 0.875 (72.560 sec) +INFO:tensorflow:global_step/sec: 1.37817 +INFO:tensorflow:step = 19901, loss = 0.666288, precision = 0.867188 (72.560 sec) +Saved checkpoint after 51 epoch(s) to data/resnet164/checkpoints/00051... +INFO:tensorflow:global_step/sec: 1.34178 +INFO:tensorflow:step = 20001, loss = 0.786395, precision = 0.859375 (74.528 sec) +INFO:tensorflow:global_step/sec: 1.37898 +INFO:tensorflow:step = 20101, loss = 0.693182, precision = 0.867188 (72.517 sec) +INFO:tensorflow:global_step/sec: 1.37843 +INFO:tensorflow:step = 20201, loss = 0.745275, precision = 0.820312 (72.547 sec) +INFO:tensorflow:global_step/sec: 1.3792 +INFO:tensorflow:step = 20301, loss = 0.547503, precision = 0.914062 (72.506 sec) +Saved checkpoint after 52 epoch(s) to data/resnet164/checkpoints/00052... +INFO:tensorflow:global_step/sec: 1.34192 +INFO:tensorflow:step = 20401, loss = 0.674193, precision = 0.84375 (74.520 sec) +INFO:tensorflow:global_step/sec: 1.37805 +INFO:tensorflow:step = 20501, loss = 0.727353, precision = 0.851562 (72.567 sec) +INFO:tensorflow:global_step/sec: 1.37921 +INFO:tensorflow:step = 20601, loss = 0.748478, precision = 0.859375 (72.505 sec) +INFO:tensorflow:global_step/sec: 1.37964 +INFO:tensorflow:step = 20701, loss = 0.634962, precision = 0.890625 (72.483 sec) +Saved checkpoint after 53 epoch(s) to data/resnet164/checkpoints/00053... +INFO:tensorflow:global_step/sec: 1.34103 +INFO:tensorflow:step = 20801, loss = 0.792848, precision = 0.828125 (74.570 sec) +INFO:tensorflow:global_step/sec: 1.3788 +INFO:tensorflow:step = 20901, loss = 0.634308, precision = 0.882812 (72.527 sec) +INFO:tensorflow:global_step/sec: 1.37857 +INFO:tensorflow:step = 21001, loss = 0.61348, precision = 0.921875 (72.539 sec) +INFO:tensorflow:global_step/sec: 1.37879 +INFO:tensorflow:step = 21101, loss = 0.694799, precision = 0.875 (72.527 sec) +Saved checkpoint after 54 epoch(s) to data/resnet164/checkpoints/00054... +INFO:tensorflow:global_step/sec: 1.34072 +INFO:tensorflow:step = 21201, loss = 0.624412, precision = 0.921875 (74.587 sec) +INFO:tensorflow:global_step/sec: 1.37896 +INFO:tensorflow:step = 21301, loss = 0.71406, precision = 0.867188 (72.518 sec) +INFO:tensorflow:global_step/sec: 1.37826 +INFO:tensorflow:step = 21401, loss = 0.654293, precision = 0.898438 (72.555 sec) +INFO:tensorflow:global_step/sec: 1.37875 +INFO:tensorflow:step = 21501, loss = 0.676674, precision = 0.890625 (72.530 sec) +Saved checkpoint after 55 epoch(s) to data/resnet164/checkpoints/00055... +INFO:tensorflow:global_step/sec: 1.34112 +INFO:tensorflow:step = 21601, loss = 0.656744, precision = 0.890625 (74.564 sec) +INFO:tensorflow:global_step/sec: 1.37893 +INFO:tensorflow:step = 21701, loss = 0.594223, precision = 0.90625 (72.520 sec) +INFO:tensorflow:global_step/sec: 1.37856 +INFO:tensorflow:step = 21801, loss = 0.721171, precision = 0.875 (72.540 sec) +Saved checkpoint after 56 epoch(s) to data/resnet164/checkpoints/00056... +INFO:tensorflow:global_step/sec: 1.34119 +INFO:tensorflow:step = 21901, loss = 0.6934, precision = 0.890625 (74.561 sec) +INFO:tensorflow:global_step/sec: 1.37933 +INFO:tensorflow:step = 22001, loss = 0.664285, precision = 0.882812 (72.499 sec) +INFO:tensorflow:global_step/sec: 1.37882 +INFO:tensorflow:step = 22101, loss = 0.617126, precision = 0.90625 (72.526 sec) +INFO:tensorflow:global_step/sec: 1.37881 +INFO:tensorflow:step = 22201, loss = 0.769066, precision = 0.8125 (72.526 sec) +Saved checkpoint after 57 epoch(s) to data/resnet164/checkpoints/00057... +INFO:tensorflow:global_step/sec: 1.34201 +INFO:tensorflow:step = 22301, loss = 0.624526, precision = 0.867188 (74.515 sec) +INFO:tensorflow:global_step/sec: 1.37919 +INFO:tensorflow:step = 22401, loss = 0.530488, precision = 0.945312 (72.506 sec) +INFO:tensorflow:global_step/sec: 1.37952 +INFO:tensorflow:step = 22501, loss = 0.640846, precision = 0.882812 (72.489 sec) +INFO:tensorflow:global_step/sec: 1.37877 +INFO:tensorflow:step = 22601, loss = 0.757148, precision = 0.859375 (72.528 sec) +Saved checkpoint after 58 epoch(s) to data/resnet164/checkpoints/00058... +INFO:tensorflow:global_step/sec: 1.34012 +INFO:tensorflow:step = 22701, loss = 0.655266, precision = 0.890625 (74.620 sec) +INFO:tensorflow:global_step/sec: 1.37741 +INFO:tensorflow:step = 22801, loss = 0.623724, precision = 0.914062 (72.600 sec) +INFO:tensorflow:global_step/sec: 1.37589 +INFO:tensorflow:step = 22901, loss = 0.666811, precision = 0.882812 (72.680 sec) +INFO:tensorflow:global_step/sec: 1.37552 +INFO:tensorflow:step = 23001, loss = 0.644588, precision = 0.890625 (72.699 sec) +Saved checkpoint after 59 epoch(s) to data/resnet164/checkpoints/00059... +INFO:tensorflow:global_step/sec: 1.33839 +INFO:tensorflow:step = 23101, loss = 0.740185, precision = 0.820312 (74.717 sec) +INFO:tensorflow:global_step/sec: 1.37661 +INFO:tensorflow:step = 23201, loss = 0.71926, precision = 0.851562 (72.642 sec) +INFO:tensorflow:global_step/sec: 1.37623 +INFO:tensorflow:step = 23301, loss = 0.746595, precision = 0.859375 (72.662 sec) +INFO:tensorflow:global_step/sec: 1.37612 +INFO:tensorflow:step = 23401, loss = 0.751183, precision = 0.820312 (72.668 sec) +Saved checkpoint after 60 epoch(s) to data/resnet164/checkpoints/00060... +INFO:tensorflow:global_step/sec: 1.33608 +INFO:tensorflow:step = 23501, loss = 0.594458, precision = 0.914062 (74.846 sec) +INFO:tensorflow:global_step/sec: 1.37631 +INFO:tensorflow:step = 23601, loss = 0.637207, precision = 0.90625 (72.658 sec) +INFO:tensorflow:global_step/sec: 1.37737 +INFO:tensorflow:step = 23701, loss = 0.77523, precision = 0.851562 (72.602 sec) +INFO:tensorflow:global_step/sec: 1.37716 +INFO:tensorflow:step = 23801, loss = 0.750571, precision = 0.875 (72.613 sec) +Saved checkpoint after 61 epoch(s) to data/resnet164/checkpoints/00061... +INFO:tensorflow:global_step/sec: 1.3402 +INFO:tensorflow:step = 23901, loss = 0.821146, precision = 0.828125 (74.616 sec) +INFO:tensorflow:global_step/sec: 1.37768 +INFO:tensorflow:step = 24001, loss = 0.583289, precision = 0.890625 (72.585 sec) +INFO:tensorflow:global_step/sec: 1.37775 +INFO:tensorflow:step = 24101, loss = 0.621442, precision = 0.898438 (72.582 sec) +INFO:tensorflow:global_step/sec: 1.37868 +INFO:tensorflow:step = 24201, loss = 0.661536, precision = 0.882812 (72.533 sec) +Saved checkpoint after 62 epoch(s) to data/resnet164/checkpoints/00062... +INFO:tensorflow:global_step/sec: 1.34093 +INFO:tensorflow:step = 24301, loss = 0.543802, precision = 0.945312 (74.575 sec) +INFO:tensorflow:global_step/sec: 1.37791 +INFO:tensorflow:step = 24401, loss = 0.690398, precision = 0.882812 (72.574 sec) +INFO:tensorflow:global_step/sec: 1.37831 +INFO:tensorflow:step = 24501, loss = 0.597685, precision = 0.90625 (72.552 sec) +INFO:tensorflow:global_step/sec: 1.37792 +INFO:tensorflow:step = 24601, loss = 0.630471, precision = 0.882812 (72.573 sec) +Saved checkpoint after 63 epoch(s) to data/resnet164/checkpoints/00063... +INFO:tensorflow:global_step/sec: 1.3402 +INFO:tensorflow:step = 24701, loss = 0.681635, precision = 0.867188 (74.616 sec) +INFO:tensorflow:global_step/sec: 1.37837 +INFO:tensorflow:step = 24801, loss = 0.677448, precision = 0.90625 (72.549 sec) +INFO:tensorflow:global_step/sec: 1.37824 +INFO:tensorflow:step = 24901, loss = 0.634026, precision = 0.921875 (72.556 sec) +INFO:tensorflow:global_step/sec: 1.37851 +INFO:tensorflow:step = 25001, loss = 0.711819, precision = 0.867188 (72.542 sec) +Saved checkpoint after 64 epoch(s) to data/resnet164/checkpoints/00064... +INFO:tensorflow:global_step/sec: 1.34078 +INFO:tensorflow:step = 25101, loss = 0.658353, precision = 0.898438 (74.584 sec) +INFO:tensorflow:global_step/sec: 1.37904 +INFO:tensorflow:step = 25201, loss = 0.599915, precision = 0.898438 (72.514 sec) +INFO:tensorflow:global_step/sec: 1.37831 +INFO:tensorflow:step = 25301, loss = 0.652107, precision = 0.914062 (72.553 sec) +INFO:tensorflow:global_step/sec: 1.37832 +INFO:tensorflow:step = 25401, loss = 0.698031, precision = 0.882812 (72.552 sec) +Saved checkpoint after 65 epoch(s) to data/resnet164/checkpoints/00065... +INFO:tensorflow:global_step/sec: 1.34066 +INFO:tensorflow:step = 25501, loss = 0.660261, precision = 0.875 (74.590 sec) +INFO:tensorflow:global_step/sec: 1.37908 +INFO:tensorflow:step = 25601, loss = 0.746789, precision = 0.890625 (72.512 sec) +INFO:tensorflow:global_step/sec: 1.37899 +INFO:tensorflow:step = 25701, loss = 0.787216, precision = 0.851562 (72.517 sec) +INFO:tensorflow:global_step/sec: 1.37867 +INFO:tensorflow:step = 25801, loss = 0.690413, precision = 0.851562 (72.534 sec) +Saved checkpoint after 66 epoch(s) to data/resnet164/checkpoints/00066... +INFO:tensorflow:global_step/sec: 1.34179 +INFO:tensorflow:step = 25901, loss = 0.659827, precision = 0.882812 (74.527 sec) +INFO:tensorflow:global_step/sec: 1.37884 +INFO:tensorflow:step = 26001, loss = 0.589913, precision = 0.929688 (72.525 sec) +INFO:tensorflow:global_step/sec: 1.37899 +INFO:tensorflow:step = 26101, loss = 0.69398, precision = 0.851562 (72.517 sec) +Saved checkpoint after 67 epoch(s) to data/resnet164/checkpoints/00067... +INFO:tensorflow:global_step/sec: 1.34173 +INFO:tensorflow:step = 26201, loss = 0.859913, precision = 0.84375 (74.531 sec) +INFO:tensorflow:global_step/sec: 1.37861 +INFO:tensorflow:step = 26301, loss = 0.629395, precision = 0.914062 (72.537 sec) +INFO:tensorflow:global_step/sec: 1.37814 +INFO:tensorflow:step = 26401, loss = 0.707004, precision = 0.859375 (72.561 sec) +INFO:tensorflow:global_step/sec: 1.37898 +INFO:tensorflow:step = 26501, loss = 0.68155, precision = 0.851562 (72.518 sec) +Saved checkpoint after 68 epoch(s) to data/resnet164/checkpoints/00068... +INFO:tensorflow:global_step/sec: 1.34025 +INFO:tensorflow:step = 26601, loss = 0.475779, precision = 0.96875 (74.613 sec) +INFO:tensorflow:global_step/sec: 1.37816 +INFO:tensorflow:step = 26701, loss = 0.648993, precision = 0.882812 (72.561 sec) +INFO:tensorflow:global_step/sec: 1.37893 +INFO:tensorflow:step = 26801, loss = 0.600669, precision = 0.890625 (72.520 sec) +INFO:tensorflow:global_step/sec: 1.3793 +INFO:tensorflow:step = 26901, loss = 0.650122, precision = 0.890625 (72.501 sec) +Saved checkpoint after 69 epoch(s) to data/resnet164/checkpoints/00069... +INFO:tensorflow:global_step/sec: 1.34109 +INFO:tensorflow:step = 27001, loss = 0.788485, precision = 0.851562 (74.566 sec) +INFO:tensorflow:global_step/sec: 1.37862 +INFO:tensorflow:step = 27101, loss = 0.655848, precision = 0.875 (72.536 sec) +INFO:tensorflow:global_step/sec: 1.3788 +INFO:tensorflow:step = 27201, loss = 0.626325, precision = 0.914062 (72.527 sec) +INFO:tensorflow:global_step/sec: 1.37861 +INFO:tensorflow:step = 27301, loss = 0.683111, precision = 0.890625 (72.537 sec) +Saved checkpoint after 70 epoch(s) to data/resnet164/checkpoints/00070... +INFO:tensorflow:global_step/sec: 1.33798 +INFO:tensorflow:step = 27401, loss = 0.594079, precision = 0.914062 (74.740 sec) +INFO:tensorflow:global_step/sec: 1.3792 +INFO:tensorflow:step = 27501, loss = 0.654749, precision = 0.890625 (72.506 sec) +INFO:tensorflow:global_step/sec: 1.37902 +INFO:tensorflow:step = 27601, loss = 0.744121, precision = 0.867188 (72.515 sec) +INFO:tensorflow:global_step/sec: 1.37857 +INFO:tensorflow:step = 27701, loss = 0.636206, precision = 0.929688 (72.539 sec) +Saved checkpoint after 71 epoch(s) to data/resnet164/checkpoints/00071... +INFO:tensorflow:global_step/sec: 1.34034 +INFO:tensorflow:step = 27801, loss = 0.593629, precision = 0.898438 (74.608 sec) +INFO:tensorflow:global_step/sec: 1.37908 +INFO:tensorflow:step = 27901, loss = 0.707738, precision = 0.875 (72.512 sec) +INFO:tensorflow:global_step/sec: 1.3791 +INFO:tensorflow:step = 28001, loss = 0.568895, precision = 0.929688 (72.511 sec) +INFO:tensorflow:global_step/sec: 1.37845 +INFO:tensorflow:step = 28101, loss = 0.819383, precision = 0.804688 (72.545 sec) +Saved checkpoint after 72 epoch(s) to data/resnet164/checkpoints/00072... +INFO:tensorflow:global_step/sec: 1.34061 +INFO:tensorflow:step = 28201, loss = 0.61783, precision = 0.921875 (74.593 sec) +INFO:tensorflow:global_step/sec: 1.37905 +INFO:tensorflow:step = 28301, loss = 0.59327, precision = 0.882812 (72.513 sec) +INFO:tensorflow:global_step/sec: 1.37913 +INFO:tensorflow:step = 28401, loss = 0.764563, precision = 0.828125 (72.510 sec) +INFO:tensorflow:global_step/sec: 1.37941 +INFO:tensorflow:step = 28501, loss = 0.84056, precision = 0.796875 (72.495 sec) +Saved checkpoint after 73 epoch(s) to data/resnet164/checkpoints/00073... +INFO:tensorflow:global_step/sec: 1.34015 +INFO:tensorflow:step = 28601, loss = 0.645253, precision = 0.90625 (74.618 sec) +INFO:tensorflow:global_step/sec: 1.37956 +INFO:tensorflow:step = 28701, loss = 0.68557, precision = 0.867188 (72.487 sec) +INFO:tensorflow:global_step/sec: 1.37872 +INFO:tensorflow:step = 28801, loss = 0.683242, precision = 0.914062 (72.531 sec) +INFO:tensorflow:global_step/sec: 1.37797 +INFO:tensorflow:step = 28901, loss = 0.714594, precision = 0.882812 (72.571 sec) +Saved checkpoint after 74 epoch(s) to data/resnet164/checkpoints/00074... +INFO:tensorflow:global_step/sec: 1.34194 +INFO:tensorflow:step = 29001, loss = 0.600449, precision = 0.882812 (74.519 sec) +INFO:tensorflow:global_step/sec: 1.37902 +INFO:tensorflow:step = 29101, loss = 0.649921, precision = 0.898438 (72.515 sec) +INFO:tensorflow:global_step/sec: 1.37868 +INFO:tensorflow:step = 29201, loss = 0.712819, precision = 0.882812 (72.533 sec) +INFO:tensorflow:global_step/sec: 1.37841 +INFO:tensorflow:step = 29301, loss = 0.76243, precision = 0.882812 (72.547 sec) +Saved checkpoint after 75 epoch(s) to data/resnet164/checkpoints/00075... +INFO:tensorflow:global_step/sec: 1.34206 +INFO:tensorflow:step = 29401, loss = 0.667819, precision = 0.914062 (74.513 sec) +INFO:tensorflow:global_step/sec: 1.37927 +INFO:tensorflow:step = 29501, loss = 0.7471, precision = 0.867188 (72.502 sec) +INFO:tensorflow:global_step/sec: 1.37894 +INFO:tensorflow:step = 29601, loss = 0.786828, precision = 0.835938 (72.519 sec) +INFO:tensorflow:global_step/sec: 1.37868 +INFO:tensorflow:step = 29701, loss = 0.723303, precision = 0.882812 (72.533 sec) +Saved checkpoint after 76 epoch(s) to data/resnet164/checkpoints/00076... +INFO:tensorflow:global_step/sec: 1.34219 +INFO:tensorflow:step = 29801, loss = 0.704925, precision = 0.875 (74.505 sec) +INFO:tensorflow:global_step/sec: 1.37897 +INFO:tensorflow:step = 29901, loss = 0.75156, precision = 0.851562 (72.518 sec) +INFO:tensorflow:global_step/sec: 1.37927 +INFO:tensorflow:step = 30001, loss = 0.615552, precision = 0.929688 (72.502 sec) +INFO:tensorflow:global_step/sec: 1.37899 +INFO:tensorflow:step = 30101, loss = 0.561019, precision = 0.929688 (72.517 sec) +Saved checkpoint after 77 epoch(s) to data/resnet164/checkpoints/00077... +INFO:tensorflow:global_step/sec: 1.34207 +INFO:tensorflow:step = 30201, loss = 0.747587, precision = 0.890625 (74.512 sec) +INFO:tensorflow:global_step/sec: 1.37868 +INFO:tensorflow:step = 30301, loss = 0.63427, precision = 0.890625 (72.533 sec) +INFO:tensorflow:global_step/sec: 1.37829 +INFO:tensorflow:step = 30401, loss = 0.719761, precision = 0.867188 (72.554 sec) +Saved checkpoint after 78 epoch(s) to data/resnet164/checkpoints/00078... +INFO:tensorflow:global_step/sec: 1.34134 +INFO:tensorflow:step = 30501, loss = 0.592332, precision = 0.921875 (74.553 sec) +INFO:tensorflow:global_step/sec: 1.37911 +INFO:tensorflow:step = 30601, loss = 0.620759, precision = 0.898438 (72.511 sec) +INFO:tensorflow:global_step/sec: 1.37959 +INFO:tensorflow:step = 30701, loss = 0.689734, precision = 0.859375 (72.486 sec) +INFO:tensorflow:global_step/sec: 1.37944 +INFO:tensorflow:step = 30801, loss = 0.592504, precision = 0.90625 (72.493 sec) +Saved checkpoint after 79 epoch(s) to data/resnet164/checkpoints/00079... +INFO:tensorflow:global_step/sec: 1.34181 +INFO:tensorflow:step = 30901, loss = 0.606076, precision = 0.890625 (74.526 sec) +INFO:tensorflow:global_step/sec: 1.37872 +INFO:tensorflow:step = 31001, loss = 0.595124, precision = 0.914062 (72.531 sec) +INFO:tensorflow:global_step/sec: 1.37856 +INFO:tensorflow:step = 31101, loss = 0.633624, precision = 0.882812 (72.539 sec) +INFO:tensorflow:global_step/sec: 1.37836 +INFO:tensorflow:step = 31201, loss = 0.751532, precision = 0.875 (72.550 sec) +Saved checkpoint after 80 epoch(s) to data/resnet164/checkpoints/00080... +INFO:tensorflow:global_step/sec: 1.33757 +INFO:tensorflow:step = 31301, loss = 0.531982, precision = 0.953125 (74.762 sec) +INFO:tensorflow:global_step/sec: 1.37934 +INFO:tensorflow:step = 31401, loss = 0.658886, precision = 0.890625 (72.498 sec) +INFO:tensorflow:global_step/sec: 1.37892 +INFO:tensorflow:step = 31501, loss = 0.698325, precision = 0.867188 (72.521 sec) +INFO:tensorflow:global_step/sec: 1.37849 +INFO:tensorflow:step = 31601, loss = 0.560457, precision = 0.921875 (72.543 sec) +Saved checkpoint after 81 epoch(s) to data/resnet164/checkpoints/00081... +INFO:tensorflow:global_step/sec: 1.34051 +INFO:tensorflow:step = 31701, loss = 0.705335, precision = 0.875 (74.598 sec) +INFO:tensorflow:global_step/sec: 1.37935 +INFO:tensorflow:step = 31801, loss = 0.580626, precision = 0.898438 (72.498 sec) +INFO:tensorflow:global_step/sec: 1.37888 +INFO:tensorflow:step = 31901, loss = 0.615262, precision = 0.882812 (72.523 sec) +INFO:tensorflow:global_step/sec: 1.37848 +INFO:tensorflow:step = 32001, loss = 0.720714, precision = 0.828125 (72.544 sec) +Saved checkpoint after 82 epoch(s) to data/resnet164/checkpoints/00082... +INFO:tensorflow:global_step/sec: 1.34147 +INFO:tensorflow:step = 32101, loss = 0.642799, precision = 0.882812 (74.546 sec) +INFO:tensorflow:global_step/sec: 1.37897 +INFO:tensorflow:step = 32201, loss = 0.613026, precision = 0.898438 (72.517 sec) +INFO:tensorflow:global_step/sec: 1.37856 +INFO:tensorflow:step = 32301, loss = 0.631813, precision = 0.929688 (72.540 sec) +INFO:tensorflow:global_step/sec: 1.37887 +INFO:tensorflow:step = 32401, loss = 0.69618, precision = 0.882812 (72.523 sec) +Saved checkpoint after 83 epoch(s) to data/resnet164/checkpoints/00083... +INFO:tensorflow:global_step/sec: 1.34204 +INFO:tensorflow:step = 32501, loss = 0.570752, precision = 0.9375 (74.513 sec) +INFO:tensorflow:global_step/sec: 1.37905 +INFO:tensorflow:step = 32601, loss = 0.638179, precision = 0.914062 (72.514 sec) +INFO:tensorflow:global_step/sec: 1.37872 +INFO:tensorflow:step = 32701, loss = 0.567763, precision = 0.90625 (72.531 sec) +INFO:tensorflow:global_step/sec: 1.37925 +INFO:tensorflow:step = 32801, loss = 0.584759, precision = 0.90625 (72.503 sec) +Saved checkpoint after 84 epoch(s) to data/resnet164/checkpoints/00084... +INFO:tensorflow:global_step/sec: 1.34119 +INFO:tensorflow:step = 32901, loss = 0.562019, precision = 0.914062 (74.561 sec) +INFO:tensorflow:global_step/sec: 1.37842 +INFO:tensorflow:step = 33001, loss = 0.644854, precision = 0.921875 (72.547 sec) +INFO:tensorflow:global_step/sec: 1.37892 +INFO:tensorflow:step = 33101, loss = 0.555516, precision = 0.9375 (72.521 sec) +INFO:tensorflow:global_step/sec: 1.37763 +INFO:tensorflow:step = 33201, loss = 0.643414, precision = 0.921875 (72.588 sec) +Saved checkpoint after 85 epoch(s) to data/resnet164/checkpoints/00085... +INFO:tensorflow:global_step/sec: 1.34189 +INFO:tensorflow:step = 33301, loss = 0.752197, precision = 0.851562 (74.522 sec) +INFO:tensorflow:global_step/sec: 1.37831 +INFO:tensorflow:step = 33401, loss = 0.628264, precision = 0.914062 (72.552 sec) +INFO:tensorflow:global_step/sec: 1.37791 +INFO:tensorflow:step = 33501, loss = 0.575388, precision = 0.90625 (72.574 sec) +INFO:tensorflow:global_step/sec: 1.37873 +INFO:tensorflow:step = 33601, loss = 0.607844, precision = 0.898438 (72.530 sec) +Saved checkpoint after 86 epoch(s) to data/resnet164/checkpoints/00086... +INFO:tensorflow:global_step/sec: 1.341 +INFO:tensorflow:step = 33701, loss = 0.598503, precision = 0.90625 (74.571 sec) +INFO:tensorflow:global_step/sec: 1.37835 +INFO:tensorflow:step = 33801, loss = 0.576267, precision = 0.921875 (72.550 sec) +INFO:tensorflow:global_step/sec: 1.37762 +INFO:tensorflow:step = 33901, loss = 0.778203, precision = 0.890625 (72.589 sec) +INFO:tensorflow:global_step/sec: 1.37833 +INFO:tensorflow:step = 34001, loss = 0.65027, precision = 0.898438 (72.552 sec) +Saved checkpoint after 87 epoch(s) to data/resnet164/checkpoints/00087... +INFO:tensorflow:global_step/sec: 1.34204 +INFO:tensorflow:step = 34101, loss = 0.634607, precision = 0.890625 (74.514 sec) +INFO:tensorflow:global_step/sec: 1.3786 +INFO:tensorflow:step = 34201, loss = 0.770081, precision = 0.851562 (72.537 sec) +INFO:tensorflow:global_step/sec: 1.37882 +INFO:tensorflow:step = 34301, loss = 0.552332, precision = 0.921875 (72.526 sec) +INFO:tensorflow:global_step/sec: 1.37955 +INFO:tensorflow:step = 34401, loss = 0.660209, precision = 0.867188 (72.487 sec) +Saved checkpoint after 88 epoch(s) to data/resnet164/checkpoints/00088... +INFO:tensorflow:global_step/sec: 1.34098 +INFO:tensorflow:step = 34501, loss = 0.656844, precision = 0.890625 (74.572 sec) +INFO:tensorflow:global_step/sec: 1.3781 +INFO:tensorflow:step = 34601, loss = 0.681649, precision = 0.875 (72.564 sec) +INFO:tensorflow:global_step/sec: 1.37837 +INFO:tensorflow:step = 34701, loss = 0.580966, precision = 0.921875 (72.550 sec) +Saved checkpoint after 89 epoch(s) to data/resnet164/checkpoints/00089... +INFO:tensorflow:global_step/sec: 1.341 +INFO:tensorflow:step = 34801, loss = 0.532375, precision = 0.945312 (74.571 sec) +INFO:tensorflow:global_step/sec: 1.37847 +INFO:tensorflow:step = 34901, loss = 0.703392, precision = 0.890625 (72.544 sec) +INFO:tensorflow:global_step/sec: 1.37909 +INFO:tensorflow:step = 35001, loss = 0.690022, precision = 0.875 (72.511 sec) +INFO:tensorflow:global_step/sec: 1.37917 +INFO:tensorflow:step = 35101, loss = 0.643245, precision = 0.90625 (72.507 sec) +Saved checkpoint after 90 epoch(s) to data/resnet164/checkpoints/00090... +INFO:tensorflow:global_step/sec: 1.34193 +INFO:tensorflow:step = 35201, loss = 0.654287, precision = 0.890625 (74.519 sec) +INFO:tensorflow:global_step/sec: 1.37942 +INFO:tensorflow:step = 35301, loss = 0.589608, precision = 0.921875 (72.494 sec) +INFO:tensorflow:global_step/sec: 1.37913 +INFO:tensorflow:step = 35401, loss = 0.663402, precision = 0.882812 (72.509 sec) +INFO:tensorflow:global_step/sec: 1.37898 +INFO:tensorflow:step = 35501, loss = 0.74295, precision = 0.867188 (72.517 sec) +Saved checkpoint after 91 epoch(s) to data/resnet164/checkpoints/00091... +INFO:tensorflow:global_step/sec: 1.33848 +INFO:tensorflow:step = 35601, loss = 0.619732, precision = 0.898438 (74.711 sec) +INFO:tensorflow:global_step/sec: 1.37928 +INFO:tensorflow:step = 35701, loss = 0.481555, precision = 0.953125 (72.502 sec) +INFO:tensorflow:global_step/sec: 1.3792 +INFO:tensorflow:step = 35801, loss = 0.423316, precision = 0.960938 (72.506 sec) +INFO:tensorflow:global_step/sec: 1.37949 +INFO:tensorflow:step = 35901, loss = 0.467324, precision = 0.953125 (72.491 sec) +Saved checkpoint after 92 epoch(s) to data/resnet164/checkpoints/00092... +INFO:tensorflow:global_step/sec: 1.34116 +INFO:tensorflow:step = 36001, loss = 0.445567, precision = 0.96875 (74.562 sec) +INFO:tensorflow:global_step/sec: 1.37934 +INFO:tensorflow:step = 36101, loss = 0.423769, precision = 0.976562 (72.498 sec) +INFO:tensorflow:global_step/sec: 1.37926 +INFO:tensorflow:step = 36201, loss = 0.426964, precision = 0.96875 (72.503 sec) +INFO:tensorflow:global_step/sec: 1.3784 +INFO:tensorflow:step = 36301, loss = 0.439296, precision = 0.953125 (72.548 sec) +Saved checkpoint after 93 epoch(s) to data/resnet164/checkpoints/00093... +INFO:tensorflow:global_step/sec: 1.342 +INFO:tensorflow:step = 36401, loss = 0.462412, precision = 0.960938 (74.516 sec) +INFO:tensorflow:global_step/sec: 1.37952 +INFO:tensorflow:step = 36501, loss = 0.467238, precision = 0.929688 (72.489 sec) +INFO:tensorflow:global_step/sec: 1.37921 +INFO:tensorflow:step = 36601, loss = 0.426008, precision = 0.976562 (72.505 sec) +INFO:tensorflow:global_step/sec: 1.37938 +INFO:tensorflow:step = 36701, loss = 0.475676, precision = 0.9375 (72.496 sec) +Saved checkpoint after 94 epoch(s) to data/resnet164/checkpoints/00094... +INFO:tensorflow:global_step/sec: 1.3414 +INFO:tensorflow:step = 36801, loss = 0.40885, precision = 0.96875 (74.549 sec) +INFO:tensorflow:global_step/sec: 1.37946 +INFO:tensorflow:step = 36901, loss = 0.365601, precision = 0.984375 (72.492 sec) +INFO:tensorflow:global_step/sec: 1.379 +INFO:tensorflow:step = 37001, loss = 0.429193, precision = 0.9375 (72.516 sec) +INFO:tensorflow:global_step/sec: 1.37987 +INFO:tensorflow:step = 37101, loss = 0.382502, precision = 0.96875 (72.471 sec) +Saved checkpoint after 95 epoch(s) to data/resnet164/checkpoints/00095... +INFO:tensorflow:global_step/sec: 1.34177 +INFO:tensorflow:step = 37201, loss = 0.382352, precision = 0.953125 (74.528 sec) +INFO:tensorflow:global_step/sec: 1.3792 +INFO:tensorflow:step = 37301, loss = 0.355848, precision = 0.976562 (72.506 sec) +INFO:tensorflow:global_step/sec: 1.37915 +INFO:tensorflow:step = 37401, loss = 0.354615, precision = 0.976562 (72.508 sec) +INFO:tensorflow:global_step/sec: 1.37922 +INFO:tensorflow:step = 37501, loss = 0.364346, precision = 0.984375 (72.505 sec) +Saved checkpoint after 96 epoch(s) to data/resnet164/checkpoints/00096... +INFO:tensorflow:global_step/sec: 1.34153 +INFO:tensorflow:step = 37601, loss = 0.372782, precision = 0.976562 (74.542 sec) +INFO:tensorflow:global_step/sec: 1.37936 +INFO:tensorflow:step = 37701, loss = 0.332243, precision = 0.992188 (72.497 sec) +INFO:tensorflow:global_step/sec: 1.3792 +INFO:tensorflow:step = 37801, loss = 0.355183, precision = 0.976562 (72.506 sec) +INFO:tensorflow:global_step/sec: 1.37903 +INFO:tensorflow:step = 37901, loss = 0.325782, precision = 0.984375 (72.515 sec) +Saved checkpoint after 97 epoch(s) to data/resnet164/checkpoints/00097... +INFO:tensorflow:global_step/sec: 1.34176 +INFO:tensorflow:step = 38001, loss = 0.361765, precision = 0.96875 (74.529 sec) +INFO:tensorflow:global_step/sec: 1.37848 +INFO:tensorflow:step = 38101, loss = 0.341869, precision = 0.984375 (72.543 sec) +INFO:tensorflow:global_step/sec: 1.37944 +INFO:tensorflow:step = 38201, loss = 0.367413, precision = 0.984375 (72.493 sec) +INFO:tensorflow:global_step/sec: 1.37857 +INFO:tensorflow:step = 38301, loss = 0.346137, precision = 0.984375 (72.539 sec) +Saved checkpoint after 98 epoch(s) to data/resnet164/checkpoints/00098... +INFO:tensorflow:global_step/sec: 1.34222 +INFO:tensorflow:step = 38401, loss = 0.333586, precision = 0.96875 (74.503 sec) +INFO:tensorflow:global_step/sec: 1.37917 +INFO:tensorflow:step = 38501, loss = 0.312043, precision = 0.984375 (72.507 sec) +INFO:tensorflow:global_step/sec: 1.37875 +INFO:tensorflow:step = 38601, loss = 0.372894, precision = 0.960938 (72.530 sec) +INFO:tensorflow:global_step/sec: 1.37867 +INFO:tensorflow:step = 38701, loss = 0.320672, precision = 0.976562 (72.534 sec) +Saved checkpoint after 99 epoch(s) to data/resnet164/checkpoints/00099... +INFO:tensorflow:global_step/sec: 1.34141 +INFO:tensorflow:step = 38801, loss = 0.317587, precision = 0.96875 (74.548 sec) +INFO:tensorflow:global_step/sec: 1.37926 +INFO:tensorflow:step = 38901, loss = 0.297556, precision = 0.984375 (72.503 sec) +INFO:tensorflow:global_step/sec: 1.37953 +INFO:tensorflow:step = 39001, loss = 0.317672, precision = 0.976562 (72.488 sec) +Saved checkpoint after 100 epoch(s) to data/resnet164/checkpoints/00100... +INFO:tensorflow:global_step/sec: 1.34231 +INFO:tensorflow:step = 39101, loss = 0.310726, precision = 0.992188 (74.498 sec) +INFO:tensorflow:global_step/sec: 1.3803 +INFO:tensorflow:step = 39201, loss = 0.367321, precision = 0.945312 (72.448 sec) +INFO:tensorflow:global_step/sec: 1.38048 +INFO:tensorflow:step = 39301, loss = 0.323489, precision = 0.96875 (72.438 sec) +INFO:tensorflow:global_step/sec: 1.38037 +INFO:tensorflow:step = 39401, loss = 0.357481, precision = 0.945312 (72.444 sec) +Saved checkpoint after 101 epoch(s) to data/resnet164/checkpoints/00101... +INFO:tensorflow:global_step/sec: 1.33845 +INFO:tensorflow:step = 39501, loss = 0.372797, precision = 0.960938 (74.713 sec) +INFO:tensorflow:global_step/sec: 1.38064 +INFO:tensorflow:step = 39601, loss = 0.294106, precision = 0.984375 (72.430 sec) +INFO:tensorflow:global_step/sec: 1.38065 +INFO:tensorflow:step = 39701, loss = 0.376446, precision = 0.953125 (72.430 sec) +INFO:tensorflow:global_step/sec: 1.3807 +INFO:tensorflow:step = 39801, loss = 0.284816, precision = 0.992188 (72.427 sec) +Saved checkpoint after 102 epoch(s) to data/resnet164/checkpoints/00102... +INFO:tensorflow:global_step/sec: 1.3428 +INFO:tensorflow:step = 39901, loss = 0.289231, precision = 0.976562 (74.472 sec) +INFO:tensorflow:global_step/sec: 1.3803 +INFO:tensorflow:step = 40001, loss = 0.270328, precision = 0.992188 (72.448 sec) +INFO:tensorflow:global_step/sec: 1.37981 +INFO:tensorflow:step = 40101, loss = 0.34249, precision = 0.953125 (72.474 sec) +INFO:tensorflow:global_step/sec: 1.38062 +INFO:tensorflow:step = 40201, loss = 0.305579, precision = 0.976562 (72.431 sec) +Saved checkpoint after 103 epoch(s) to data/resnet164/checkpoints/00103... +INFO:tensorflow:global_step/sec: 1.34176 +INFO:tensorflow:step = 40301, loss = 0.254434, precision = 1.0 (74.529 sec) +INFO:tensorflow:global_step/sec: 1.38042 +INFO:tensorflow:step = 40401, loss = 0.270511, precision = 0.992188 (72.441 sec) +INFO:tensorflow:global_step/sec: 1.38022 +INFO:tensorflow:step = 40501, loss = 0.284316, precision = 0.984375 (72.453 sec) +INFO:tensorflow:global_step/sec: 1.38053 +INFO:tensorflow:step = 40601, loss = 0.275436, precision = 0.96875 (72.436 sec) +Saved checkpoint after 104 epoch(s) to data/resnet164/checkpoints/00104... +INFO:tensorflow:global_step/sec: 1.34349 +INFO:tensorflow:step = 40701, loss = 0.261646, precision = 0.992188 (74.433 sec) +INFO:tensorflow:global_step/sec: 1.38007 +INFO:tensorflow:step = 40801, loss = 0.257787, precision = 0.992188 (72.460 sec) +INFO:tensorflow:global_step/sec: 1.38019 +INFO:tensorflow:step = 40901, loss = 0.256961, precision = 0.992188 (72.454 sec) +INFO:tensorflow:global_step/sec: 1.38057 +INFO:tensorflow:step = 41001, loss = 0.302682, precision = 0.96875 (72.434 sec) +Saved checkpoint after 105 epoch(s) to data/resnet164/checkpoints/00105... +INFO:tensorflow:global_step/sec: 1.34394 +INFO:tensorflow:step = 41101, loss = 0.261003, precision = 0.984375 (74.408 sec) +INFO:tensorflow:global_step/sec: 1.38062 +INFO:tensorflow:step = 41201, loss = 0.255877, precision = 0.992188 (72.431 sec) +INFO:tensorflow:global_step/sec: 1.38027 +INFO:tensorflow:step = 41301, loss = 0.261708, precision = 0.976562 (72.450 sec) +INFO:tensorflow:global_step/sec: 1.38096 +INFO:tensorflow:step = 41401, loss = 0.286167, precision = 0.976562 (72.413 sec) +Saved checkpoint after 106 epoch(s) to data/resnet164/checkpoints/00106... +INFO:tensorflow:global_step/sec: 1.34311 +INFO:tensorflow:step = 41501, loss = 0.264997, precision = 0.992188 (74.454 sec) +INFO:tensorflow:global_step/sec: 1.38071 +INFO:tensorflow:step = 41601, loss = 0.226899, precision = 1.0 (72.426 sec) +INFO:tensorflow:global_step/sec: 1.38131 +INFO:tensorflow:step = 41701, loss = 0.310962, precision = 0.960938 (72.395 sec) +INFO:tensorflow:global_step/sec: 1.38023 +INFO:tensorflow:step = 41801, loss = 0.247084, precision = 0.984375 (72.452 sec) +Saved checkpoint after 107 epoch(s) to data/resnet164/checkpoints/00107... +INFO:tensorflow:global_step/sec: 1.34249 +INFO:tensorflow:step = 41901, loss = 0.308377, precision = 0.953125 (74.488 sec) +INFO:tensorflow:global_step/sec: 1.38073 +INFO:tensorflow:step = 42001, loss = 0.276382, precision = 0.976562 (72.425 sec) +INFO:tensorflow:global_step/sec: 1.38003 +INFO:tensorflow:step = 42101, loss = 0.251112, precision = 0.976562 (72.462 sec) +INFO:tensorflow:global_step/sec: 1.3809 +INFO:tensorflow:step = 42201, loss = 0.235442, precision = 0.992188 (72.417 sec) +Saved checkpoint after 108 epoch(s) to data/resnet164/checkpoints/00108... +INFO:tensorflow:global_step/sec: 1.34355 +INFO:tensorflow:step = 42301, loss = 0.231923, precision = 0.992188 (74.430 sec) +INFO:tensorflow:global_step/sec: 1.38113 +INFO:tensorflow:step = 42401, loss = 0.246266, precision = 0.992188 (72.404 sec) +INFO:tensorflow:global_step/sec: 1.38007 +INFO:tensorflow:step = 42501, loss = 0.243915, precision = 0.992188 (72.460 sec) +INFO:tensorflow:global_step/sec: 1.38037 +INFO:tensorflow:step = 42601, loss = 0.216456, precision = 0.992188 (72.444 sec) +Saved checkpoint after 109 epoch(s) to data/resnet164/checkpoints/00109... +INFO:tensorflow:global_step/sec: 1.34328 +INFO:tensorflow:step = 42701, loss = 0.236366, precision = 0.992188 (74.445 sec) +INFO:tensorflow:global_step/sec: 1.38005 +INFO:tensorflow:step = 42801, loss = 0.213624, precision = 1.0 (72.461 sec) +INFO:tensorflow:global_step/sec: 1.37969 +INFO:tensorflow:step = 42901, loss = 0.272287, precision = 0.96875 (72.480 sec) +INFO:tensorflow:global_step/sec: 1.37945 +INFO:tensorflow:step = 43001, loss = 0.29783, precision = 0.96875 (72.493 sec) +Saved checkpoint after 110 epoch(s) to data/resnet164/checkpoints/00110... +INFO:tensorflow:global_step/sec: 1.34253 +INFO:tensorflow:step = 43101, loss = 0.252152, precision = 0.96875 (74.486 sec) +INFO:tensorflow:global_step/sec: 1.38049 +INFO:tensorflow:step = 43201, loss = 0.231989, precision = 0.992188 (72.438 sec) +INFO:tensorflow:global_step/sec: 1.38001 +INFO:tensorflow:step = 43301, loss = 0.261933, precision = 0.96875 (72.463 sec) +Saved checkpoint after 111 epoch(s) to data/resnet164/checkpoints/00111... +INFO:tensorflow:global_step/sec: 1.33917 +INFO:tensorflow:step = 43401, loss = 0.207446, precision = 1.0 (74.673 sec) +INFO:tensorflow:global_step/sec: 1.38001 +INFO:tensorflow:step = 43501, loss = 0.20669, precision = 1.0 (72.463 sec) +INFO:tensorflow:global_step/sec: 1.38007 +INFO:tensorflow:step = 43601, loss = 0.203212, precision = 1.0 (72.460 sec) +INFO:tensorflow:global_step/sec: 1.38061 +INFO:tensorflow:step = 43701, loss = 0.250676, precision = 0.976562 (72.432 sec) +Saved checkpoint after 112 epoch(s) to data/resnet164/checkpoints/00112... +INFO:tensorflow:global_step/sec: 1.34304 +INFO:tensorflow:step = 43801, loss = 0.218182, precision = 1.0 (74.458 sec) +INFO:tensorflow:global_step/sec: 1.38047 +INFO:tensorflow:step = 43901, loss = 0.25003, precision = 0.976562 (72.439 sec) +INFO:tensorflow:global_step/sec: 1.37927 +INFO:tensorflow:step = 44001, loss = 0.238393, precision = 0.976562 (72.502 sec) +INFO:tensorflow:global_step/sec: 1.38053 +INFO:tensorflow:step = 44101, loss = 0.279076, precision = 0.953125 (72.436 sec) +Saved checkpoint after 113 epoch(s) to data/resnet164/checkpoints/00113... +INFO:tensorflow:global_step/sec: 1.34179 +INFO:tensorflow:step = 44201, loss = 0.203329, precision = 0.992188 (74.527 sec) +INFO:tensorflow:global_step/sec: 1.37958 +INFO:tensorflow:step = 44301, loss = 0.219183, precision = 0.984375 (72.486 sec) +INFO:tensorflow:global_step/sec: 1.37886 +INFO:tensorflow:step = 44401, loss = 0.19264, precision = 1.0 (72.524 sec) +INFO:tensorflow:global_step/sec: 1.37959 +INFO:tensorflow:step = 44501, loss = 0.245206, precision = 0.960938 (72.485 sec) +Saved checkpoint after 114 epoch(s) to data/resnet164/checkpoints/00114... +INFO:tensorflow:global_step/sec: 1.34288 +INFO:tensorflow:step = 44601, loss = 0.214979, precision = 0.984375 (74.467 sec) +INFO:tensorflow:global_step/sec: 1.38031 +INFO:tensorflow:step = 44701, loss = 0.246489, precision = 0.976562 (72.448 sec) +INFO:tensorflow:global_step/sec: 1.37935 +INFO:tensorflow:step = 44801, loss = 0.238816, precision = 0.992188 (72.498 sec) +INFO:tensorflow:global_step/sec: 1.37983 +INFO:tensorflow:step = 44901, loss = 0.216306, precision = 0.984375 (72.473 sec) +Saved checkpoint after 115 epoch(s) to data/resnet164/checkpoints/00115... +INFO:tensorflow:global_step/sec: 1.34204 +INFO:tensorflow:step = 45001, loss = 0.231852, precision = 0.992188 (74.514 sec) +INFO:tensorflow:global_step/sec: 1.37984 +INFO:tensorflow:step = 45101, loss = 0.21096, precision = 0.992188 (72.472 sec) +INFO:tensorflow:global_step/sec: 1.37939 +INFO:tensorflow:step = 45201, loss = 0.203829, precision = 0.992188 (72.496 sec) +INFO:tensorflow:global_step/sec: 1.37977 +INFO:tensorflow:step = 45301, loss = 0.211305, precision = 0.984375 (72.476 sec) +Saved checkpoint after 116 epoch(s) to data/resnet164/checkpoints/00116... +INFO:tensorflow:global_step/sec: 1.34231 +INFO:tensorflow:step = 45401, loss = 0.275882, precision = 0.976562 (74.499 sec) +INFO:tensorflow:global_step/sec: 1.37971 +INFO:tensorflow:step = 45501, loss = 0.225356, precision = 0.976562 (72.479 sec) +INFO:tensorflow:global_step/sec: 1.37993 +INFO:tensorflow:step = 45601, loss = 0.207037, precision = 0.992188 (72.468 sec) +INFO:tensorflow:global_step/sec: 1.37983 +INFO:tensorflow:step = 45701, loss = 0.221635, precision = 1.0 (72.473 sec) +Saved checkpoint after 117 epoch(s) to data/resnet164/checkpoints/00117... +INFO:tensorflow:global_step/sec: 1.34255 +INFO:tensorflow:step = 45801, loss = 0.194522, precision = 1.0 (74.485 sec) +INFO:tensorflow:global_step/sec: 1.37912 +INFO:tensorflow:step = 45901, loss = 0.219012, precision = 0.976562 (72.510 sec) +INFO:tensorflow:global_step/sec: 1.37862 +INFO:tensorflow:step = 46001, loss = 0.272174, precision = 0.960938 (72.536 sec) +INFO:tensorflow:global_step/sec: 1.37969 +INFO:tensorflow:step = 46101, loss = 0.208349, precision = 0.984375 (72.480 sec) +Saved checkpoint after 118 epoch(s) to data/resnet164/checkpoints/00118... +INFO:tensorflow:global_step/sec: 1.34313 +INFO:tensorflow:step = 46201, loss = 0.206592, precision = 0.992188 (74.453 sec) +INFO:tensorflow:global_step/sec: 1.38013 +INFO:tensorflow:step = 46301, loss = 0.261781, precision = 0.96875 (72.457 sec) +INFO:tensorflow:global_step/sec: 1.38045 +INFO:tensorflow:step = 46401, loss = 0.18767, precision = 1.0 (72.440 sec) +INFO:tensorflow:global_step/sec: 1.37982 +INFO:tensorflow:step = 46501, loss = 0.184939, precision = 1.0 (72.473 sec) +Saved checkpoint after 119 epoch(s) to data/resnet164/checkpoints/00119... +INFO:tensorflow:global_step/sec: 1.34303 +INFO:tensorflow:step = 46601, loss = 0.264502, precision = 0.960938 (74.458 sec) +INFO:tensorflow:global_step/sec: 1.37954 +INFO:tensorflow:step = 46701, loss = 0.202994, precision = 0.992188 (72.488 sec) +INFO:tensorflow:global_step/sec: 1.38008 +INFO:tensorflow:step = 46801, loss = 0.184741, precision = 0.992188 (72.460 sec) +INFO:tensorflow:global_step/sec: 1.37987 +INFO:tensorflow:step = 46901, loss = 0.236777, precision = 0.976562 (72.471 sec) +Saved checkpoint after 120 epoch(s) to data/resnet164/checkpoints/00120... +INFO:tensorflow:global_step/sec: 1.34146 +INFO:tensorflow:step = 47001, loss = 0.240601, precision = 0.960938 (74.546 sec) +INFO:tensorflow:global_step/sec: 1.38018 +INFO:tensorflow:step = 47101, loss = 0.261198, precision = 0.96875 (72.454 sec) +INFO:tensorflow:global_step/sec: 1.37958 +INFO:tensorflow:step = 47201, loss = 0.22062, precision = 0.976562 (72.486 sec) +INFO:tensorflow:global_step/sec: 1.37955 +INFO:tensorflow:step = 47301, loss = 0.200534, precision = 0.976562 (72.488 sec) +Saved checkpoint after 121 epoch(s) to data/resnet164/checkpoints/00121... +INFO:tensorflow:global_step/sec: 1.34253 +INFO:tensorflow:step = 47401, loss = 0.185214, precision = 1.0 (74.486 sec) +INFO:tensorflow:global_step/sec: 1.38012 +INFO:tensorflow:step = 47501, loss = 0.221454, precision = 0.96875 (72.457 sec) +INFO:tensorflow:global_step/sec: 1.3799 +INFO:tensorflow:step = 47601, loss = 0.173938, precision = 1.0 (72.469 sec) +INFO:tensorflow:global_step/sec: 1.38032 +INFO:tensorflow:step = 47701, loss = 0.198441, precision = 0.992188 (72.447 sec) +Saved checkpoint after 122 epoch(s) to data/resnet164/checkpoints/00122... +INFO:tensorflow:global_step/sec: 1.34008 +INFO:tensorflow:step = 47801, loss = 0.213993, precision = 0.976562 (74.622 sec) +INFO:tensorflow:global_step/sec: 1.37989 +INFO:tensorflow:step = 47901, loss = 0.217169, precision = 0.96875 (72.469 sec) +INFO:tensorflow:global_step/sec: 1.38053 +INFO:tensorflow:step = 48001, loss = 0.18865, precision = 0.992188 (72.436 sec) +Saved checkpoint after 123 epoch(s) to data/resnet164/checkpoints/00123... +INFO:tensorflow:global_step/sec: 1.34134 +INFO:tensorflow:step = 48101, loss = 0.255777, precision = 0.96875 (74.552 sec) +INFO:tensorflow:global_step/sec: 1.38063 +INFO:tensorflow:step = 48201, loss = 0.192308, precision = 0.992188 (72.431 sec) +INFO:tensorflow:global_step/sec: 1.37984 +INFO:tensorflow:step = 48301, loss = 0.224998, precision = 0.976562 (72.472 sec) +INFO:tensorflow:global_step/sec: 1.37987 +INFO:tensorflow:step = 48401, loss = 0.198497, precision = 0.976562 (72.471 sec) +Saved checkpoint after 124 epoch(s) to data/resnet164/checkpoints/00124... +INFO:tensorflow:global_step/sec: 1.34178 +INFO:tensorflow:step = 48501, loss = 0.213281, precision = 0.984375 (74.528 sec) +INFO:tensorflow:global_step/sec: 1.38012 +INFO:tensorflow:step = 48601, loss = 0.207232, precision = 0.96875 (72.457 sec) +INFO:tensorflow:global_step/sec: 1.37983 +INFO:tensorflow:step = 48701, loss = 0.235287, precision = 0.96875 (72.473 sec) +INFO:tensorflow:global_step/sec: 1.38005 +INFO:tensorflow:step = 48801, loss = 0.184656, precision = 0.992188 (72.461 sec) +Saved checkpoint after 125 epoch(s) to data/resnet164/checkpoints/00125... +INFO:tensorflow:global_step/sec: 1.34249 +INFO:tensorflow:step = 48901, loss = 0.174493, precision = 1.0 (74.489 sec) +INFO:tensorflow:global_step/sec: 1.38012 +INFO:tensorflow:step = 49001, loss = 0.233085, precision = 0.976562 (72.457 sec) +INFO:tensorflow:global_step/sec: 1.37992 +INFO:tensorflow:step = 49101, loss = 0.214592, precision = 0.976562 (72.468 sec) +INFO:tensorflow:global_step/sec: 1.3798 +INFO:tensorflow:step = 49201, loss = 0.223424, precision = 0.96875 (72.474 sec) +Saved checkpoint after 126 epoch(s) to data/resnet164/checkpoints/00126... +INFO:tensorflow:global_step/sec: 1.34253 +INFO:tensorflow:step = 49301, loss = 0.171683, precision = 1.0 (74.487 sec) +INFO:tensorflow:global_step/sec: 1.38009 +INFO:tensorflow:step = 49401, loss = 0.195505, precision = 0.992188 (72.459 sec) +INFO:tensorflow:global_step/sec: 1.3801 +INFO:tensorflow:step = 49501, loss = 0.195481, precision = 0.984375 (72.459 sec) +INFO:tensorflow:global_step/sec: 1.37985 +INFO:tensorflow:step = 49601, loss = 0.176351, precision = 0.992188 (72.472 sec) +Saved checkpoint after 127 epoch(s) to data/resnet164/checkpoints/00127... +INFO:tensorflow:global_step/sec: 1.34265 +INFO:tensorflow:step = 49701, loss = 0.197871, precision = 0.984375 (74.480 sec) +INFO:tensorflow:global_step/sec: 1.38027 +INFO:tensorflow:step = 49801, loss = 0.197897, precision = 0.984375 (72.449 sec) +INFO:tensorflow:global_step/sec: 1.37883 +INFO:tensorflow:step = 49901, loss = 0.196275, precision = 0.984375 (72.525 sec) +INFO:tensorflow:global_step/sec: 1.37939 +INFO:tensorflow:step = 50001, loss = 0.233134, precision = 0.953125 (72.496 sec) +Saved checkpoint after 128 epoch(s) to data/resnet164/checkpoints/00128... +INFO:tensorflow:global_step/sec: 1.34236 +INFO:tensorflow:step = 50101, loss = 0.174463, precision = 1.0 (74.496 sec) +INFO:tensorflow:global_step/sec: 1.38016 +INFO:tensorflow:step = 50201, loss = 0.186685, precision = 0.992188 (72.455 sec) +INFO:tensorflow:global_step/sec: 1.3796 +INFO:tensorflow:step = 50301, loss = 0.224603, precision = 0.96875 (72.485 sec) +INFO:tensorflow:global_step/sec: 1.38002 +INFO:tensorflow:step = 50401, loss = 0.217714, precision = 0.976562 (72.463 sec) +Saved checkpoint after 129 epoch(s) to data/resnet164/checkpoints/00129... +INFO:tensorflow:global_step/sec: 1.34229 +INFO:tensorflow:step = 50501, loss = 0.172745, precision = 0.992188 (74.500 sec) +INFO:tensorflow:global_step/sec: 1.38005 +INFO:tensorflow:step = 50601, loss = 0.169074, precision = 0.992188 (72.461 sec) +INFO:tensorflow:global_step/sec: 1.38016 +INFO:tensorflow:step = 50701, loss = 0.242936, precision = 0.960938 (72.455 sec) +INFO:tensorflow:global_step/sec: 1.37948 +INFO:tensorflow:step = 50801, loss = 0.219039, precision = 0.976562 (72.491 sec) +Saved checkpoint after 130 epoch(s) to data/resnet164/checkpoints/00130... +INFO:tensorflow:global_step/sec: 1.34269 +INFO:tensorflow:step = 50901, loss = 0.210132, precision = 0.984375 (74.478 sec) +INFO:tensorflow:global_step/sec: 1.38008 +INFO:tensorflow:step = 51001, loss = 0.241535, precision = 0.976562 (72.459 sec) +INFO:tensorflow:global_step/sec: 1.37996 +INFO:tensorflow:step = 51101, loss = 0.209235, precision = 0.976562 (72.466 sec) +INFO:tensorflow:global_step/sec: 1.38042 +INFO:tensorflow:step = 51201, loss = 0.206396, precision = 0.984375 (72.441 sec) +Saved checkpoint after 131 epoch(s) to data/resnet164/checkpoints/00131... +INFO:tensorflow:global_step/sec: 1.34233 +INFO:tensorflow:step = 51301, loss = 0.197363, precision = 0.984375 (74.497 sec) +INFO:tensorflow:global_step/sec: 1.38019 +INFO:tensorflow:step = 51401, loss = 0.192948, precision = 0.984375 (72.454 sec) +INFO:tensorflow:global_step/sec: 1.37952 +INFO:tensorflow:step = 51501, loss = 0.156492, precision = 1.0 (72.489 sec) +INFO:tensorflow:global_step/sec: 1.37977 +INFO:tensorflow:step = 51601, loss = 0.163582, precision = 1.0 (72.476 sec) +Saved checkpoint after 132 epoch(s) to data/resnet164/checkpoints/00132... +INFO:tensorflow:global_step/sec: 1.3384 +INFO:tensorflow:step = 51701, loss = 0.171711, precision = 0.992188 (74.716 sec) +INFO:tensorflow:global_step/sec: 1.37985 +INFO:tensorflow:step = 51801, loss = 0.184238, precision = 0.984375 (72.472 sec) +INFO:tensorflow:global_step/sec: 1.37977 +INFO:tensorflow:step = 51901, loss = 0.188997, precision = 0.992188 (72.476 sec) +INFO:tensorflow:global_step/sec: 1.37938 +INFO:tensorflow:step = 52001, loss = 0.21168, precision = 0.992188 (72.497 sec) +Saved checkpoint after 133 epoch(s) to data/resnet164/checkpoints/00133... +INFO:tensorflow:global_step/sec: 1.34245 +INFO:tensorflow:step = 52101, loss = 0.199282, precision = 0.976562 (74.491 sec) +INFO:tensorflow:global_step/sec: 1.37992 +INFO:tensorflow:step = 52201, loss = 0.282383, precision = 0.976562 (72.468 sec) +INFO:tensorflow:global_step/sec: 1.37999 +INFO:tensorflow:step = 52301, loss = 0.183288, precision = 0.984375 (72.464 sec) +Saved checkpoint after 134 epoch(s) to data/resnet164/checkpoints/00134... +INFO:tensorflow:global_step/sec: 1.3421 +INFO:tensorflow:step = 52401, loss = 0.192742, precision = 0.984375 (74.510 sec) +INFO:tensorflow:global_step/sec: 1.3802 +INFO:tensorflow:step = 52501, loss = 0.178399, precision = 0.984375 (72.453 sec) +INFO:tensorflow:global_step/sec: 1.37943 +INFO:tensorflow:step = 52601, loss = 0.190701, precision = 0.976562 (72.494 sec) +INFO:tensorflow:global_step/sec: 1.37989 +INFO:tensorflow:step = 52701, loss = 0.187779, precision = 0.984375 (72.470 sec) +Saved checkpoint after 135 epoch(s) to data/resnet164/checkpoints/00135... +INFO:tensorflow:global_step/sec: 1.34239 +INFO:tensorflow:step = 52801, loss = 0.190229, precision = 0.992188 (74.494 sec) +INFO:tensorflow:global_step/sec: 1.37942 +INFO:tensorflow:step = 52901, loss = 0.197764, precision = 0.984375 (72.494 sec) +INFO:tensorflow:global_step/sec: 1.37977 +INFO:tensorflow:step = 53001, loss = 0.171323, precision = 0.992188 (72.476 sec) +INFO:tensorflow:global_step/sec: 1.38047 +INFO:tensorflow:step = 53101, loss = 0.223882, precision = 0.984375 (72.439 sec) +Saved checkpoint after 136 epoch(s) to data/resnet164/checkpoints/00136... +INFO:tensorflow:global_step/sec: 1.34209 +INFO:tensorflow:step = 53201, loss = 0.163383, precision = 0.992188 (74.511 sec) +INFO:tensorflow:global_step/sec: 1.38012 +INFO:tensorflow:step = 53301, loss = 0.20037, precision = 0.976562 (72.458 sec) +INFO:tensorflow:global_step/sec: 1.38008 +INFO:tensorflow:step = 53401, loss = 0.162264, precision = 1.0 (72.460 sec) +INFO:tensorflow:global_step/sec: 1.38063 +INFO:tensorflow:step = 53501, loss = 0.162631, precision = 1.0 (72.431 sec) +Saved checkpoint after 137 epoch(s) to data/resnet164/checkpoints/00137... +INFO:tensorflow:global_step/sec: 1.3422 +INFO:tensorflow:step = 53601, loss = 0.149306, precision = 1.0 (74.505 sec) +INFO:tensorflow:global_step/sec: 1.38039 +INFO:tensorflow:step = 53701, loss = 0.155236, precision = 1.0 (72.443 sec) +INFO:tensorflow:global_step/sec: 1.37984 +INFO:tensorflow:step = 53801, loss = 0.147392, precision = 1.0 (72.472 sec) +INFO:tensorflow:global_step/sec: 1.38054 +INFO:tensorflow:step = 53901, loss = 0.15061, precision = 1.0 (72.435 sec) +Saved checkpoint after 138 epoch(s) to data/resnet164/checkpoints/00138... +INFO:tensorflow:global_step/sec: 1.3424 +INFO:tensorflow:step = 54001, loss = 0.147906, precision = 1.0 (74.493 sec) +INFO:tensorflow:global_step/sec: 1.37997 +INFO:tensorflow:step = 54101, loss = 0.153978, precision = 1.0 (72.465 sec) +INFO:tensorflow:global_step/sec: 1.37973 +INFO:tensorflow:step = 54201, loss = 0.173278, precision = 0.992188 (72.478 sec) +INFO:tensorflow:global_step/sec: 1.38029 +INFO:tensorflow:step = 54301, loss = 0.156943, precision = 0.992188 (72.449 sec) +Saved checkpoint after 139 epoch(s) to data/resnet164/checkpoints/00139... +INFO:tensorflow:global_step/sec: 1.34232 +INFO:tensorflow:step = 54401, loss = 0.14756, precision = 1.0 (74.498 sec) +INFO:tensorflow:global_step/sec: 1.37942 +INFO:tensorflow:step = 54501, loss = 0.16017, precision = 0.992188 (72.494 sec) +INFO:tensorflow:global_step/sec: 1.37957 +INFO:tensorflow:step = 54601, loss = 0.163587, precision = 0.992188 (72.486 sec) +INFO:tensorflow:global_step/sec: 1.37975 +INFO:tensorflow:step = 54701, loss = 0.1517, precision = 1.0 (72.477 sec) +Saved checkpoint after 140 epoch(s) to data/resnet164/checkpoints/00140... +INFO:tensorflow:global_step/sec: 1.342 +INFO:tensorflow:step = 54801, loss = 0.156709, precision = 1.0 (74.516 sec) +INFO:tensorflow:global_step/sec: 1.38025 +INFO:tensorflow:step = 54901, loss = 0.151172, precision = 1.0 (72.451 sec) +INFO:tensorflow:global_step/sec: 1.37972 +INFO:tensorflow:step = 55001, loss = 0.151846, precision = 1.0 (72.478 sec) +INFO:tensorflow:global_step/sec: 1.3798 +INFO:tensorflow:step = 55101, loss = 0.145245, precision = 1.0 (72.474 sec) +Saved checkpoint after 141 epoch(s) to data/resnet164/checkpoints/00141... +INFO:tensorflow:global_step/sec: 1.34302 +INFO:tensorflow:step = 55201, loss = 0.145915, precision = 1.0 (74.459 sec) +INFO:tensorflow:global_step/sec: 1.38044 +INFO:tensorflow:step = 55301, loss = 0.16522, precision = 0.992188 (72.441 sec) +INFO:tensorflow:global_step/sec: 1.38002 +INFO:tensorflow:step = 55401, loss = 0.164583, precision = 0.992188 (72.463 sec) +INFO:tensorflow:global_step/sec: 1.38031 +INFO:tensorflow:step = 55501, loss = 0.149506, precision = 1.0 (72.448 sec) +Saved checkpoint after 142 epoch(s) to data/resnet164/checkpoints/00142... +INFO:tensorflow:global_step/sec: 1.33809 +INFO:tensorflow:step = 55601, loss = 0.146593, precision = 1.0 (74.733 sec) +INFO:tensorflow:global_step/sec: 1.3796 +INFO:tensorflow:step = 55701, loss = 0.152829, precision = 0.992188 (72.484 sec) +INFO:tensorflow:global_step/sec: 1.3801 +INFO:tensorflow:step = 55801, loss = 0.149571, precision = 1.0 (72.458 sec) +INFO:tensorflow:global_step/sec: 1.37995 +INFO:tensorflow:step = 55901, loss = 0.146656, precision = 1.0 (72.466 sec) +Saved checkpoint after 143 epoch(s) to data/resnet164/checkpoints/00143... +INFO:tensorflow:global_step/sec: 1.3422 +INFO:tensorflow:step = 56001, loss = 0.153983, precision = 1.0 (74.505 sec) +INFO:tensorflow:global_step/sec: 1.37993 +INFO:tensorflow:step = 56101, loss = 0.143817, precision = 1.0 (72.468 sec) +INFO:tensorflow:global_step/sec: 1.38074 +INFO:tensorflow:step = 56201, loss = 0.149217, precision = 1.0 (72.425 sec) +INFO:tensorflow:global_step/sec: 1.37998 +INFO:tensorflow:step = 56301, loss = 0.147841, precision = 1.0 (72.465 sec) +Saved checkpoint after 144 epoch(s) to data/resnet164/checkpoints/00144... +INFO:tensorflow:global_step/sec: 1.34199 +INFO:tensorflow:step = 56401, loss = 0.145371, precision = 1.0 (74.516 sec) +INFO:tensorflow:global_step/sec: 1.37959 +INFO:tensorflow:step = 56501, loss = 0.146306, precision = 1.0 (72.485 sec) +INFO:tensorflow:global_step/sec: 1.37995 +INFO:tensorflow:step = 56601, loss = 0.14793, precision = 1.0 (72.466 sec) +Saved checkpoint after 145 epoch(s) to data/resnet164/checkpoints/00145... +INFO:tensorflow:global_step/sec: 1.34361 +INFO:tensorflow:step = 56701, loss = 0.143309, precision = 1.0 (74.426 sec) +INFO:tensorflow:global_step/sec: 1.38017 +INFO:tensorflow:step = 56801, loss = 0.144068, precision = 1.0 (72.455 sec) +INFO:tensorflow:global_step/sec: 1.3797 +INFO:tensorflow:step = 56901, loss = 0.145012, precision = 1.0 (72.480 sec) +INFO:tensorflow:global_step/sec: 1.38005 +INFO:tensorflow:step = 57001, loss = 0.142365, precision = 1.0 (72.461 sec) +Saved checkpoint after 146 epoch(s) to data/resnet164/checkpoints/00146... +INFO:tensorflow:global_step/sec: 1.34178 +INFO:tensorflow:step = 57101, loss = 0.146322, precision = 1.0 (74.528 sec) +INFO:tensorflow:global_step/sec: 1.3802 +INFO:tensorflow:step = 57201, loss = 0.140394, precision = 1.0 (72.453 sec) +INFO:tensorflow:global_step/sec: 1.37987 +INFO:tensorflow:step = 57301, loss = 0.14302, precision = 1.0 (72.470 sec) +INFO:tensorflow:global_step/sec: 1.3801 +INFO:tensorflow:step = 57401, loss = 0.142653, precision = 1.0 (72.458 sec) +Saved checkpoint after 147 epoch(s) to data/resnet164/checkpoints/00147... +INFO:tensorflow:global_step/sec: 1.34279 +INFO:tensorflow:step = 57501, loss = 0.142091, precision = 1.0 (74.472 sec) +INFO:tensorflow:global_step/sec: 1.37927 +INFO:tensorflow:step = 57601, loss = 0.145931, precision = 1.0 (72.502 sec) +INFO:tensorflow:global_step/sec: 1.38034 +INFO:tensorflow:step = 57701, loss = 0.147896, precision = 1.0 (72.446 sec) +INFO:tensorflow:global_step/sec: 1.38003 +INFO:tensorflow:step = 57801, loss = 0.143395, precision = 1.0 (72.462 sec) +Saved checkpoint after 148 epoch(s) to data/resnet164/checkpoints/00148... +INFO:tensorflow:global_step/sec: 1.34128 +INFO:tensorflow:step = 57901, loss = 0.148966, precision = 0.992188 (74.556 sec) +INFO:tensorflow:global_step/sec: 1.38073 +INFO:tensorflow:step = 58001, loss = 0.142076, precision = 1.0 (72.425 sec) +INFO:tensorflow:global_step/sec: 1.38081 +INFO:tensorflow:step = 58101, loss = 0.141362, precision = 1.0 (72.421 sec) +INFO:tensorflow:global_step/sec: 1.38065 +INFO:tensorflow:step = 58201, loss = 0.144317, precision = 1.0 (72.429 sec) +Saved checkpoint after 149 epoch(s) to data/resnet164/checkpoints/00149... +INFO:tensorflow:global_step/sec: 1.34222 +INFO:tensorflow:step = 58301, loss = 0.138353, precision = 1.0 (74.504 sec) +INFO:tensorflow:global_step/sec: 1.38047 +INFO:tensorflow:step = 58401, loss = 0.140506, precision = 1.0 (72.439 sec) +INFO:tensorflow:global_step/sec: 1.38111 +INFO:tensorflow:step = 58501, loss = 0.140917, precision = 1.0 (72.406 sec) +INFO:tensorflow:global_step/sec: 1.37997 +INFO:tensorflow:step = 58601, loss = 0.138601, precision = 1.0 (72.466 sec) +Saved checkpoint after 150 epoch(s) to data/resnet164/checkpoints/00150... +INFO:tensorflow:global_step/sec: 1.34217 +INFO:tensorflow:step = 58701, loss = 0.140843, precision = 1.0 (74.506 sec) +INFO:tensorflow:global_step/sec: 1.38136 +INFO:tensorflow:step = 58801, loss = 0.13943, precision = 1.0 (72.392 sec) +INFO:tensorflow:global_step/sec: 1.38009 +INFO:tensorflow:step = 58901, loss = 0.138747, precision = 1.0 (72.459 sec) +INFO:tensorflow:global_step/sec: 1.38111 +INFO:tensorflow:step = 59001, loss = 0.141136, precision = 1.0 (72.406 sec) +Saved checkpoint after 151 epoch(s) to data/resnet164/checkpoints/00151... +INFO:tensorflow:global_step/sec: 1.34409 +INFO:tensorflow:step = 59101, loss = 0.143382, precision = 1.0 (74.400 sec) +INFO:tensorflow:global_step/sec: 1.38066 +INFO:tensorflow:step = 59201, loss = 0.138805, precision = 1.0 (72.429 sec) +INFO:tensorflow:global_step/sec: 1.38044 +INFO:tensorflow:step = 59301, loss = 0.144236, precision = 1.0 (72.441 sec) +INFO:tensorflow:global_step/sec: 1.38069 +INFO:tensorflow:step = 59401, loss = 0.139455, precision = 1.0 (72.428 sec) +Saved checkpoint after 152 epoch(s) to data/resnet164/checkpoints/00152... +INFO:tensorflow:global_step/sec: 1.34049 +INFO:tensorflow:step = 59501, loss = 0.13953, precision = 1.0 (74.599 sec) +INFO:tensorflow:global_step/sec: 1.38072 +INFO:tensorflow:step = 59601, loss = 0.140441, precision = 1.0 (72.426 sec) +INFO:tensorflow:global_step/sec: 1.38101 +INFO:tensorflow:step = 59701, loss = 0.140954, precision = 1.0 (72.411 sec) +INFO:tensorflow:global_step/sec: 1.38049 +INFO:tensorflow:step = 59801, loss = 0.138693, precision = 1.0 (72.438 sec) +Saved checkpoint after 153 epoch(s) to data/resnet164/checkpoints/00153... +INFO:tensorflow:global_step/sec: 1.34214 +INFO:tensorflow:step = 59901, loss = 0.136399, precision = 1.0 (74.508 sec) +INFO:tensorflow:global_step/sec: 1.38035 +INFO:tensorflow:step = 60001, loss = 0.141033, precision = 1.0 (72.445 sec) +INFO:tensorflow:global_step/sec: 1.38065 +INFO:tensorflow:step = 60101, loss = 0.140297, precision = 1.0 (72.430 sec) +INFO:tensorflow:global_step/sec: 1.38006 +INFO:tensorflow:step = 60201, loss = 0.144203, precision = 1.0 (72.460 sec) +Saved checkpoint after 154 epoch(s) to data/resnet164/checkpoints/00154... +INFO:tensorflow:global_step/sec: 1.34296 +INFO:tensorflow:step = 60301, loss = 0.136698, precision = 1.0 (74.462 sec) +INFO:tensorflow:global_step/sec: 1.38024 +INFO:tensorflow:step = 60401, loss = 0.135943, precision = 1.0 (72.451 sec) +INFO:tensorflow:global_step/sec: 1.37987 +INFO:tensorflow:step = 60501, loss = 0.137866, precision = 1.0 (72.471 sec) +INFO:tensorflow:global_step/sec: 1.38045 +INFO:tensorflow:step = 60601, loss = 0.136414, precision = 1.0 (72.440 sec) +Saved checkpoint after 155 epoch(s) to data/resnet164/checkpoints/00155... +INFO:tensorflow:global_step/sec: 1.34183 +INFO:tensorflow:step = 60701, loss = 0.136635, precision = 1.0 (74.525 sec) +INFO:tensorflow:global_step/sec: 1.38012 +INFO:tensorflow:step = 60801, loss = 0.138931, precision = 1.0 (72.457 sec) +INFO:tensorflow:global_step/sec: 1.38027 +INFO:tensorflow:step = 60901, loss = 0.137717, precision = 1.0 (72.450 sec) +Saved checkpoint after 156 epoch(s) to data/resnet164/checkpoints/00156... +INFO:tensorflow:global_step/sec: 1.34231 +INFO:tensorflow:step = 61001, loss = 0.136982, precision = 1.0 (74.499 sec) +INFO:tensorflow:global_step/sec: 1.37991 +INFO:tensorflow:step = 61101, loss = 0.139826, precision = 1.0 (72.469 sec) +INFO:tensorflow:global_step/sec: 1.37981 +INFO:tensorflow:step = 61201, loss = 0.135672, precision = 1.0 (72.474 sec) +INFO:tensorflow:global_step/sec: 1.37938 +INFO:tensorflow:step = 61301, loss = 0.156501, precision = 0.992188 (72.496 sec) +Saved checkpoint after 157 epoch(s) to data/resnet164/checkpoints/00157... +INFO:tensorflow:global_step/sec: 1.34156 +INFO:tensorflow:step = 61401, loss = 0.135838, precision = 1.0 (74.540 sec) +INFO:tensorflow:global_step/sec: 1.38045 +INFO:tensorflow:step = 61501, loss = 0.141953, precision = 1.0 (72.440 sec) +INFO:tensorflow:global_step/sec: 1.3794 +INFO:tensorflow:step = 61601, loss = 0.13449, precision = 1.0 (72.495 sec) +INFO:tensorflow:global_step/sec: 1.37966 +INFO:tensorflow:step = 61701, loss = 0.139027, precision = 1.0 (72.482 sec) +Saved checkpoint after 158 epoch(s) to data/resnet164/checkpoints/00158... +INFO:tensorflow:global_step/sec: 1.34207 +INFO:tensorflow:step = 61801, loss = 0.135659, precision = 1.0 (74.512 sec) +INFO:tensorflow:global_step/sec: 1.37958 +INFO:tensorflow:step = 61901, loss = 0.137448, precision = 1.0 (72.486 sec) +INFO:tensorflow:global_step/sec: 1.38043 +INFO:tensorflow:step = 62001, loss = 0.134966, precision = 1.0 (72.441 sec) +INFO:tensorflow:global_step/sec: 1.38021 +INFO:tensorflow:step = 62101, loss = 0.137159, precision = 1.0 (72.453 sec) +Saved checkpoint after 159 epoch(s) to data/resnet164/checkpoints/00159... +INFO:tensorflow:global_step/sec: 1.34174 +INFO:tensorflow:step = 62201, loss = 0.134425, precision = 1.0 (74.530 sec) +INFO:tensorflow:global_step/sec: 1.38021 +INFO:tensorflow:step = 62301, loss = 0.134616, precision = 1.0 (72.453 sec) +INFO:tensorflow:global_step/sec: 1.37996 +INFO:tensorflow:step = 62401, loss = 0.13402, precision = 1.0 (72.466 sec) +INFO:tensorflow:global_step/sec: 1.38063 +INFO:tensorflow:step = 62501, loss = 0.144149, precision = 1.0 (72.431 sec) +Saved checkpoint after 160 epoch(s) to data/resnet164/checkpoints/00160... +INFO:tensorflow:global_step/sec: 1.34293 +INFO:tensorflow:step = 62601, loss = 0.134148, precision = 1.0 (74.464 sec) +INFO:tensorflow:global_step/sec: 1.38057 +INFO:tensorflow:step = 62701, loss = 0.134293, precision = 1.0 (72.434 sec) +INFO:tensorflow:global_step/sec: 1.38039 +INFO:tensorflow:step = 62801, loss = 0.134053, precision = 1.0 (72.443 sec) +INFO:tensorflow:global_step/sec: 1.37988 +INFO:tensorflow:step = 62901, loss = 0.134996, precision = 1.0 (72.470 sec) +Saved checkpoint after 161 epoch(s) to data/resnet164/checkpoints/00161... +INFO:tensorflow:global_step/sec: 1.34385 +INFO:tensorflow:step = 63001, loss = 0.132756, precision = 1.0 (74.413 sec) +INFO:tensorflow:global_step/sec: 1.38071 +INFO:tensorflow:step = 63101, loss = 0.132362, precision = 1.0 (72.426 sec) +INFO:tensorflow:global_step/sec: 1.38076 +INFO:tensorflow:step = 63201, loss = 0.139944, precision = 1.0 (72.424 sec) +INFO:tensorflow:global_step/sec: 1.38121 +INFO:tensorflow:step = 63301, loss = 0.132115, precision = 1.0 (72.400 sec) +Saved checkpoint after 162 epoch(s) to data/resnet164/checkpoints/00162... +INFO:tensorflow:global_step/sec: 1.34307 +INFO:tensorflow:step = 63401, loss = 0.183218, precision = 0.992188 (74.457 sec) +INFO:tensorflow:global_step/sec: 1.38054 +INFO:tensorflow:step = 63501, loss = 0.13395, precision = 1.0 (72.436 sec) +INFO:tensorflow:global_step/sec: 1.38025 +INFO:tensorflow:step = 63601, loss = 0.136109, precision = 1.0 (72.451 sec) +INFO:tensorflow:global_step/sec: 1.38052 +INFO:tensorflow:step = 63701, loss = 0.1342, precision = 1.0 (72.437 sec) +Saved checkpoint after 163 epoch(s) to data/resnet164/checkpoints/00163... +INFO:tensorflow:global_step/sec: 1.33907 +INFO:tensorflow:step = 63801, loss = 0.132322, precision = 1.0 (74.679 sec) +INFO:tensorflow:global_step/sec: 1.37999 +INFO:tensorflow:step = 63901, loss = 0.136683, precision = 1.0 (72.464 sec) +INFO:tensorflow:global_step/sec: 1.38012 +INFO:tensorflow:step = 64001, loss = 0.134241, precision = 1.0 (72.457 sec) +INFO:tensorflow:global_step/sec: 1.37987 +INFO:tensorflow:step = 64101, loss = 0.13512, precision = 1.0 (72.471 sec) +Saved checkpoint after 164 epoch(s) to data/resnet164/checkpoints/00164... +INFO:tensorflow:global_step/sec: 1.34216 +INFO:tensorflow:step = 64201, loss = 0.130926, precision = 1.0 (74.507 sec) +INFO:tensorflow:global_step/sec: 1.38018 +INFO:tensorflow:step = 64301, loss = 0.130578, precision = 1.0 (72.454 sec) +INFO:tensorflow:global_step/sec: 1.38023 +INFO:tensorflow:step = 64401, loss = 0.1314, precision = 1.0 (72.451 sec) +INFO:tensorflow:global_step/sec: 1.3792 +INFO:tensorflow:step = 64501, loss = 0.145365, precision = 0.992188 (72.506 sec) +Saved checkpoint after 165 epoch(s) to data/resnet164/checkpoints/00165... +INFO:tensorflow:global_step/sec: 1.34306 +INFO:tensorflow:step = 64601, loss = 0.130826, precision = 1.0 (74.457 sec) +INFO:tensorflow:global_step/sec: 1.38 +INFO:tensorflow:step = 64701, loss = 0.131676, precision = 1.0 (72.464 sec) +INFO:tensorflow:global_step/sec: 1.37945 +INFO:tensorflow:step = 64801, loss = 0.136718, precision = 1.0 (72.492 sec) +INFO:tensorflow:global_step/sec: 1.37925 +INFO:tensorflow:step = 64901, loss = 0.133956, precision = 1.0 (72.503 sec) +Saved checkpoint after 166 epoch(s) to data/resnet164/checkpoints/00166... +INFO:tensorflow:global_step/sec: 1.34207 +INFO:tensorflow:step = 65001, loss = 0.130224, precision = 1.0 (74.512 sec) +INFO:tensorflow:global_step/sec: 1.38032 +INFO:tensorflow:step = 65101, loss = 0.130544, precision = 1.0 (72.447 sec) +INFO:tensorflow:global_step/sec: 1.37966 +INFO:tensorflow:step = 65201, loss = 0.130836, precision = 1.0 (72.481 sec) +Saved checkpoint after 167 epoch(s) to data/resnet164/checkpoints/00167... +INFO:tensorflow:global_step/sec: 1.34252 +INFO:tensorflow:step = 65301, loss = 0.13118, precision = 1.0 (74.487 sec) +INFO:tensorflow:global_step/sec: 1.37969 +INFO:tensorflow:step = 65401, loss = 0.129618, precision = 1.0 (72.480 sec) +INFO:tensorflow:global_step/sec: 1.37998 +INFO:tensorflow:step = 65501, loss = 0.129269, precision = 1.0 (72.465 sec) +INFO:tensorflow:global_step/sec: 1.37959 +INFO:tensorflow:step = 65601, loss = 0.129481, precision = 1.0 (72.485 sec) +Saved checkpoint after 168 epoch(s) to data/resnet164/checkpoints/00168... +INFO:tensorflow:global_step/sec: 1.34234 +INFO:tensorflow:step = 65701, loss = 0.130263, precision = 1.0 (74.497 sec) +INFO:tensorflow:global_step/sec: 1.37983 +INFO:tensorflow:step = 65801, loss = 0.129327, precision = 1.0 (72.472 sec) +INFO:tensorflow:global_step/sec: 1.38025 +INFO:tensorflow:step = 65901, loss = 0.12906, precision = 1.0 (72.451 sec) +INFO:tensorflow:global_step/sec: 1.38056 +INFO:tensorflow:step = 66001, loss = 0.132457, precision = 1.0 (72.434 sec) +Saved checkpoint after 169 epoch(s) to data/resnet164/checkpoints/00169... +INFO:tensorflow:global_step/sec: 1.3425 +INFO:tensorflow:step = 66101, loss = 0.130364, precision = 1.0 (74.488 sec) +INFO:tensorflow:global_step/sec: 1.38047 +INFO:tensorflow:step = 66201, loss = 0.133962, precision = 1.0 (72.439 sec) +INFO:tensorflow:global_step/sec: 1.37989 +INFO:tensorflow:step = 66301, loss = 0.13121, precision = 1.0 (72.470 sec) +INFO:tensorflow:global_step/sec: 1.3801 +INFO:tensorflow:step = 66401, loss = 0.128366, precision = 1.0 (72.458 sec) +Saved checkpoint after 170 epoch(s) to data/resnet164/checkpoints/00170... +INFO:tensorflow:global_step/sec: 1.34308 +INFO:tensorflow:step = 66501, loss = 0.128467, precision = 1.0 (74.456 sec) +INFO:tensorflow:global_step/sec: 1.38048 +INFO:tensorflow:step = 66601, loss = 0.12817, precision = 1.0 (72.438 sec) +INFO:tensorflow:global_step/sec: 1.3802 +INFO:tensorflow:step = 66701, loss = 0.129547, precision = 1.0 (72.453 sec) +INFO:tensorflow:global_step/sec: 1.3799 +INFO:tensorflow:step = 66801, loss = 0.130289, precision = 1.0 (72.469 sec) +Saved checkpoint after 171 epoch(s) to data/resnet164/checkpoints/00171... +INFO:tensorflow:global_step/sec: 1.34272 +INFO:tensorflow:step = 66901, loss = 0.127497, precision = 1.0 (74.476 sec) +INFO:tensorflow:global_step/sec: 1.37998 +INFO:tensorflow:step = 67001, loss = 0.127866, precision = 1.0 (72.465 sec) +INFO:tensorflow:global_step/sec: 1.38062 +INFO:tensorflow:step = 67101, loss = 0.131174, precision = 1.0 (72.431 sec) +INFO:tensorflow:global_step/sec: 1.38091 +INFO:tensorflow:step = 67201, loss = 0.130517, precision = 1.0 (72.416 sec) +Saved checkpoint after 172 epoch(s) to data/resnet164/checkpoints/00172... +INFO:tensorflow:global_step/sec: 1.34219 +INFO:tensorflow:step = 67301, loss = 0.127718, precision = 1.0 (74.505 sec) +INFO:tensorflow:global_step/sec: 1.38074 +INFO:tensorflow:step = 67401, loss = 0.127816, precision = 1.0 (72.425 sec) +INFO:tensorflow:global_step/sec: 1.37973 +INFO:tensorflow:step = 67501, loss = 0.12677, precision = 1.0 (72.478 sec) +INFO:tensorflow:global_step/sec: 1.37927 +INFO:tensorflow:step = 67601, loss = 0.127132, precision = 1.0 (72.502 sec) +Saved checkpoint after 173 epoch(s) to data/resnet164/checkpoints/00173... +INFO:tensorflow:global_step/sec: 1.33835 +INFO:tensorflow:step = 67701, loss = 0.12913, precision = 1.0 (74.719 sec) +INFO:tensorflow:global_step/sec: 1.37948 +INFO:tensorflow:step = 67801, loss = 0.128085, precision = 1.0 (72.491 sec) +INFO:tensorflow:global_step/sec: 1.37908 +INFO:tensorflow:step = 67901, loss = 0.133165, precision = 1.0 (72.512 sec) +INFO:tensorflow:global_step/sec: 1.37898 +INFO:tensorflow:step = 68001, loss = 0.145059, precision = 0.992188 (72.518 sec) +Saved checkpoint after 174 epoch(s) to data/resnet164/checkpoints/00174... +INFO:tensorflow:global_step/sec: 1.34252 +INFO:tensorflow:step = 68101, loss = 0.128805, precision = 1.0 (74.487 sec) +INFO:tensorflow:global_step/sec: 1.37942 +INFO:tensorflow:step = 68201, loss = 0.126559, precision = 1.0 (72.494 sec) +INFO:tensorflow:global_step/sec: 1.37931 +INFO:tensorflow:step = 68301, loss = 0.129718, precision = 1.0 (72.500 sec) +INFO:tensorflow:global_step/sec: 1.37923 +INFO:tensorflow:step = 68401, loss = 0.127166, precision = 1.0 (72.504 sec) +Saved checkpoint after 175 epoch(s) to data/resnet164/checkpoints/00175... +INFO:tensorflow:global_step/sec: 1.34148 +INFO:tensorflow:step = 68501, loss = 0.126302, precision = 1.0 (74.545 sec) +INFO:tensorflow:global_step/sec: 1.37989 +INFO:tensorflow:step = 68601, loss = 0.12554, precision = 1.0 (72.469 sec) +INFO:tensorflow:global_step/sec: 1.38007 +INFO:tensorflow:step = 68701, loss = 0.128204, precision = 1.0 (72.460 sec) +INFO:tensorflow:global_step/sec: 1.3792 +INFO:tensorflow:step = 68801, loss = 0.136946, precision = 1.0 (72.506 sec) +Saved checkpoint after 176 epoch(s) to data/resnet164/checkpoints/00176... +INFO:tensorflow:global_step/sec: 1.34292 +INFO:tensorflow:step = 68901, loss = 0.125704, precision = 1.0 (74.465 sec) +INFO:tensorflow:global_step/sec: 1.37989 +INFO:tensorflow:step = 69001, loss = 0.126611, precision = 1.0 (72.469 sec) +INFO:tensorflow:global_step/sec: 1.38001 +INFO:tensorflow:step = 69101, loss = 0.129123, precision = 1.0 (72.464 sec) +INFO:tensorflow:global_step/sec: 1.3798 +INFO:tensorflow:step = 69201, loss = 0.124895, precision = 1.0 (72.474 sec) +Saved checkpoint after 177 epoch(s) to data/resnet164/checkpoints/00177... +INFO:tensorflow:global_step/sec: 1.34193 +INFO:tensorflow:step = 69301, loss = 0.124874, precision = 1.0 (74.520 sec) +INFO:tensorflow:global_step/sec: 1.37983 +INFO:tensorflow:step = 69401, loss = 0.12561, precision = 1.0 (72.473 sec) +INFO:tensorflow:global_step/sec: 1.38032 +INFO:tensorflow:step = 69501, loss = 0.126745, precision = 1.0 (72.447 sec) +Saved checkpoint after 178 epoch(s) to data/resnet164/checkpoints/00178... +INFO:tensorflow:global_step/sec: 1.34222 +INFO:tensorflow:step = 69601, loss = 0.126426, precision = 1.0 (74.504 sec) +INFO:tensorflow:global_step/sec: 1.38039 +INFO:tensorflow:step = 69701, loss = 0.124839, precision = 1.0 (72.444 sec) +INFO:tensorflow:global_step/sec: 1.38082 +INFO:tensorflow:step = 69801, loss = 0.124205, precision = 1.0 (72.421 sec) +INFO:tensorflow:global_step/sec: 1.37968 +INFO:tensorflow:step = 69901, loss = 0.125906, precision = 1.0 (72.481 sec) +Saved checkpoint after 179 epoch(s) to data/resnet164/checkpoints/00179... +INFO:tensorflow:global_step/sec: 1.34301 +INFO:tensorflow:step = 70001, loss = 0.123676, precision = 1.0 (74.459 sec) +INFO:tensorflow:global_step/sec: 1.37962 +INFO:tensorflow:step = 70101, loss = 0.123378, precision = 1.0 (72.484 sec) +INFO:tensorflow:global_step/sec: 1.37965 +INFO:tensorflow:step = 70201, loss = 0.123294, precision = 1.0 (72.482 sec) +INFO:tensorflow:global_step/sec: 1.3805 +INFO:tensorflow:step = 70301, loss = 0.12342, precision = 1.0 (72.438 sec) +Saved checkpoint after 180 epoch(s) to data/resnet164/checkpoints/00180... +INFO:tensorflow:global_step/sec: 1.34256 +INFO:tensorflow:step = 70401, loss = 0.123247, precision = 1.0 (74.485 sec) +INFO:tensorflow:global_step/sec: 1.38003 +INFO:tensorflow:step = 70501, loss = 0.125684, precision = 1.0 (72.462 sec) +INFO:tensorflow:global_step/sec: 1.3798 +INFO:tensorflow:step = 70601, loss = 0.128968, precision = 1.0 (72.474 sec) +INFO:tensorflow:global_step/sec: 1.37976 +INFO:tensorflow:step = 70701, loss = 0.123338, precision = 1.0 (72.476 sec) +Saved checkpoint after 181 epoch(s) to data/resnet164/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1k80_ec2/resnet164_nb_train.log b/tensorflow/CIFAR10/logs/1k80_ec2/resnet164_nb_train.log new file mode 100644 index 0000000..770a573 --- /dev/null +++ b/tensorflow/CIFAR10/logs/1k80_ec2/resnet164_nb_train.log @@ -0,0 +1,2162 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/2.60m params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_10/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/sub2/bn2/beta (16, 16/16 params) + unit_1_10/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_11/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/sub2/bn2/beta (16, 16/16 params) + unit_1_11/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_12/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/sub2/bn2/beta (16, 16/16 params) + unit_1_12/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_13/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/sub2/bn2/beta (16, 16/16 params) + unit_1_13/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_14/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/sub2/bn2/beta (16, 16/16 params) + unit_1_14/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_15/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/sub2/bn2/beta (16, 16/16 params) + unit_1_15/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_16/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/sub2/bn2/beta (16, 16/16 params) + unit_1_16/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_17/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/sub2/bn2/beta (16, 16/16 params) + unit_1_17/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_18/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_18/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_18/sub2/bn2/beta (16, 16/16 params) + unit_1_18/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_19/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_19/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_19/sub2/bn2/beta (16, 16/16 params) + unit_1_19/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_20/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_20/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_20/sub2/bn2/beta (16, 16/16 params) + unit_1_20/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_21/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_21/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_21/sub2/bn2/beta (16, 16/16 params) + unit_1_21/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_22/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_22/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_22/sub2/bn2/beta (16, 16/16 params) + unit_1_22/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_23/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_23/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_23/sub2/bn2/beta (16, 16/16 params) + unit_1_23/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_24/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_24/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_24/sub2/bn2/beta (16, 16/16 params) + unit_1_24/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_25/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_25/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_25/sub2/bn2/beta (16, 16/16 params) + unit_1_25/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_26/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_26/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_26/sub2/bn2/beta (16, 16/16 params) + unit_1_26/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_3/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_4/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_5/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_6/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_7/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_8/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_9/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/sub2/bn2/beta (16, 16/16 params) + unit_1_9/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_10/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/sub2/bn2/beta (32, 32/32 params) + unit_2_10/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_11/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/sub2/bn2/beta (32, 32/32 params) + unit_2_11/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_12/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/sub2/bn2/beta (32, 32/32 params) + unit_2_12/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_13/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/sub2/bn2/beta (32, 32/32 params) + unit_2_13/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_14/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/sub2/bn2/beta (32, 32/32 params) + unit_2_14/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_15/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/sub2/bn2/beta (32, 32/32 params) + unit_2_15/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_16/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/sub2/bn2/beta (32, 32/32 params) + unit_2_16/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_17/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/sub2/bn2/beta (32, 32/32 params) + unit_2_17/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_18/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_18/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_18/sub2/bn2/beta (32, 32/32 params) + unit_2_18/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_19/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_19/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_19/sub2/bn2/beta (32, 32/32 params) + unit_2_19/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_20/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_20/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_20/sub2/bn2/beta (32, 32/32 params) + unit_2_20/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_21/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_21/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_21/sub2/bn2/beta (32, 32/32 params) + unit_2_21/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_22/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_22/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_22/sub2/bn2/beta (32, 32/32 params) + unit_2_22/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_23/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_23/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_23/sub2/bn2/beta (32, 32/32 params) + unit_2_23/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_24/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_24/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_24/sub2/bn2/beta (32, 32/32 params) + unit_2_24/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_25/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_25/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_25/sub2/bn2/beta (32, 32/32 params) + unit_2_25/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_26/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_26/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_26/sub2/bn2/beta (32, 32/32 params) + unit_2_26/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_3/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_4/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_5/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_6/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_7/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_8/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_9/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/sub2/bn2/beta (32, 32/32 params) + unit_2_9/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_10/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/sub2/bn2/beta (64, 64/64 params) + unit_3_10/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_11/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/sub2/bn2/beta (64, 64/64 params) + unit_3_11/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_12/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/sub2/bn2/beta (64, 64/64 params) + unit_3_12/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_13/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/sub2/bn2/beta (64, 64/64 params) + unit_3_13/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_14/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/sub2/bn2/beta (64, 64/64 params) + unit_3_14/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_15/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/sub2/bn2/beta (64, 64/64 params) + unit_3_15/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_16/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/sub2/bn2/beta (64, 64/64 params) + unit_3_16/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_17/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/sub2/bn2/beta (64, 64/64 params) + unit_3_17/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_18/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_18/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_18/sub2/bn2/beta (64, 64/64 params) + unit_3_18/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_19/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_19/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_19/sub2/bn2/beta (64, 64/64 params) + unit_3_19/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_20/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_20/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_20/sub2/bn2/beta (64, 64/64 params) + unit_3_20/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_21/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_21/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_21/sub2/bn2/beta (64, 64/64 params) + unit_3_21/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_22/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_22/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_22/sub2/bn2/beta (64, 64/64 params) + unit_3_22/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_23/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_23/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_23/sub2/bn2/beta (64, 64/64 params) + unit_3_23/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_24/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_24/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_24/sub2/bn2/beta (64, 64/64 params) + unit_3_24/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_25/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_25/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_25/sub2/bn2/beta (64, 64/64 params) + unit_3_25/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_26/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_26/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_26/sub2/bn2/beta (64, 64/64 params) + unit_3_26/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_3/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_4/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_5/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_6/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_7/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_8/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_9/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/sub2/bn2/beta (64, 64/64 params) + unit_3_9/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/97.35b flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + +======================End of Report========================== +2017-07-30 07:39:56.718375: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero +2017-07-30 07:39:56.718895: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla K80 +major: 3 minor: 7 memoryClockRate (GHz) 0.8235 +pciBusID 0000:00:1e.0 +Total memory: 11.17GiB +Free memory: 11.11GiB +2017-07-30 07:39:56.718928: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-07-30 07:39:56.718940: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-07-30 07:39:56.718962: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:00:1e.0) +2017-07-30 07:39:59.003499: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 07:39:59.003551: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 4 visible devices +2017-07-30 07:39:59.004497: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x7dd4970 executing computations on platform Host. Devices: +2017-07-30 07:39:59.004515: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-07-30 07:39:59.005185: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 07:39:59.005209: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 4 visible devices +2017-07-30 07:39:59.005675: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x92cb4b0 executing computations on platform CUDA. Devices: +2017-07-30 07:39:59.005695: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla K80, Compute Capability 3.7 +2017-07-30 07:40:11.348041: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:247] PoolAllocator: After 1146 get requests, put_count=1100 evicted_count=1000 eviction_rate=0.909091 and unsatisfied allocation rate=1 +2017-07-30 07:40:11.348106: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:259] Raising pool_size_limit_ from 100 to 110 +INFO:tensorflow:step = 1, loss = 5.27797, precision = 0.125 +INFO:tensorflow:global_step/sec: 1.5135 +INFO:tensorflow:step = 101, loss = 4.67691, precision = 0.335938 (66.073 sec) +INFO:tensorflow:global_step/sec: 1.5382 +INFO:tensorflow:step = 201, loss = 4.43108, precision = 0.507812 (65.011 sec) +INFO:tensorflow:global_step/sec: 1.53584 +INFO:tensorflow:step = 301, loss = 4.43132, precision = 0.476562 (65.111 sec) +total_params: 2596842 +Saved checkpoint after 1 epoch(s) to data/resnet164/checkpoints/00001... +INFO:tensorflow:global_step/sec: 1.47387 +INFO:tensorflow:step = 401, loss = 4.69635, precision = 0.257812 (67.848 sec) +INFO:tensorflow:global_step/sec: 1.5363 +INFO:tensorflow:step = 501, loss = 4.08161, precision = 0.476562 (65.091 sec) +INFO:tensorflow:global_step/sec: 1.53527 +INFO:tensorflow:step = 601, loss = 3.65396, precision = 0.523438 (65.135 sec) +INFO:tensorflow:global_step/sec: 1.53611 +INFO:tensorflow:step = 701, loss = 3.29999, precision = 0.578125 (65.100 sec) +Saved checkpoint after 2 epoch(s) to data/resnet164/checkpoints/00002... +INFO:tensorflow:global_step/sec: 1.48225 +INFO:tensorflow:step = 801, loss = 3.07529, precision = 0.609375 (67.465 sec) +INFO:tensorflow:global_step/sec: 1.53433 +INFO:tensorflow:step = 901, loss = 2.74084, precision = 0.710938 (65.175 sec) +INFO:tensorflow:global_step/sec: 1.53442 +INFO:tensorflow:step = 1001, loss = 2.45254, precision = 0.742188 (65.171 sec) +INFO:tensorflow:global_step/sec: 1.53425 +INFO:tensorflow:step = 1101, loss = 2.36702, precision = 0.710938 (65.178 sec) +Saved checkpoint after 3 epoch(s) to data/resnet164/checkpoints/00003... +INFO:tensorflow:global_step/sec: 1.48202 +INFO:tensorflow:step = 1201, loss = 2.06267, precision = 0.773438 (67.476 sec) +INFO:tensorflow:global_step/sec: 1.53414 +INFO:tensorflow:step = 1301, loss = 2.12234, precision = 0.710938 (65.183 sec) +INFO:tensorflow:global_step/sec: 1.5344 +INFO:tensorflow:step = 1401, loss = 1.97477, precision = 0.703125 (65.172 sec) +INFO:tensorflow:global_step/sec: 1.53399 +INFO:tensorflow:step = 1501, loss = 1.84352, precision = 0.742188 (65.190 sec) +Saved checkpoint after 4 epoch(s) to data/resnet164/checkpoints/00004... +INFO:tensorflow:global_step/sec: 1.48232 +INFO:tensorflow:step = 1601, loss = 1.84976, precision = 0.71875 (67.462 sec) +INFO:tensorflow:global_step/sec: 1.53453 +INFO:tensorflow:step = 1701, loss = 1.60527, precision = 0.765625 (65.166 sec) +INFO:tensorflow:global_step/sec: 1.53467 +INFO:tensorflow:step = 1801, loss = 1.53635, precision = 0.789062 (65.160 sec) +INFO:tensorflow:global_step/sec: 1.53457 +INFO:tensorflow:step = 1901, loss = 1.54984, precision = 0.78125 (65.165 sec) +Saved checkpoint after 5 epoch(s) to data/resnet164/checkpoints/00005... +INFO:tensorflow:global_step/sec: 1.48232 +INFO:tensorflow:step = 2001, loss = 1.36699, precision = 0.804688 (67.462 sec) +INFO:tensorflow:global_step/sec: 1.53462 +INFO:tensorflow:step = 2101, loss = 1.30302, precision = 0.789062 (65.163 sec) +INFO:tensorflow:global_step/sec: 1.5342 +INFO:tensorflow:step = 2201, loss = 1.36564, precision = 0.789062 (65.181 sec) +INFO:tensorflow:global_step/sec: 1.53384 +INFO:tensorflow:step = 2301, loss = 1.445, precision = 0.742188 (65.196 sec) +Saved checkpoint after 6 epoch(s) to data/resnet164/checkpoints/00006... +INFO:tensorflow:global_step/sec: 1.48274 +INFO:tensorflow:step = 2401, loss = 1.36562, precision = 0.78125 (67.443 sec) +INFO:tensorflow:global_step/sec: 1.53561 +INFO:tensorflow:step = 2501, loss = 1.04024, precision = 0.8125 (65.121 sec) +INFO:tensorflow:global_step/sec: 1.53582 +INFO:tensorflow:step = 2601, loss = 1.19101, precision = 0.75 (65.112 sec) +INFO:tensorflow:global_step/sec: 1.53428 +INFO:tensorflow:step = 2701, loss = 0.894505, precision = 0.875 (65.177 sec) +Saved checkpoint after 7 epoch(s) to data/resnet164/checkpoints/00007... +INFO:tensorflow:global_step/sec: 1.48285 +INFO:tensorflow:step = 2801, loss = 1.31629, precision = 0.710938 (67.438 sec) +INFO:tensorflow:global_step/sec: 1.53483 +INFO:tensorflow:step = 2901, loss = 1.20145, precision = 0.742188 (65.154 sec) +INFO:tensorflow:global_step/sec: 1.53454 +INFO:tensorflow:step = 3001, loss = 1.08642, precision = 0.773438 (65.166 sec) +INFO:tensorflow:global_step/sec: 1.53467 +INFO:tensorflow:step = 3101, loss = 0.904929, precision = 0.8125 (65.161 sec) +Saved checkpoint after 8 epoch(s) to data/resnet164/checkpoints/00008... +INFO:tensorflow:global_step/sec: 1.47818 +INFO:tensorflow:step = 3201, loss = 0.848977, precision = 0.859375 (67.651 sec) +INFO:tensorflow:global_step/sec: 1.53477 +INFO:tensorflow:step = 3301, loss = 0.901198, precision = 0.84375 (65.156 sec) +INFO:tensorflow:global_step/sec: 1.53539 +INFO:tensorflow:step = 3401, loss = 0.925471, precision = 0.820312 (65.130 sec) +INFO:tensorflow:global_step/sec: 1.53507 +INFO:tensorflow:step = 3501, loss = 1.04695, precision = 0.757812 (65.144 sec) +Saved checkpoint after 9 epoch(s) to data/resnet164/checkpoints/00009... +INFO:tensorflow:global_step/sec: 1.48362 +INFO:tensorflow:step = 3601, loss = 0.785341, precision = 0.851562 (67.403 sec) +INFO:tensorflow:global_step/sec: 1.53572 +INFO:tensorflow:step = 3701, loss = 0.841788, precision = 0.851562 (65.116 sec) +INFO:tensorflow:global_step/sec: 1.53443 +INFO:tensorflow:step = 3801, loss = 0.774006, precision = 0.84375 (65.171 sec) +INFO:tensorflow:global_step/sec: 1.5347 +INFO:tensorflow:step = 3901, loss = 0.776803, precision = 0.867188 (65.160 sec) +Saved checkpoint after 10 epoch(s) to data/resnet164/checkpoints/00010... +INFO:tensorflow:global_step/sec: 1.48413 +INFO:tensorflow:step = 4001, loss = 0.838103, precision = 0.835938 (67.379 sec) +INFO:tensorflow:global_step/sec: 1.53533 +INFO:tensorflow:step = 4101, loss = 0.886179, precision = 0.734375 (65.132 sec) +INFO:tensorflow:global_step/sec: 1.53559 +INFO:tensorflow:step = 4201, loss = 0.911013, precision = 0.8125 (65.121 sec) +Saved checkpoint after 11 epoch(s) to data/resnet164/checkpoints/00011... +INFO:tensorflow:global_step/sec: 1.48495 +INFO:tensorflow:step = 4301, loss = 0.696146, precision = 0.890625 (67.342 sec) +INFO:tensorflow:global_step/sec: 1.53587 +INFO:tensorflow:step = 4401, loss = 0.690876, precision = 0.898438 (65.110 sec) +INFO:tensorflow:global_step/sec: 1.53586 +INFO:tensorflow:step = 4501, loss = 0.862145, precision = 0.820312 (65.110 sec) +INFO:tensorflow:global_step/sec: 1.536 +INFO:tensorflow:step = 4601, loss = 0.7374, precision = 0.859375 (65.104 sec) +Saved checkpoint after 12 epoch(s) to data/resnet164/checkpoints/00012... +INFO:tensorflow:global_step/sec: 1.4849 +INFO:tensorflow:step = 4701, loss = 0.807036, precision = 0.820312 (67.345 sec) +INFO:tensorflow:global_step/sec: 1.53607 +INFO:tensorflow:step = 4801, loss = 0.797251, precision = 0.828125 (65.101 sec) +INFO:tensorflow:global_step/sec: 1.53551 +INFO:tensorflow:step = 4901, loss = 0.880211, precision = 0.796875 (65.125 sec) +INFO:tensorflow:global_step/sec: 1.53513 +INFO:tensorflow:step = 5001, loss = 0.762536, precision = 0.867188 (65.141 sec) +Saved checkpoint after 13 epoch(s) to data/resnet164/checkpoints/00013... +INFO:tensorflow:global_step/sec: 1.48489 +INFO:tensorflow:step = 5101, loss = 0.846517, precision = 0.820312 (67.345 sec) +INFO:tensorflow:global_step/sec: 1.5359 +INFO:tensorflow:step = 5201, loss = 0.78557, precision = 0.820312 (65.108 sec) +INFO:tensorflow:global_step/sec: 1.53625 +INFO:tensorflow:step = 5301, loss = 0.831104, precision = 0.804688 (65.094 sec) +INFO:tensorflow:global_step/sec: 1.53604 +INFO:tensorflow:step = 5401, loss = 0.803595, precision = 0.828125 (65.102 sec) +Saved checkpoint after 14 epoch(s) to data/resnet164/checkpoints/00014... +INFO:tensorflow:global_step/sec: 1.48557 +INFO:tensorflow:step = 5501, loss = 0.65618, precision = 0.890625 (67.314 sec) +INFO:tensorflow:global_step/sec: 1.53642 +INFO:tensorflow:step = 5601, loss = 0.727976, precision = 0.882812 (65.086 sec) +INFO:tensorflow:global_step/sec: 1.53644 +INFO:tensorflow:step = 5701, loss = 0.908297, precision = 0.835938 (65.085 sec) +INFO:tensorflow:global_step/sec: 1.53712 +INFO:tensorflow:step = 5801, loss = 0.604747, precision = 0.898438 (65.057 sec) +Saved checkpoint after 15 epoch(s) to data/resnet164/checkpoints/00015... +INFO:tensorflow:global_step/sec: 1.48504 +INFO:tensorflow:step = 5901, loss = 0.645811, precision = 0.914062 (67.338 sec) +INFO:tensorflow:global_step/sec: 1.53722 +INFO:tensorflow:step = 6001, loss = 0.699435, precision = 0.867188 (65.053 sec) +INFO:tensorflow:global_step/sec: 1.53648 +INFO:tensorflow:step = 6101, loss = 0.741522, precision = 0.851562 (65.084 sec) +INFO:tensorflow:global_step/sec: 1.53619 +INFO:tensorflow:step = 6201, loss = 0.63104, precision = 0.882812 (65.096 sec) +Saved checkpoint after 16 epoch(s) to data/resnet164/checkpoints/00016... +INFO:tensorflow:global_step/sec: 1.4855 +INFO:tensorflow:step = 6301, loss = 0.723512, precision = 0.867188 (67.317 sec) +INFO:tensorflow:global_step/sec: 1.53615 +INFO:tensorflow:step = 6401, loss = 0.709199, precision = 0.851562 (65.097 sec) +INFO:tensorflow:global_step/sec: 1.53653 +INFO:tensorflow:step = 6501, loss = 0.714786, precision = 0.835938 (65.082 sec) +INFO:tensorflow:global_step/sec: 1.5371 +INFO:tensorflow:step = 6601, loss = 0.79697, precision = 0.789062 (65.058 sec) +Saved checkpoint after 17 epoch(s) to data/resnet164/checkpoints/00017... +INFO:tensorflow:global_step/sec: 1.48545 +INFO:tensorflow:step = 6701, loss = 0.528376, precision = 0.929688 (67.320 sec) +INFO:tensorflow:global_step/sec: 1.53591 +INFO:tensorflow:step = 6801, loss = 0.704986, precision = 0.875 (65.108 sec) +INFO:tensorflow:global_step/sec: 1.537 +INFO:tensorflow:step = 6901, loss = 0.713222, precision = 0.851562 (65.062 sec) +INFO:tensorflow:global_step/sec: 1.53712 +INFO:tensorflow:step = 7001, loss = 0.846533, precision = 0.84375 (65.056 sec) +Saved checkpoint after 18 epoch(s) to data/resnet164/checkpoints/00018... +INFO:tensorflow:global_step/sec: 1.48005 +INFO:tensorflow:step = 7101, loss = 0.681155, precision = 0.851562 (67.566 sec) +INFO:tensorflow:global_step/sec: 1.53694 +INFO:tensorflow:step = 7201, loss = 0.862231, precision = 0.84375 (65.064 sec) +INFO:tensorflow:global_step/sec: 1.53637 +INFO:tensorflow:step = 7301, loss = 0.661776, precision = 0.875 (65.088 sec) +INFO:tensorflow:global_step/sec: 1.53721 +INFO:tensorflow:step = 7401, loss = 0.725801, precision = 0.835938 (65.053 sec) +Saved checkpoint after 19 epoch(s) to data/resnet164/checkpoints/00019... +INFO:tensorflow:global_step/sec: 1.48566 +INFO:tensorflow:step = 7501, loss = 0.770155, precision = 0.835938 (67.310 sec) +INFO:tensorflow:global_step/sec: 1.53575 +INFO:tensorflow:step = 7601, loss = 0.728164, precision = 0.835938 (65.114 sec) +INFO:tensorflow:global_step/sec: 1.53673 +INFO:tensorflow:step = 7701, loss = 0.91495, precision = 0.835938 (65.073 sec) +INFO:tensorflow:global_step/sec: 1.53584 +INFO:tensorflow:step = 7801, loss = 0.787203, precision = 0.820312 (65.111 sec) +Saved checkpoint after 20 epoch(s) to data/resnet164/checkpoints/00020... +INFO:tensorflow:global_step/sec: 1.48575 +INFO:tensorflow:step = 7901, loss = 0.768345, precision = 0.789062 (67.306 sec) +INFO:tensorflow:global_step/sec: 1.53646 +INFO:tensorflow:step = 8001, loss = 0.723351, precision = 0.851562 (65.084 sec) +INFO:tensorflow:global_step/sec: 1.53615 +INFO:tensorflow:step = 8101, loss = 0.669478, precision = 0.875 (65.098 sec) +INFO:tensorflow:global_step/sec: 1.53676 +INFO:tensorflow:step = 8201, loss = 0.606217, precision = 0.90625 (65.072 sec) +Saved checkpoint after 21 epoch(s) to data/resnet164/checkpoints/00021... +INFO:tensorflow:global_step/sec: 1.48675 +INFO:tensorflow:step = 8301, loss = 0.725701, precision = 0.859375 (67.261 sec) +INFO:tensorflow:global_step/sec: 1.53628 +INFO:tensorflow:step = 8401, loss = 0.625222, precision = 0.898438 (65.092 sec) +INFO:tensorflow:global_step/sec: 1.53714 +INFO:tensorflow:step = 8501, loss = 0.870172, precision = 0.773438 (65.056 sec) +INFO:tensorflow:global_step/sec: 1.53694 +INFO:tensorflow:step = 8601, loss = 0.601821, precision = 0.90625 (65.064 sec) +Saved checkpoint after 22 epoch(s) to data/resnet164/checkpoints/00022... +INFO:tensorflow:global_step/sec: 1.48578 +INFO:tensorflow:step = 8701, loss = 0.584834, precision = 0.898438 (67.305 sec) +INFO:tensorflow:global_step/sec: 1.53646 +INFO:tensorflow:step = 8801, loss = 0.697999, precision = 0.882812 (65.085 sec) +INFO:tensorflow:global_step/sec: 1.53611 +INFO:tensorflow:step = 8901, loss = 0.745883, precision = 0.867188 (65.099 sec) +Saved checkpoint after 23 epoch(s) to data/resnet164/checkpoints/00023... +INFO:tensorflow:global_step/sec: 1.48599 +INFO:tensorflow:step = 9001, loss = 0.707493, precision = 0.875 (67.295 sec) +INFO:tensorflow:global_step/sec: 1.53723 +INFO:tensorflow:step = 9101, loss = 0.652854, precision = 0.859375 (65.052 sec) +INFO:tensorflow:global_step/sec: 1.53567 +INFO:tensorflow:step = 9201, loss = 0.62858, precision = 0.890625 (65.118 sec) +INFO:tensorflow:global_step/sec: 1.53696 +INFO:tensorflow:step = 9301, loss = 0.909386, precision = 0.789062 (65.064 sec) +Saved checkpoint after 24 epoch(s) to data/resnet164/checkpoints/00024... +INFO:tensorflow:global_step/sec: 1.48541 +INFO:tensorflow:step = 9401, loss = 0.694637, precision = 0.859375 (67.321 sec) +INFO:tensorflow:global_step/sec: 1.53703 +INFO:tensorflow:step = 9501, loss = 0.629115, precision = 0.875 (65.060 sec) +INFO:tensorflow:global_step/sec: 1.53653 +INFO:tensorflow:step = 9601, loss = 0.728139, precision = 0.859375 (65.082 sec) +INFO:tensorflow:global_step/sec: 1.53555 +INFO:tensorflow:step = 9701, loss = 0.647743, precision = 0.882812 (65.123 sec) +Saved checkpoint after 25 epoch(s) to data/resnet164/checkpoints/00025... +INFO:tensorflow:global_step/sec: 1.48612 +INFO:tensorflow:step = 9801, loss = 0.769875, precision = 0.828125 (67.289 sec) +INFO:tensorflow:global_step/sec: 1.53684 +INFO:tensorflow:step = 9901, loss = 0.714496, precision = 0.875 (65.069 sec) +INFO:tensorflow:global_step/sec: 1.53678 +INFO:tensorflow:step = 10001, loss = 0.731602, precision = 0.84375 (65.071 sec) +INFO:tensorflow:global_step/sec: 1.53658 +INFO:tensorflow:step = 10101, loss = 0.748687, precision = 0.851562 (65.080 sec) +Saved checkpoint after 26 epoch(s) to data/resnet164/checkpoints/00026... +INFO:tensorflow:global_step/sec: 1.4855 +INFO:tensorflow:step = 10201, loss = 0.773319, precision = 0.828125 (67.317 sec) +INFO:tensorflow:global_step/sec: 1.53662 +INFO:tensorflow:step = 10301, loss = 0.736553, precision = 0.835938 (65.078 sec) +INFO:tensorflow:global_step/sec: 1.53672 +INFO:tensorflow:step = 10401, loss = 0.672988, precision = 0.859375 (65.074 sec) +INFO:tensorflow:global_step/sec: 1.53648 +INFO:tensorflow:step = 10501, loss = 0.747988, precision = 0.851562 (65.084 sec) +Saved checkpoint after 27 epoch(s) to data/resnet164/checkpoints/00027... +INFO:tensorflow:global_step/sec: 1.48545 +INFO:tensorflow:step = 10601, loss = 0.628114, precision = 0.898438 (67.320 sec) +INFO:tensorflow:global_step/sec: 1.53697 +INFO:tensorflow:step = 10701, loss = 0.678666, precision = 0.84375 (65.063 sec) +INFO:tensorflow:global_step/sec: 1.53685 +INFO:tensorflow:step = 10801, loss = 0.871938, precision = 0.8125 (65.068 sec) +INFO:tensorflow:global_step/sec: 1.53644 +INFO:tensorflow:step = 10901, loss = 0.736933, precision = 0.835938 (65.086 sec) +Saved checkpoint after 28 epoch(s) to data/resnet164/checkpoints/00028... +INFO:tensorflow:global_step/sec: 1.47879 +INFO:tensorflow:step = 11001, loss = 0.672743, precision = 0.859375 (67.623 sec) +INFO:tensorflow:global_step/sec: 1.53633 +INFO:tensorflow:step = 11101, loss = 0.659466, precision = 0.914062 (65.090 sec) +INFO:tensorflow:global_step/sec: 1.5372 +INFO:tensorflow:step = 11201, loss = 0.716083, precision = 0.875 (65.053 sec) +INFO:tensorflow:global_step/sec: 1.53646 +INFO:tensorflow:step = 11301, loss = 0.655226, precision = 0.890625 (65.085 sec) +Saved checkpoint after 29 epoch(s) to data/resnet164/checkpoints/00029... +INFO:tensorflow:global_step/sec: 1.4842 +INFO:tensorflow:step = 11401, loss = 0.667466, precision = 0.859375 (67.376 sec) +INFO:tensorflow:global_step/sec: 1.53616 +INFO:tensorflow:step = 11501, loss = 0.644609, precision = 0.835938 (65.098 sec) +INFO:tensorflow:global_step/sec: 1.53606 +INFO:tensorflow:step = 11601, loss = 0.614121, precision = 0.898438 (65.101 sec) +INFO:tensorflow:global_step/sec: 1.53625 +INFO:tensorflow:step = 11701, loss = 0.632079, precision = 0.875 (65.094 sec) +Saved checkpoint after 30 epoch(s) to data/resnet164/checkpoints/00030... +INFO:tensorflow:global_step/sec: 1.48538 +INFO:tensorflow:step = 11801, loss = 0.759186, precision = 0.84375 (67.323 sec) +INFO:tensorflow:global_step/sec: 1.5366 +INFO:tensorflow:step = 11901, loss = 0.69399, precision = 0.867188 (65.079 sec) +INFO:tensorflow:global_step/sec: 1.5365 +INFO:tensorflow:step = 12001, loss = 0.603905, precision = 0.890625 (65.083 sec) +INFO:tensorflow:global_step/sec: 1.53658 +INFO:tensorflow:step = 12101, loss = 0.591376, precision = 0.898438 (65.080 sec) +Saved checkpoint after 31 epoch(s) to data/resnet164/checkpoints/00031... +INFO:tensorflow:global_step/sec: 1.48437 +INFO:tensorflow:step = 12201, loss = 0.75086, precision = 0.875 (67.369 sec) +INFO:tensorflow:global_step/sec: 1.5361 +INFO:tensorflow:step = 12301, loss = 0.838497, precision = 0.835938 (65.100 sec) +INFO:tensorflow:global_step/sec: 1.5367 +INFO:tensorflow:step = 12401, loss = 0.656594, precision = 0.882812 (65.075 sec) +INFO:tensorflow:global_step/sec: 1.53531 +INFO:tensorflow:step = 12501, loss = 0.718711, precision = 0.851562 (65.133 sec) +Saved checkpoint after 32 epoch(s) to data/resnet164/checkpoints/00032... +INFO:tensorflow:global_step/sec: 1.48589 +INFO:tensorflow:step = 12601, loss = 0.70332, precision = 0.875 (67.300 sec) +INFO:tensorflow:global_step/sec: 1.53615 +INFO:tensorflow:step = 12701, loss = 0.681721, precision = 0.828125 (65.098 sec) +INFO:tensorflow:global_step/sec: 1.53605 +INFO:tensorflow:step = 12801, loss = 0.675054, precision = 0.882812 (65.102 sec) +INFO:tensorflow:global_step/sec: 1.53583 +INFO:tensorflow:step = 12901, loss = 0.648492, precision = 0.875 (65.112 sec) +Saved checkpoint after 33 epoch(s) to data/resnet164/checkpoints/00033... +INFO:tensorflow:global_step/sec: 1.48551 +INFO:tensorflow:step = 13001, loss = 0.607357, precision = 0.90625 (67.317 sec) +INFO:tensorflow:global_step/sec: 1.53675 +INFO:tensorflow:step = 13101, loss = 0.673828, precision = 0.875 (65.072 sec) +INFO:tensorflow:global_step/sec: 1.53581 +INFO:tensorflow:step = 13201, loss = 0.543284, precision = 0.929688 (65.112 sec) +Saved checkpoint after 34 epoch(s) to data/resnet164/checkpoints/00034... +INFO:tensorflow:global_step/sec: 1.48602 +INFO:tensorflow:step = 13301, loss = 0.787341, precision = 0.8125 (67.294 sec) +INFO:tensorflow:global_step/sec: 1.5367 +INFO:tensorflow:step = 13401, loss = 0.646139, precision = 0.882812 (65.075 sec) +INFO:tensorflow:global_step/sec: 1.53695 +INFO:tensorflow:step = 13501, loss = 0.606904, precision = 0.875 (65.064 sec) +INFO:tensorflow:global_step/sec: 1.5359 +INFO:tensorflow:step = 13601, loss = 0.698694, precision = 0.882812 (65.108 sec) +Saved checkpoint after 35 epoch(s) to data/resnet164/checkpoints/00035... +INFO:tensorflow:global_step/sec: 1.48464 +INFO:tensorflow:step = 13701, loss = 0.676161, precision = 0.898438 (67.357 sec) +INFO:tensorflow:global_step/sec: 1.53633 +INFO:tensorflow:step = 13801, loss = 0.606655, precision = 0.90625 (65.090 sec) +INFO:tensorflow:global_step/sec: 1.53728 +INFO:tensorflow:step = 13901, loss = 0.611004, precision = 0.914062 (65.050 sec) +INFO:tensorflow:global_step/sec: 1.53672 +INFO:tensorflow:step = 14001, loss = 0.74484, precision = 0.859375 (65.074 sec) +Saved checkpoint after 36 epoch(s) to data/resnet164/checkpoints/00036... +INFO:tensorflow:global_step/sec: 1.48579 +INFO:tensorflow:step = 14101, loss = 0.668738, precision = 0.867188 (67.304 sec) +INFO:tensorflow:global_step/sec: 1.53665 +INFO:tensorflow:step = 14201, loss = 0.70191, precision = 0.867188 (65.076 sec) +INFO:tensorflow:global_step/sec: 1.53678 +INFO:tensorflow:step = 14301, loss = 0.730742, precision = 0.859375 (65.071 sec) +INFO:tensorflow:global_step/sec: 1.53704 +INFO:tensorflow:step = 14401, loss = 0.582803, precision = 0.914062 (65.060 sec) +Saved checkpoint after 37 epoch(s) to data/resnet164/checkpoints/00037... +INFO:tensorflow:global_step/sec: 1.48725 +INFO:tensorflow:step = 14501, loss = 0.600601, precision = 0.90625 (67.239 sec) +INFO:tensorflow:global_step/sec: 1.53741 +INFO:tensorflow:step = 14601, loss = 0.582864, precision = 0.898438 (65.044 sec) +INFO:tensorflow:global_step/sec: 1.53667 +INFO:tensorflow:step = 14701, loss = 0.711707, precision = 0.851562 (65.076 sec) +INFO:tensorflow:global_step/sec: 1.53715 +INFO:tensorflow:step = 14801, loss = 0.601084, precision = 0.898438 (65.055 sec) +Saved checkpoint after 38 epoch(s) to data/resnet164/checkpoints/00038... +INFO:tensorflow:global_step/sec: 1.48056 +INFO:tensorflow:step = 14901, loss = 0.648659, precision = 0.875 (67.551 sec) +INFO:tensorflow:global_step/sec: 1.53678 +INFO:tensorflow:step = 15001, loss = 0.649462, precision = 0.898438 (65.062 sec) +INFO:tensorflow:global_step/sec: 1.5367 +INFO:tensorflow:step = 15101, loss = 0.587804, precision = 0.875 (65.075 sec) +INFO:tensorflow:global_step/sec: 1.53769 +INFO:tensorflow:step = 15201, loss = 0.663032, precision = 0.890625 (65.033 sec) +Saved checkpoint after 39 epoch(s) to data/resnet164/checkpoints/00039... +INFO:tensorflow:global_step/sec: 1.48578 +INFO:tensorflow:step = 15301, loss = 0.771988, precision = 0.875 (67.305 sec) +INFO:tensorflow:global_step/sec: 1.53678 +INFO:tensorflow:step = 15401, loss = 0.725884, precision = 0.898438 (65.071 sec) +INFO:tensorflow:global_step/sec: 1.53646 +INFO:tensorflow:step = 15501, loss = 0.682203, precision = 0.851562 (65.085 sec) +INFO:tensorflow:global_step/sec: 1.53754 +INFO:tensorflow:step = 15601, loss = 0.624827, precision = 0.90625 (65.039 sec) +Saved checkpoint after 40 epoch(s) to data/resnet164/checkpoints/00040... +INFO:tensorflow:global_step/sec: 1.4846 +INFO:tensorflow:step = 15701, loss = 0.73851, precision = 0.835938 (67.358 sec) +INFO:tensorflow:global_step/sec: 1.53697 +INFO:tensorflow:step = 15801, loss = 0.600834, precision = 0.898438 (65.063 sec) +INFO:tensorflow:global_step/sec: 1.5365 +INFO:tensorflow:step = 15901, loss = 0.626756, precision = 0.851562 (65.083 sec) +INFO:tensorflow:global_step/sec: 1.53675 +INFO:tensorflow:step = 16001, loss = 0.678833, precision = 0.851562 (65.072 sec) +Saved checkpoint after 41 epoch(s) to data/resnet164/checkpoints/00041... +INFO:tensorflow:global_step/sec: 1.48671 +INFO:tensorflow:step = 16101, loss = 0.593004, precision = 0.890625 (67.263 sec) +INFO:tensorflow:global_step/sec: 1.53806 +INFO:tensorflow:step = 16201, loss = 0.600034, precision = 0.898438 (65.017 sec) +INFO:tensorflow:global_step/sec: 1.53726 +INFO:tensorflow:step = 16301, loss = 0.725094, precision = 0.851562 (65.051 sec) +INFO:tensorflow:global_step/sec: 1.53717 +INFO:tensorflow:step = 16401, loss = 0.68474, precision = 0.890625 (65.055 sec) +Saved checkpoint after 42 epoch(s) to data/resnet164/checkpoints/00042... +INFO:tensorflow:global_step/sec: 1.48628 +INFO:tensorflow:step = 16501, loss = 0.681057, precision = 0.890625 (67.282 sec) +INFO:tensorflow:global_step/sec: 1.53706 +INFO:tensorflow:step = 16601, loss = 0.635163, precision = 0.882812 (65.059 sec) +INFO:tensorflow:global_step/sec: 1.53747 +INFO:tensorflow:step = 16701, loss = 0.878721, precision = 0.8125 (65.042 sec) +INFO:tensorflow:global_step/sec: 1.53702 +INFO:tensorflow:step = 16801, loss = 0.717201, precision = 0.875 (65.061 sec) +Saved checkpoint after 43 epoch(s) to data/resnet164/checkpoints/00043... +INFO:tensorflow:global_step/sec: 1.48585 +INFO:tensorflow:step = 16901, loss = 0.721952, precision = 0.867188 (67.302 sec) +INFO:tensorflow:global_step/sec: 1.53762 +INFO:tensorflow:step = 17001, loss = 0.578133, precision = 0.929688 (65.036 sec) +INFO:tensorflow:global_step/sec: 1.53698 +INFO:tensorflow:step = 17101, loss = 0.58972, precision = 0.898438 (65.063 sec) +INFO:tensorflow:global_step/sec: 1.53785 +INFO:tensorflow:step = 17201, loss = 0.708688, precision = 0.859375 (65.026 sec) +Saved checkpoint after 44 epoch(s) to data/resnet164/checkpoints/00044... +INFO:tensorflow:global_step/sec: 1.48572 +INFO:tensorflow:step = 17301, loss = 0.562343, precision = 0.921875 (67.308 sec) +INFO:tensorflow:global_step/sec: 1.53682 +INFO:tensorflow:step = 17401, loss = 0.562922, precision = 0.914062 (65.069 sec) +INFO:tensorflow:global_step/sec: 1.53645 +INFO:tensorflow:step = 17501, loss = 0.638552, precision = 0.890625 (65.085 sec) +Saved checkpoint after 45 epoch(s) to data/resnet164/checkpoints/00045... +INFO:tensorflow:global_step/sec: 1.48495 +INFO:tensorflow:step = 17601, loss = 0.579459, precision = 0.914062 (67.342 sec) +INFO:tensorflow:global_step/sec: 1.53691 +INFO:tensorflow:step = 17701, loss = 0.638716, precision = 0.875 (65.066 sec) +INFO:tensorflow:global_step/sec: 1.53683 +INFO:tensorflow:step = 17801, loss = 0.651956, precision = 0.898438 (65.069 sec) +INFO:tensorflow:global_step/sec: 1.53741 +INFO:tensorflow:step = 17901, loss = 0.540421, precision = 0.929688 (65.045 sec) +Saved checkpoint after 46 epoch(s) to data/resnet164/checkpoints/00046... +INFO:tensorflow:global_step/sec: 1.48598 +INFO:tensorflow:step = 18001, loss = 0.641053, precision = 0.867188 (67.296 sec) +INFO:tensorflow:global_step/sec: 1.53696 +INFO:tensorflow:step = 18101, loss = 0.458401, precision = 0.960938 (65.064 sec) +INFO:tensorflow:global_step/sec: 1.53774 +INFO:tensorflow:step = 18201, loss = 0.644515, precision = 0.875 (65.030 sec) +INFO:tensorflow:global_step/sec: 1.53636 +INFO:tensorflow:step = 18301, loss = 0.739301, precision = 0.835938 (65.089 sec) +Saved checkpoint after 47 epoch(s) to data/resnet164/checkpoints/00047... +INFO:tensorflow:global_step/sec: 1.48593 +INFO:tensorflow:step = 18401, loss = 0.532901, precision = 0.945312 (67.298 sec) +INFO:tensorflow:global_step/sec: 1.5367 +INFO:tensorflow:step = 18501, loss = 0.778647, precision = 0.8125 (65.074 sec) +INFO:tensorflow:global_step/sec: 1.53598 +INFO:tensorflow:step = 18601, loss = 0.812563, precision = 0.859375 (65.105 sec) +INFO:tensorflow:global_step/sec: 1.53681 +INFO:tensorflow:step = 18701, loss = 0.613423, precision = 0.898438 (65.070 sec) +Saved checkpoint after 48 epoch(s) to data/resnet164/checkpoints/00048... +INFO:tensorflow:global_step/sec: 1.48457 +INFO:tensorflow:step = 18801, loss = 0.649234, precision = 0.890625 (67.360 sec) +INFO:tensorflow:global_step/sec: 1.53644 +INFO:tensorflow:step = 18901, loss = 0.612509, precision = 0.890625 (65.086 sec) +INFO:tensorflow:global_step/sec: 1.53658 +INFO:tensorflow:step = 19001, loss = 0.687881, precision = 0.890625 (65.080 sec) +INFO:tensorflow:global_step/sec: 1.53701 +INFO:tensorflow:step = 19101, loss = 0.68333, precision = 0.851562 (65.062 sec) +Saved checkpoint after 49 epoch(s) to data/resnet164/checkpoints/00049... +INFO:tensorflow:global_step/sec: 1.47963 +INFO:tensorflow:step = 19201, loss = 0.562811, precision = 0.914062 (67.585 sec) +INFO:tensorflow:global_step/sec: 1.53582 +INFO:tensorflow:step = 19301, loss = 0.593888, precision = 0.914062 (65.112 sec) +INFO:tensorflow:global_step/sec: 1.53699 +INFO:tensorflow:step = 19401, loss = 0.624757, precision = 0.859375 (65.062 sec) +INFO:tensorflow:global_step/sec: 1.53641 +INFO:tensorflow:step = 19501, loss = 0.679996, precision = 0.867188 (65.087 sec) +Saved checkpoint after 50 epoch(s) to data/resnet164/checkpoints/00050... +INFO:tensorflow:global_step/sec: 1.48449 +INFO:tensorflow:step = 19601, loss = 0.730495, precision = 0.875 (67.363 sec) +INFO:tensorflow:global_step/sec: 1.53565 +INFO:tensorflow:step = 19701, loss = 0.677334, precision = 0.898438 (65.119 sec) +INFO:tensorflow:global_step/sec: 1.53685 +INFO:tensorflow:step = 19801, loss = 0.777845, precision = 0.859375 (65.068 sec) +INFO:tensorflow:global_step/sec: 1.53627 +INFO:tensorflow:step = 19901, loss = 0.603952, precision = 0.90625 (65.093 sec) +Saved checkpoint after 51 epoch(s) to data/resnet164/checkpoints/00051... +INFO:tensorflow:global_step/sec: 1.48526 +INFO:tensorflow:step = 20001, loss = 0.526926, precision = 0.9375 (67.328 sec) +INFO:tensorflow:global_step/sec: 1.53574 +INFO:tensorflow:step = 20101, loss = 0.671032, precision = 0.875 (65.115 sec) +INFO:tensorflow:global_step/sec: 1.53653 +INFO:tensorflow:step = 20201, loss = 0.656096, precision = 0.875 (65.082 sec) +INFO:tensorflow:global_step/sec: 1.53487 +INFO:tensorflow:step = 20301, loss = 0.655471, precision = 0.890625 (65.152 sec) +Saved checkpoint after 52 epoch(s) to data/resnet164/checkpoints/00052... +INFO:tensorflow:global_step/sec: 1.48586 +INFO:tensorflow:step = 20401, loss = 0.59329, precision = 0.898438 (67.301 sec) +INFO:tensorflow:global_step/sec: 1.53685 +INFO:tensorflow:step = 20501, loss = 0.714309, precision = 0.875 (65.068 sec) +INFO:tensorflow:global_step/sec: 1.53641 +INFO:tensorflow:step = 20601, loss = 0.673456, precision = 0.882812 (65.087 sec) +INFO:tensorflow:global_step/sec: 1.53543 +INFO:tensorflow:step = 20701, loss = 0.607041, precision = 0.882812 (65.129 sec) +Saved checkpoint after 53 epoch(s) to data/resnet164/checkpoints/00053... +INFO:tensorflow:global_step/sec: 1.48466 +INFO:tensorflow:step = 20801, loss = 0.735606, precision = 0.84375 (67.356 sec) +INFO:tensorflow:global_step/sec: 1.5355 +INFO:tensorflow:step = 20901, loss = 0.613577, precision = 0.882812 (65.125 sec) +INFO:tensorflow:global_step/sec: 1.53652 +INFO:tensorflow:step = 21001, loss = 0.763079, precision = 0.84375 (65.082 sec) +INFO:tensorflow:global_step/sec: 1.53606 +INFO:tensorflow:step = 21101, loss = 0.640737, precision = 0.875 (65.102 sec) +Saved checkpoint after 54 epoch(s) to data/resnet164/checkpoints/00054... +INFO:tensorflow:global_step/sec: 1.4859 +INFO:tensorflow:step = 21201, loss = 0.661024, precision = 0.890625 (67.299 sec) +INFO:tensorflow:global_step/sec: 1.5356 +INFO:tensorflow:step = 21301, loss = 0.577548, precision = 0.921875 (65.121 sec) +INFO:tensorflow:global_step/sec: 1.53596 +INFO:tensorflow:step = 21401, loss = 0.682296, precision = 0.867188 (65.106 sec) +INFO:tensorflow:global_step/sec: 1.53617 +INFO:tensorflow:step = 21501, loss = 0.606051, precision = 0.882812 (65.097 sec) +Saved checkpoint after 55 epoch(s) to data/resnet164/checkpoints/00055... +INFO:tensorflow:global_step/sec: 1.4843 +INFO:tensorflow:step = 21601, loss = 0.700824, precision = 0.867188 (67.372 sec) +INFO:tensorflow:global_step/sec: 1.53563 +INFO:tensorflow:step = 21701, loss = 0.627267, precision = 0.882812 (65.120 sec) +INFO:tensorflow:global_step/sec: 1.53618 +INFO:tensorflow:step = 21801, loss = 0.595174, precision = 0.914062 (65.097 sec) +Saved checkpoint after 56 epoch(s) to data/resnet164/checkpoints/00056... +INFO:tensorflow:global_step/sec: 1.48429 +INFO:tensorflow:step = 21901, loss = 0.58022, precision = 0.914062 (67.373 sec) +INFO:tensorflow:global_step/sec: 1.53658 +INFO:tensorflow:step = 22001, loss = 0.596949, precision = 0.890625 (65.079 sec) +INFO:tensorflow:global_step/sec: 1.53583 +INFO:tensorflow:step = 22101, loss = 0.666202, precision = 0.875 (65.111 sec) +INFO:tensorflow:global_step/sec: 1.53612 +INFO:tensorflow:step = 22201, loss = 0.626375, precision = 0.882812 (65.099 sec) +Saved checkpoint after 57 epoch(s) to data/resnet164/checkpoints/00057... +INFO:tensorflow:global_step/sec: 1.4851 +INFO:tensorflow:step = 22301, loss = 0.712273, precision = 0.882812 (67.336 sec) +INFO:tensorflow:global_step/sec: 1.53755 +INFO:tensorflow:step = 22401, loss = 0.731736, precision = 0.851562 (65.038 sec) +INFO:tensorflow:global_step/sec: 1.53632 +INFO:tensorflow:step = 22501, loss = 0.616017, precision = 0.882812 (65.091 sec) +INFO:tensorflow:global_step/sec: 1.53612 +INFO:tensorflow:step = 22601, loss = 0.615794, precision = 0.914062 (65.099 sec) +Saved checkpoint after 58 epoch(s) to data/resnet164/checkpoints/00058... +INFO:tensorflow:global_step/sec: 1.48488 +INFO:tensorflow:step = 22701, loss = 0.527375, precision = 0.9375 (67.346 sec) +INFO:tensorflow:global_step/sec: 1.53556 +INFO:tensorflow:step = 22801, loss = 0.646324, precision = 0.890625 (65.123 sec) +INFO:tensorflow:global_step/sec: 1.53613 +INFO:tensorflow:step = 22901, loss = 0.83237, precision = 0.828125 (65.099 sec) +INFO:tensorflow:global_step/sec: 1.53533 +INFO:tensorflow:step = 23001, loss = 0.515942, precision = 0.90625 (65.132 sec) +Saved checkpoint after 59 epoch(s) to data/resnet164/checkpoints/00059... +INFO:tensorflow:global_step/sec: 1.47909 +INFO:tensorflow:step = 23101, loss = 0.658317, precision = 0.898438 (67.609 sec) +INFO:tensorflow:global_step/sec: 1.53656 +INFO:tensorflow:step = 23201, loss = 0.684257, precision = 0.882812 (65.080 sec) +INFO:tensorflow:global_step/sec: 1.53649 +INFO:tensorflow:step = 23301, loss = 0.765305, precision = 0.851562 (65.083 sec) +INFO:tensorflow:global_step/sec: 1.53638 +INFO:tensorflow:step = 23401, loss = 0.675543, precision = 0.898438 (65.088 sec) +Saved checkpoint after 60 epoch(s) to data/resnet164/checkpoints/00060... +INFO:tensorflow:global_step/sec: 1.4849 +INFO:tensorflow:step = 23501, loss = 0.671727, precision = 0.867188 (67.345 sec) +INFO:tensorflow:global_step/sec: 1.53682 +INFO:tensorflow:step = 23601, loss = 0.608586, precision = 0.882812 (65.069 sec) +INFO:tensorflow:global_step/sec: 1.53661 +INFO:tensorflow:step = 23701, loss = 0.61999, precision = 0.867188 (65.078 sec) +INFO:tensorflow:global_step/sec: 1.5359 +INFO:tensorflow:step = 23801, loss = 0.676512, precision = 0.90625 (65.108 sec) +Saved checkpoint after 61 epoch(s) to data/resnet164/checkpoints/00061... +INFO:tensorflow:global_step/sec: 1.48643 +INFO:tensorflow:step = 23901, loss = 0.608181, precision = 0.914062 (67.276 sec) +INFO:tensorflow:global_step/sec: 1.53592 +INFO:tensorflow:step = 24001, loss = 0.703835, precision = 0.882812 (65.108 sec) +INFO:tensorflow:global_step/sec: 1.53726 +INFO:tensorflow:step = 24101, loss = 0.640651, precision = 0.882812 (65.051 sec) +INFO:tensorflow:global_step/sec: 1.53666 +INFO:tensorflow:step = 24201, loss = 0.812827, precision = 0.84375 (65.076 sec) +Saved checkpoint after 62 epoch(s) to data/resnet164/checkpoints/00062... +INFO:tensorflow:global_step/sec: 1.48566 +INFO:tensorflow:step = 24301, loss = 0.614567, precision = 0.898438 (67.310 sec) +INFO:tensorflow:global_step/sec: 1.53637 +INFO:tensorflow:step = 24401, loss = 0.618375, precision = 0.90625 (65.088 sec) +INFO:tensorflow:global_step/sec: 1.53661 +INFO:tensorflow:step = 24501, loss = 0.591891, precision = 0.90625 (65.078 sec) +INFO:tensorflow:global_step/sec: 1.53758 +INFO:tensorflow:step = 24601, loss = 0.615695, precision = 0.914062 (65.037 sec) +Saved checkpoint after 63 epoch(s) to data/resnet164/checkpoints/00063... +INFO:tensorflow:global_step/sec: 1.48601 +INFO:tensorflow:step = 24701, loss = 0.686707, precision = 0.890625 (67.294 sec) +INFO:tensorflow:global_step/sec: 1.5368 +INFO:tensorflow:step = 24801, loss = 0.535973, precision = 0.945312 (65.070 sec) +INFO:tensorflow:global_step/sec: 1.53684 +INFO:tensorflow:step = 24901, loss = 0.679071, precision = 0.835938 (65.068 sec) +INFO:tensorflow:global_step/sec: 1.53579 +INFO:tensorflow:step = 25001, loss = 0.569047, precision = 0.90625 (65.113 sec) +Saved checkpoint after 64 epoch(s) to data/resnet164/checkpoints/00064... +INFO:tensorflow:global_step/sec: 1.48599 +INFO:tensorflow:step = 25101, loss = 0.734459, precision = 0.851562 (67.295 sec) +INFO:tensorflow:global_step/sec: 1.53477 +INFO:tensorflow:step = 25201, loss = 0.619124, precision = 0.890625 (65.156 sec) +INFO:tensorflow:global_step/sec: 1.53434 +INFO:tensorflow:step = 25301, loss = 0.741744, precision = 0.882812 (65.175 sec) +INFO:tensorflow:global_step/sec: 1.5328 +INFO:tensorflow:step = 25401, loss = 0.560725, precision = 0.921875 (65.240 sec) +Saved checkpoint after 65 epoch(s) to data/resnet164/checkpoints/00065... +INFO:tensorflow:global_step/sec: 1.48355 +INFO:tensorflow:step = 25501, loss = 0.627995, precision = 0.875 (67.406 sec) +INFO:tensorflow:global_step/sec: 1.53422 +INFO:tensorflow:step = 25601, loss = 0.595327, precision = 0.929688 (65.179 sec) +INFO:tensorflow:global_step/sec: 1.53319 +INFO:tensorflow:step = 25701, loss = 0.835489, precision = 0.828125 (65.224 sec) +INFO:tensorflow:global_step/sec: 1.53336 +INFO:tensorflow:step = 25801, loss = 0.588673, precision = 0.898438 (65.216 sec) +Saved checkpoint after 66 epoch(s) to data/resnet164/checkpoints/00066... +INFO:tensorflow:global_step/sec: 1.48385 +INFO:tensorflow:step = 25901, loss = 0.533161, precision = 0.929688 (67.392 sec) +INFO:tensorflow:global_step/sec: 1.5334 +INFO:tensorflow:step = 26001, loss = 0.726996, precision = 0.851562 (65.215 sec) +INFO:tensorflow:global_step/sec: 1.53453 +INFO:tensorflow:step = 26101, loss = 0.697504, precision = 0.890625 (65.166 sec) +Saved checkpoint after 67 epoch(s) to data/resnet164/checkpoints/00067... +INFO:tensorflow:global_step/sec: 1.48264 +INFO:tensorflow:step = 26201, loss = 0.598153, precision = 0.898438 (67.447 sec) +INFO:tensorflow:global_step/sec: 1.53543 +INFO:tensorflow:step = 26301, loss = 0.643059, precision = 0.875 (65.128 sec) +INFO:tensorflow:global_step/sec: 1.53476 +INFO:tensorflow:step = 26401, loss = 0.585417, precision = 0.898438 (65.157 sec) +INFO:tensorflow:global_step/sec: 1.53443 +INFO:tensorflow:step = 26501, loss = 0.610808, precision = 0.914062 (65.171 sec) +Saved checkpoint after 68 epoch(s) to data/resnet164/checkpoints/00068... +INFO:tensorflow:global_step/sec: 1.47983 +INFO:tensorflow:step = 26601, loss = 0.646767, precision = 0.890625 (67.575 sec) +INFO:tensorflow:global_step/sec: 1.53559 +INFO:tensorflow:step = 26701, loss = 0.600827, precision = 0.914062 (65.122 sec) +INFO:tensorflow:global_step/sec: 1.53542 +INFO:tensorflow:step = 26801, loss = 0.619399, precision = 0.867188 (65.128 sec) +INFO:tensorflow:global_step/sec: 1.53562 +INFO:tensorflow:step = 26901, loss = 0.576587, precision = 0.90625 (65.120 sec) +Saved checkpoint after 69 epoch(s) to data/resnet164/checkpoints/00069... +INFO:tensorflow:global_step/sec: 1.48623 +INFO:tensorflow:step = 27001, loss = 0.60461, precision = 0.898438 (67.284 sec) +INFO:tensorflow:global_step/sec: 1.53579 +INFO:tensorflow:step = 27101, loss = 0.578157, precision = 0.90625 (65.113 sec) +INFO:tensorflow:global_step/sec: 1.53613 +INFO:tensorflow:step = 27201, loss = 0.692511, precision = 0.898438 (65.098 sec) +INFO:tensorflow:global_step/sec: 1.53576 +INFO:tensorflow:step = 27301, loss = 0.519012, precision = 0.929688 (65.115 sec) +Saved checkpoint after 70 epoch(s) to data/resnet164/checkpoints/00070... +INFO:tensorflow:global_step/sec: 1.48648 +INFO:tensorflow:step = 27401, loss = 0.580644, precision = 0.90625 (67.273 sec) +INFO:tensorflow:global_step/sec: 1.53668 +INFO:tensorflow:step = 27501, loss = 0.549367, precision = 0.9375 (65.075 sec) +INFO:tensorflow:global_step/sec: 1.53564 +INFO:tensorflow:step = 27601, loss = 0.627777, precision = 0.890625 (65.119 sec) +INFO:tensorflow:global_step/sec: 1.53616 +INFO:tensorflow:step = 27701, loss = 0.592922, precision = 0.867188 (65.097 sec) +Saved checkpoint after 71 epoch(s) to data/resnet164/checkpoints/00071... +INFO:tensorflow:global_step/sec: 1.48545 +INFO:tensorflow:step = 27801, loss = 0.6318, precision = 0.929688 (67.320 sec) +INFO:tensorflow:global_step/sec: 1.53609 +INFO:tensorflow:step = 27901, loss = 0.695166, precision = 0.867188 (65.100 sec) +INFO:tensorflow:global_step/sec: 1.53703 +INFO:tensorflow:step = 28001, loss = 0.67606, precision = 0.890625 (65.060 sec) +INFO:tensorflow:global_step/sec: 1.53677 +INFO:tensorflow:step = 28101, loss = 0.62994, precision = 0.882812 (65.071 sec) +Saved checkpoint after 72 epoch(s) to data/resnet164/checkpoints/00072... +INFO:tensorflow:global_step/sec: 1.48756 +INFO:tensorflow:step = 28201, loss = 0.724036, precision = 0.875 (67.224 sec) +INFO:tensorflow:global_step/sec: 1.53652 +INFO:tensorflow:step = 28301, loss = 0.589986, precision = 0.921875 (65.082 sec) +INFO:tensorflow:global_step/sec: 1.53661 +INFO:tensorflow:step = 28401, loss = 0.713445, precision = 0.867188 (65.078 sec) +INFO:tensorflow:global_step/sec: 1.53661 +INFO:tensorflow:step = 28501, loss = 0.595454, precision = 0.898438 (65.078 sec) +Saved checkpoint after 73 epoch(s) to data/resnet164/checkpoints/00073... +INFO:tensorflow:global_step/sec: 1.48737 +INFO:tensorflow:step = 28601, loss = 0.573251, precision = 0.921875 (67.233 sec) +INFO:tensorflow:global_step/sec: 1.5378 +INFO:tensorflow:step = 28701, loss = 0.596389, precision = 0.890625 (65.028 sec) +INFO:tensorflow:global_step/sec: 1.53655 +INFO:tensorflow:step = 28801, loss = 0.667376, precision = 0.875 (65.081 sec) +INFO:tensorflow:global_step/sec: 1.53731 +INFO:tensorflow:step = 28901, loss = 0.55282, precision = 0.914062 (65.049 sec) +Saved checkpoint after 74 epoch(s) to data/resnet164/checkpoints/00074... +INFO:tensorflow:global_step/sec: 1.48713 +INFO:tensorflow:step = 29001, loss = 0.602912, precision = 0.914062 (67.244 sec) +INFO:tensorflow:global_step/sec: 1.53746 +INFO:tensorflow:step = 29101, loss = 0.55281, precision = 0.90625 (65.042 sec) +INFO:tensorflow:global_step/sec: 1.53825 +INFO:tensorflow:step = 29201, loss = 0.684066, precision = 0.890625 (65.009 sec) +INFO:tensorflow:global_step/sec: 1.53706 +INFO:tensorflow:step = 29301, loss = 0.583383, precision = 0.921875 (65.059 sec) +Saved checkpoint after 75 epoch(s) to data/resnet164/checkpoints/00075... +INFO:tensorflow:global_step/sec: 1.48807 +INFO:tensorflow:step = 29401, loss = 0.591098, precision = 0.898438 (67.201 sec) +INFO:tensorflow:global_step/sec: 1.53753 +INFO:tensorflow:step = 29501, loss = 0.606966, precision = 0.90625 (65.040 sec) +INFO:tensorflow:global_step/sec: 1.53733 +INFO:tensorflow:step = 29601, loss = 0.63906, precision = 0.867188 (65.048 sec) +INFO:tensorflow:global_step/sec: 1.53694 +INFO:tensorflow:step = 29701, loss = 0.706484, precision = 0.867188 (65.065 sec) +Saved checkpoint after 76 epoch(s) to data/resnet164/checkpoints/00076... +INFO:tensorflow:global_step/sec: 1.48766 +INFO:tensorflow:step = 29801, loss = 0.617723, precision = 0.90625 (67.220 sec) +INFO:tensorflow:global_step/sec: 1.53789 +INFO:tensorflow:step = 29901, loss = 0.540172, precision = 0.921875 (65.024 sec) +INFO:tensorflow:global_step/sec: 1.53784 +INFO:tensorflow:step = 30001, loss = 0.530819, precision = 0.929688 (65.026 sec) +INFO:tensorflow:global_step/sec: 1.53736 +INFO:tensorflow:step = 30101, loss = 0.537004, precision = 0.921875 (65.046 sec) +Saved checkpoint after 77 epoch(s) to data/resnet164/checkpoints/00077... +INFO:tensorflow:global_step/sec: 1.48741 +INFO:tensorflow:step = 30201, loss = 0.626214, precision = 0.898438 (67.231 sec) +INFO:tensorflow:global_step/sec: 1.5384 +INFO:tensorflow:step = 30301, loss = 0.69338, precision = 0.859375 (65.003 sec) +INFO:tensorflow:global_step/sec: 1.53232 +INFO:tensorflow:step = 30401, loss = 0.600037, precision = 0.90625 (65.260 sec) +Saved checkpoint after 78 epoch(s) to data/resnet164/checkpoints/00078... +INFO:tensorflow:global_step/sec: 1.48637 +INFO:tensorflow:step = 30501, loss = 0.586588, precision = 0.898438 (67.278 sec) +INFO:tensorflow:global_step/sec: 1.53904 +INFO:tensorflow:step = 30601, loss = 0.681159, precision = 0.875 (64.975 sec) +INFO:tensorflow:global_step/sec: 1.53799 +INFO:tensorflow:step = 30701, loss = 0.694549, precision = 0.875 (65.020 sec) +INFO:tensorflow:global_step/sec: 1.53757 +INFO:tensorflow:step = 30801, loss = 0.676913, precision = 0.851562 (65.037 sec) +Saved checkpoint after 79 epoch(s) to data/resnet164/checkpoints/00079... +INFO:tensorflow:global_step/sec: 1.48373 +INFO:tensorflow:step = 30901, loss = 0.615521, precision = 0.898438 (67.398 sec) +INFO:tensorflow:global_step/sec: 1.53852 +INFO:tensorflow:step = 31001, loss = 0.637895, precision = 0.882812 (64.998 sec) +INFO:tensorflow:global_step/sec: 1.53734 +INFO:tensorflow:step = 31101, loss = 0.588658, precision = 0.898438 (65.047 sec) +INFO:tensorflow:global_step/sec: 1.53788 +INFO:tensorflow:step = 31201, loss = 0.59791, precision = 0.898438 (65.025 sec) +Saved checkpoint after 80 epoch(s) to data/resnet164/checkpoints/00080... +INFO:tensorflow:global_step/sec: 1.48806 +INFO:tensorflow:step = 31301, loss = 0.579561, precision = 0.898438 (67.202 sec) +INFO:tensorflow:global_step/sec: 1.53841 +INFO:tensorflow:step = 31401, loss = 0.698907, precision = 0.882812 (65.002 sec) +INFO:tensorflow:global_step/sec: 1.53673 +INFO:tensorflow:step = 31501, loss = 0.619984, precision = 0.90625 (65.073 sec) +INFO:tensorflow:global_step/sec: 1.53801 +INFO:tensorflow:step = 31601, loss = 0.541995, precision = 0.921875 (65.019 sec) +Saved checkpoint after 81 epoch(s) to data/resnet164/checkpoints/00081... +INFO:tensorflow:global_step/sec: 1.48817 +INFO:tensorflow:step = 31701, loss = 0.714598, precision = 0.84375 (67.197 sec) +INFO:tensorflow:global_step/sec: 1.5385 +INFO:tensorflow:step = 31801, loss = 0.571965, precision = 0.921875 (64.998 sec) +INFO:tensorflow:global_step/sec: 1.53771 +INFO:tensorflow:step = 31901, loss = 0.674813, precision = 0.867188 (65.032 sec) +INFO:tensorflow:global_step/sec: 1.53795 +INFO:tensorflow:step = 32001, loss = 0.629702, precision = 0.875 (65.022 sec) +Saved checkpoint after 82 epoch(s) to data/resnet164/checkpoints/00082... +INFO:tensorflow:global_step/sec: 1.48759 +INFO:tensorflow:step = 32101, loss = 0.610066, precision = 0.914062 (67.223 sec) +INFO:tensorflow:global_step/sec: 1.53752 +INFO:tensorflow:step = 32201, loss = 0.604756, precision = 0.914062 (65.040 sec) +INFO:tensorflow:global_step/sec: 1.5373 +INFO:tensorflow:step = 32301, loss = 0.664176, precision = 0.890625 (65.049 sec) +INFO:tensorflow:global_step/sec: 1.53789 +INFO:tensorflow:step = 32401, loss = 0.523876, precision = 0.929688 (65.024 sec) +Saved checkpoint after 83 epoch(s) to data/resnet164/checkpoints/00083... +INFO:tensorflow:global_step/sec: 1.48853 +INFO:tensorflow:step = 32501, loss = 0.6705, precision = 0.890625 (67.181 sec) +INFO:tensorflow:global_step/sec: 1.53808 +INFO:tensorflow:step = 32601, loss = 0.575041, precision = 0.921875 (65.016 sec) +INFO:tensorflow:global_step/sec: 1.53729 +INFO:tensorflow:step = 32701, loss = 0.59817, precision = 0.898438 (65.050 sec) +INFO:tensorflow:global_step/sec: 1.53868 +INFO:tensorflow:step = 32801, loss = 0.484091, precision = 0.953125 (64.991 sec) +Saved checkpoint after 84 epoch(s) to data/resnet164/checkpoints/00084... +INFO:tensorflow:global_step/sec: 1.4886 +INFO:tensorflow:step = 32901, loss = 0.573995, precision = 0.90625 (67.177 sec) +INFO:tensorflow:global_step/sec: 1.53783 +INFO:tensorflow:step = 33001, loss = 0.578946, precision = 0.921875 (65.027 sec) +INFO:tensorflow:global_step/sec: 1.53813 +INFO:tensorflow:step = 33101, loss = 0.558529, precision = 0.914062 (65.014 sec) +INFO:tensorflow:global_step/sec: 1.53815 +INFO:tensorflow:step = 33201, loss = 0.715649, precision = 0.84375 (65.013 sec) +Saved checkpoint after 85 epoch(s) to data/resnet164/checkpoints/00085... +INFO:tensorflow:global_step/sec: 1.48816 +INFO:tensorflow:step = 33301, loss = 0.615297, precision = 0.882812 (67.197 sec) +INFO:tensorflow:global_step/sec: 1.53854 +INFO:tensorflow:step = 33401, loss = 0.9221, precision = 0.789062 (64.996 sec) +INFO:tensorflow:global_step/sec: 1.53805 +INFO:tensorflow:step = 33501, loss = 0.575363, precision = 0.898438 (65.018 sec) +INFO:tensorflow:global_step/sec: 1.53839 +INFO:tensorflow:step = 33601, loss = 0.609659, precision = 0.90625 (65.003 sec) +Saved checkpoint after 86 epoch(s) to data/resnet164/checkpoints/00086... +INFO:tensorflow:global_step/sec: 1.48794 +INFO:tensorflow:step = 33701, loss = 0.647603, precision = 0.914062 (67.207 sec) +INFO:tensorflow:global_step/sec: 1.53863 +INFO:tensorflow:step = 33801, loss = 0.604528, precision = 0.890625 (64.993 sec) +INFO:tensorflow:global_step/sec: 1.53752 +INFO:tensorflow:step = 33901, loss = 0.622788, precision = 0.898438 (65.040 sec) +INFO:tensorflow:global_step/sec: 1.53906 +INFO:tensorflow:step = 34001, loss = 0.71129, precision = 0.867188 (64.975 sec) +Saved checkpoint after 87 epoch(s) to data/resnet164/checkpoints/00087... +INFO:tensorflow:global_step/sec: 1.48785 +INFO:tensorflow:step = 34101, loss = 0.623885, precision = 0.90625 (67.211 sec) +INFO:tensorflow:global_step/sec: 1.53772 +INFO:tensorflow:step = 34201, loss = 0.641328, precision = 0.867188 (65.031 sec) +INFO:tensorflow:global_step/sec: 1.53811 +INFO:tensorflow:step = 34301, loss = 0.827174, precision = 0.835938 (65.015 sec) +INFO:tensorflow:global_step/sec: 1.53708 +INFO:tensorflow:step = 34401, loss = 0.690846, precision = 0.859375 (65.058 sec) +Saved checkpoint after 88 epoch(s) to data/resnet164/checkpoints/00088... +INFO:tensorflow:global_step/sec: 1.4877 +INFO:tensorflow:step = 34501, loss = 0.726781, precision = 0.859375 (67.218 sec) +INFO:tensorflow:global_step/sec: 1.53852 +INFO:tensorflow:step = 34601, loss = 0.490325, precision = 0.953125 (64.998 sec) +INFO:tensorflow:global_step/sec: 1.53815 +INFO:tensorflow:step = 34701, loss = 0.66919, precision = 0.867188 (65.013 sec) +Saved checkpoint after 89 epoch(s) to data/resnet164/checkpoints/00089... +INFO:tensorflow:global_step/sec: 1.48747 +INFO:tensorflow:step = 34801, loss = 0.601915, precision = 0.882812 (67.228 sec) +INFO:tensorflow:global_step/sec: 1.53793 +INFO:tensorflow:step = 34901, loss = 0.662251, precision = 0.875 (65.022 sec) +INFO:tensorflow:global_step/sec: 1.53778 +INFO:tensorflow:step = 35001, loss = 0.542129, precision = 0.929688 (65.029 sec) +INFO:tensorflow:global_step/sec: 1.53852 +INFO:tensorflow:step = 35101, loss = 0.585279, precision = 0.914062 (64.997 sec) +Saved checkpoint after 90 epoch(s) to data/resnet164/checkpoints/00090... +INFO:tensorflow:global_step/sec: 1.48291 +INFO:tensorflow:step = 35201, loss = 0.647449, precision = 0.875 (67.435 sec) +INFO:tensorflow:global_step/sec: 1.53912 +INFO:tensorflow:step = 35301, loss = 0.700691, precision = 0.875 (64.972 sec) +INFO:tensorflow:global_step/sec: 1.5384 +INFO:tensorflow:step = 35401, loss = 0.576982, precision = 0.929688 (65.002 sec) +INFO:tensorflow:global_step/sec: 1.53792 +INFO:tensorflow:step = 35501, loss = 0.590659, precision = 0.914062 (65.023 sec) +Saved checkpoint after 91 epoch(s) to data/resnet164/checkpoints/00091... +INFO:tensorflow:global_step/sec: 1.48734 +INFO:tensorflow:step = 35601, loss = 0.547712, precision = 0.929688 (67.234 sec) +INFO:tensorflow:global_step/sec: 1.5377 +INFO:tensorflow:step = 35701, loss = 0.527249, precision = 0.921875 (65.032 sec) +INFO:tensorflow:global_step/sec: 1.53727 +INFO:tensorflow:step = 35801, loss = 0.467066, precision = 0.960938 (65.050 sec) +INFO:tensorflow:global_step/sec: 1.53772 +INFO:tensorflow:step = 35901, loss = 0.526003, precision = 0.929688 (65.031 sec) +Saved checkpoint after 92 epoch(s) to data/resnet164/checkpoints/00092... +INFO:tensorflow:global_step/sec: 1.48769 +INFO:tensorflow:step = 36001, loss = 0.461445, precision = 0.953125 (67.218 sec) +INFO:tensorflow:global_step/sec: 1.53774 +INFO:tensorflow:step = 36101, loss = 0.422585, precision = 0.984375 (65.031 sec) +INFO:tensorflow:global_step/sec: 1.53794 +INFO:tensorflow:step = 36201, loss = 0.452637, precision = 0.953125 (65.022 sec) +INFO:tensorflow:global_step/sec: 1.53686 +INFO:tensorflow:step = 36301, loss = 0.398298, precision = 0.96875 (65.068 sec) +Saved checkpoint after 93 epoch(s) to data/resnet164/checkpoints/00093... +INFO:tensorflow:global_step/sec: 1.48726 +INFO:tensorflow:step = 36401, loss = 0.407796, precision = 0.976562 (67.238 sec) +INFO:tensorflow:global_step/sec: 1.53783 +INFO:tensorflow:step = 36501, loss = 0.381719, precision = 0.96875 (65.027 sec) +INFO:tensorflow:global_step/sec: 1.53745 +INFO:tensorflow:step = 36601, loss = 0.4043, precision = 0.953125 (65.043 sec) +INFO:tensorflow:global_step/sec: 1.53741 +INFO:tensorflow:step = 36701, loss = 0.325274, precision = 0.992188 (65.045 sec) +Saved checkpoint after 94 epoch(s) to data/resnet164/checkpoints/00094... +INFO:tensorflow:global_step/sec: 1.48747 +INFO:tensorflow:step = 36801, loss = 0.353671, precision = 0.992188 (67.228 sec) +INFO:tensorflow:global_step/sec: 1.53851 +INFO:tensorflow:step = 36901, loss = 0.365548, precision = 0.976562 (64.998 sec) +INFO:tensorflow:global_step/sec: 1.53739 +INFO:tensorflow:step = 37001, loss = 0.465176, precision = 0.929688 (65.045 sec) +INFO:tensorflow:global_step/sec: 1.53685 +INFO:tensorflow:step = 37101, loss = 0.40852, precision = 0.953125 (65.068 sec) +Saved checkpoint after 95 epoch(s) to data/resnet164/checkpoints/00095... +INFO:tensorflow:global_step/sec: 1.48796 +INFO:tensorflow:step = 37201, loss = 0.380133, precision = 0.960938 (67.206 sec) +INFO:tensorflow:global_step/sec: 1.53768 +INFO:tensorflow:step = 37301, loss = 0.312698, precision = 1.0 (65.033 sec) +INFO:tensorflow:global_step/sec: 1.53807 +INFO:tensorflow:step = 37401, loss = 0.380572, precision = 0.960938 (65.016 sec) +INFO:tensorflow:global_step/sec: 1.53708 +INFO:tensorflow:step = 37501, loss = 0.364437, precision = 0.96875 (65.058 sec) +Saved checkpoint after 96 epoch(s) to data/resnet164/checkpoints/00096... +INFO:tensorflow:global_step/sec: 1.48724 +INFO:tensorflow:step = 37601, loss = 0.35312, precision = 0.976562 (67.239 sec) +INFO:tensorflow:global_step/sec: 1.53748 +INFO:tensorflow:step = 37701, loss = 0.326605, precision = 0.984375 (65.042 sec) +INFO:tensorflow:global_step/sec: 1.53733 +INFO:tensorflow:step = 37801, loss = 0.367262, precision = 0.96875 (65.048 sec) +INFO:tensorflow:global_step/sec: 1.53777 +INFO:tensorflow:step = 37901, loss = 0.328368, precision = 0.984375 (65.029 sec) +Saved checkpoint after 97 epoch(s) to data/resnet164/checkpoints/00097... +INFO:tensorflow:global_step/sec: 1.48775 +INFO:tensorflow:step = 38001, loss = 0.292248, precision = 1.0 (67.216 sec) +INFO:tensorflow:global_step/sec: 1.53784 +INFO:tensorflow:step = 38101, loss = 0.318555, precision = 0.984375 (65.026 sec) +INFO:tensorflow:global_step/sec: 1.53745 +INFO:tensorflow:step = 38201, loss = 0.285471, precision = 1.0 (65.043 sec) +INFO:tensorflow:global_step/sec: 1.53803 +INFO:tensorflow:step = 38301, loss = 0.335056, precision = 0.976562 (65.018 sec) +Saved checkpoint after 98 epoch(s) to data/resnet164/checkpoints/00098... +INFO:tensorflow:global_step/sec: 1.48727 +INFO:tensorflow:step = 38401, loss = 0.356182, precision = 0.945312 (67.238 sec) +INFO:tensorflow:global_step/sec: 1.53647 +INFO:tensorflow:step = 38501, loss = 0.32673, precision = 0.992188 (65.084 sec) +INFO:tensorflow:global_step/sec: 1.53682 +INFO:tensorflow:step = 38601, loss = 0.333199, precision = 0.976562 (65.069 sec) +INFO:tensorflow:global_step/sec: 1.53792 +INFO:tensorflow:step = 38701, loss = 0.401517, precision = 0.945312 (65.023 sec) +Saved checkpoint after 99 epoch(s) to data/resnet164/checkpoints/00099... +INFO:tensorflow:global_step/sec: 1.48787 +INFO:tensorflow:step = 38801, loss = 0.310998, precision = 0.976562 (67.210 sec) +INFO:tensorflow:global_step/sec: 1.5371 +INFO:tensorflow:step = 38901, loss = 0.31737, precision = 0.96875 (65.057 sec) +INFO:tensorflow:global_step/sec: 1.5374 +INFO:tensorflow:step = 39001, loss = 0.32711, precision = 0.984375 (65.045 sec) +Saved checkpoint after 100 epoch(s) to data/resnet164/checkpoints/00100... +INFO:tensorflow:global_step/sec: 1.48178 +INFO:tensorflow:step = 39101, loss = 0.326505, precision = 0.96875 (67.486 sec) +INFO:tensorflow:global_step/sec: 1.53752 +INFO:tensorflow:step = 39201, loss = 0.298131, precision = 0.976562 (65.040 sec) +INFO:tensorflow:global_step/sec: 1.53665 +INFO:tensorflow:step = 39301, loss = 0.275153, precision = 0.976562 (65.077 sec) +INFO:tensorflow:global_step/sec: 1.5373 +INFO:tensorflow:step = 39401, loss = 0.298815, precision = 0.976562 (65.049 sec) +Saved checkpoint after 101 epoch(s) to data/resnet164/checkpoints/00101... +INFO:tensorflow:global_step/sec: 1.48797 +INFO:tensorflow:step = 39501, loss = 0.304136, precision = 0.984375 (67.206 sec) +INFO:tensorflow:global_step/sec: 1.53816 +INFO:tensorflow:step = 39601, loss = 0.281316, precision = 0.976562 (65.012 sec) +INFO:tensorflow:global_step/sec: 1.53706 +INFO:tensorflow:step = 39701, loss = 0.275897, precision = 0.992188 (65.059 sec) +INFO:tensorflow:global_step/sec: 1.53833 +INFO:tensorflow:step = 39801, loss = 0.33823, precision = 0.96875 (65.006 sec) +Saved checkpoint after 102 epoch(s) to data/resnet164/checkpoints/00102... +INFO:tensorflow:global_step/sec: 1.48777 +INFO:tensorflow:step = 39901, loss = 0.263481, precision = 0.992188 (67.215 sec) +INFO:tensorflow:global_step/sec: 1.53835 +INFO:tensorflow:step = 40001, loss = 0.289812, precision = 0.96875 (65.004 sec) +INFO:tensorflow:global_step/sec: 1.53742 +INFO:tensorflow:step = 40101, loss = 0.274278, precision = 0.984375 (65.044 sec) +INFO:tensorflow:global_step/sec: 1.53749 +INFO:tensorflow:step = 40201, loss = 0.310522, precision = 0.953125 (65.041 sec) +Saved checkpoint after 103 epoch(s) to data/resnet164/checkpoints/00103... +INFO:tensorflow:global_step/sec: 1.48781 +INFO:tensorflow:step = 40301, loss = 0.3062, precision = 0.96875 (67.213 sec) +INFO:tensorflow:global_step/sec: 1.53736 +INFO:tensorflow:step = 40401, loss = 0.290786, precision = 0.953125 (65.047 sec) +INFO:tensorflow:global_step/sec: 1.53765 +INFO:tensorflow:step = 40501, loss = 0.280181, precision = 0.992188 (65.034 sec) +INFO:tensorflow:global_step/sec: 1.53752 +INFO:tensorflow:step = 40601, loss = 0.264947, precision = 0.984375 (65.040 sec) +Saved checkpoint after 104 epoch(s) to data/resnet164/checkpoints/00104... +INFO:tensorflow:global_step/sec: 1.48759 +INFO:tensorflow:step = 40701, loss = 0.264948, precision = 0.976562 (67.223 sec) +INFO:tensorflow:global_step/sec: 1.53756 +INFO:tensorflow:step = 40801, loss = 0.32631, precision = 0.953125 (65.038 sec) +INFO:tensorflow:global_step/sec: 1.53737 +INFO:tensorflow:step = 40901, loss = 0.241206, precision = 1.0 (65.046 sec) +INFO:tensorflow:global_step/sec: 1.53685 +INFO:tensorflow:step = 41001, loss = 0.243662, precision = 0.992188 (65.068 sec) +Saved checkpoint after 105 epoch(s) to data/resnet164/checkpoints/00105... +INFO:tensorflow:global_step/sec: 1.48732 +INFO:tensorflow:step = 41101, loss = 0.29101, precision = 0.976562 (67.235 sec) +INFO:tensorflow:global_step/sec: 1.53773 +INFO:tensorflow:step = 41201, loss = 0.250219, precision = 0.984375 (65.031 sec) +INFO:tensorflow:global_step/sec: 1.53631 +INFO:tensorflow:step = 41301, loss = 0.236117, precision = 0.984375 (65.091 sec) +INFO:tensorflow:global_step/sec: 1.53787 +INFO:tensorflow:step = 41401, loss = 0.291215, precision = 0.976562 (65.025 sec) +Saved checkpoint after 106 epoch(s) to data/resnet164/checkpoints/00106... +INFO:tensorflow:global_step/sec: 1.48725 +INFO:tensorflow:step = 41501, loss = 0.241246, precision = 0.984375 (67.238 sec) +INFO:tensorflow:global_step/sec: 1.53714 +INFO:tensorflow:step = 41601, loss = 0.287592, precision = 0.953125 (65.056 sec) +INFO:tensorflow:global_step/sec: 1.53726 +INFO:tensorflow:step = 41701, loss = 0.263299, precision = 0.976562 (65.051 sec) +INFO:tensorflow:global_step/sec: 1.53785 +INFO:tensorflow:step = 41801, loss = 0.267142, precision = 0.984375 (65.026 sec) +Saved checkpoint after 107 epoch(s) to data/resnet164/checkpoints/00107... +INFO:tensorflow:global_step/sec: 1.48746 +INFO:tensorflow:step = 41901, loss = 0.25487, precision = 0.976562 (67.229 sec) +INFO:tensorflow:global_step/sec: 1.53766 +INFO:tensorflow:step = 42001, loss = 0.217929, precision = 0.992188 (65.034 sec) +INFO:tensorflow:global_step/sec: 1.53719 +INFO:tensorflow:step = 42101, loss = 0.243461, precision = 0.976562 (65.054 sec) +INFO:tensorflow:global_step/sec: 1.53665 +INFO:tensorflow:step = 42201, loss = 0.221499, precision = 0.992188 (65.077 sec) +Saved checkpoint after 108 epoch(s) to data/resnet164/checkpoints/00108... +INFO:tensorflow:global_step/sec: 1.48717 +INFO:tensorflow:step = 42301, loss = 0.249487, precision = 0.976562 (67.242 sec) +INFO:tensorflow:global_step/sec: 1.53683 +INFO:tensorflow:step = 42401, loss = 0.207245, precision = 1.0 (65.069 sec) +INFO:tensorflow:global_step/sec: 1.53751 +INFO:tensorflow:step = 42501, loss = 0.27597, precision = 0.976562 (65.040 sec) +INFO:tensorflow:global_step/sec: 1.53737 +INFO:tensorflow:step = 42601, loss = 0.249345, precision = 0.976562 (65.046 sec) +Saved checkpoint after 109 epoch(s) to data/resnet164/checkpoints/00109... +INFO:tensorflow:global_step/sec: 1.48721 +INFO:tensorflow:step = 42701, loss = 0.22232, precision = 0.984375 (67.240 sec) +INFO:tensorflow:global_step/sec: 1.5369 +INFO:tensorflow:step = 42801, loss = 0.205983, precision = 0.992188 (65.066 sec) +INFO:tensorflow:global_step/sec: 1.53706 +INFO:tensorflow:step = 42901, loss = 0.209755, precision = 1.0 (65.059 sec) +INFO:tensorflow:global_step/sec: 1.53694 +INFO:tensorflow:step = 43001, loss = 0.239174, precision = 0.976562 (65.064 sec) +Saved checkpoint after 110 epoch(s) to data/resnet164/checkpoints/00110... +INFO:tensorflow:global_step/sec: 1.48704 +INFO:tensorflow:step = 43101, loss = 0.265903, precision = 0.96875 (67.248 sec) +INFO:tensorflow:global_step/sec: 1.5372 +INFO:tensorflow:step = 43201, loss = 0.237779, precision = 0.96875 (65.053 sec) +INFO:tensorflow:global_step/sec: 1.53705 +INFO:tensorflow:step = 43301, loss = 0.215969, precision = 0.992188 (65.060 sec) +Saved checkpoint after 111 epoch(s) to data/resnet164/checkpoints/00111... +INFO:tensorflow:global_step/sec: 1.48141 +INFO:tensorflow:step = 43401, loss = 0.207816, precision = 0.992188 (67.503 sec) +INFO:tensorflow:global_step/sec: 1.53782 +INFO:tensorflow:step = 43501, loss = 0.275785, precision = 0.960938 (65.027 sec) +INFO:tensorflow:global_step/sec: 1.53671 +INFO:tensorflow:step = 43601, loss = 0.246326, precision = 0.976562 (65.074 sec) +INFO:tensorflow:global_step/sec: 1.53605 +INFO:tensorflow:step = 43701, loss = 0.336145, precision = 0.945312 (65.102 sec) +Saved checkpoint after 112 epoch(s) to data/resnet164/checkpoints/00112... +INFO:tensorflow:global_step/sec: 1.48674 +INFO:tensorflow:step = 43801, loss = 0.224973, precision = 0.976562 (67.261 sec) +INFO:tensorflow:global_step/sec: 1.53837 +INFO:tensorflow:step = 43901, loss = 0.203918, precision = 0.992188 (65.004 sec) +INFO:tensorflow:global_step/sec: 1.53736 +INFO:tensorflow:step = 44001, loss = 0.225746, precision = 0.96875 (65.047 sec) +INFO:tensorflow:global_step/sec: 1.53794 +INFO:tensorflow:step = 44101, loss = 0.209245, precision = 0.992188 (65.022 sec) +Saved checkpoint after 113 epoch(s) to data/resnet164/checkpoints/00113... +INFO:tensorflow:global_step/sec: 1.48808 +INFO:tensorflow:step = 44201, loss = 0.193683, precision = 0.992188 (67.201 sec) +INFO:tensorflow:global_step/sec: 1.53898 +INFO:tensorflow:step = 44301, loss = 0.217642, precision = 0.976562 (64.978 sec) +INFO:tensorflow:global_step/sec: 1.53853 +INFO:tensorflow:step = 44401, loss = 0.263248, precision = 0.984375 (64.997 sec) +INFO:tensorflow:global_step/sec: 1.53896 +INFO:tensorflow:step = 44501, loss = 0.241328, precision = 0.984375 (64.979 sec) +Saved checkpoint after 114 epoch(s) to data/resnet164/checkpoints/00114... +INFO:tensorflow:global_step/sec: 1.48987 +INFO:tensorflow:step = 44601, loss = 0.192058, precision = 0.992188 (67.120 sec) +INFO:tensorflow:global_step/sec: 1.53953 +INFO:tensorflow:step = 44701, loss = 0.227114, precision = 0.976562 (64.955 sec) +INFO:tensorflow:global_step/sec: 1.53991 +INFO:tensorflow:step = 44801, loss = 0.209883, precision = 0.984375 (64.939 sec) +INFO:tensorflow:global_step/sec: 1.53989 +INFO:tensorflow:step = 44901, loss = 0.224748, precision = 0.976562 (64.940 sec) +Saved checkpoint after 115 epoch(s) to data/resnet164/checkpoints/00115... +INFO:tensorflow:global_step/sec: 1.49048 +INFO:tensorflow:step = 45001, loss = 0.211792, precision = 0.984375 (67.093 sec) +INFO:tensorflow:global_step/sec: 1.5404 +INFO:tensorflow:step = 45101, loss = 0.226179, precision = 0.984375 (64.918 sec) +INFO:tensorflow:global_step/sec: 1.54074 +INFO:tensorflow:step = 45201, loss = 0.188213, precision = 0.992188 (64.904 sec) +INFO:tensorflow:global_step/sec: 1.54062 +INFO:tensorflow:step = 45301, loss = 0.193554, precision = 0.992188 (64.909 sec) +Saved checkpoint after 116 epoch(s) to data/resnet164/checkpoints/00116... +INFO:tensorflow:global_step/sec: 1.49062 +INFO:tensorflow:step = 45401, loss = 0.221339, precision = 0.984375 (67.086 sec) +INFO:tensorflow:global_step/sec: 1.54042 +INFO:tensorflow:step = 45501, loss = 0.196198, precision = 0.984375 (64.917 sec) +INFO:tensorflow:global_step/sec: 1.54071 +INFO:tensorflow:step = 45601, loss = 0.204959, precision = 0.984375 (64.905 sec) +INFO:tensorflow:global_step/sec: 1.54091 +INFO:tensorflow:step = 45701, loss = 0.225076, precision = 0.976562 (64.897 sec) +Saved checkpoint after 117 epoch(s) to data/resnet164/checkpoints/00117... +INFO:tensorflow:global_step/sec: 1.48969 +INFO:tensorflow:step = 45801, loss = 0.225426, precision = 0.984375 (67.128 sec) +INFO:tensorflow:global_step/sec: 1.54117 +INFO:tensorflow:step = 45901, loss = 0.246229, precision = 0.976562 (64.886 sec) +INFO:tensorflow:global_step/sec: 1.5416 +INFO:tensorflow:step = 46001, loss = 0.195941, precision = 0.992188 (64.868 sec) +INFO:tensorflow:global_step/sec: 1.54105 +INFO:tensorflow:step = 46101, loss = 0.207296, precision = 0.976562 (64.891 sec) +Saved checkpoint after 118 epoch(s) to data/resnet164/checkpoints/00118... +INFO:tensorflow:global_step/sec: 1.48859 +INFO:tensorflow:step = 46201, loss = 0.178315, precision = 0.992188 (67.178 sec) +INFO:tensorflow:global_step/sec: 1.54164 +INFO:tensorflow:step = 46301, loss = 0.230019, precision = 0.984375 (64.866 sec) +INFO:tensorflow:global_step/sec: 1.54091 +INFO:tensorflow:step = 46401, loss = 0.171436, precision = 1.0 (64.897 sec) +INFO:tensorflow:global_step/sec: 1.54047 +INFO:tensorflow:step = 46501, loss = 0.171255, precision = 0.992188 (64.915 sec) +Saved checkpoint after 119 epoch(s) to data/resnet164/checkpoints/00119... +INFO:tensorflow:global_step/sec: 1.49119 +INFO:tensorflow:step = 46601, loss = 0.188434, precision = 0.992188 (67.061 sec) +INFO:tensorflow:global_step/sec: 1.54124 +INFO:tensorflow:step = 46701, loss = 0.232863, precision = 0.96875 (64.883 sec) +INFO:tensorflow:global_step/sec: 1.54104 +INFO:tensorflow:step = 46801, loss = 0.210805, precision = 0.984375 (64.891 sec) +INFO:tensorflow:global_step/sec: 1.54085 +INFO:tensorflow:step = 46901, loss = 0.206052, precision = 0.984375 (64.899 sec) +Saved checkpoint after 120 epoch(s) to data/resnet164/checkpoints/00120... +INFO:tensorflow:global_step/sec: 1.49023 +INFO:tensorflow:step = 47001, loss = 0.195099, precision = 0.992188 (67.104 sec) +INFO:tensorflow:global_step/sec: 1.54114 +INFO:tensorflow:step = 47101, loss = 0.233702, precision = 0.976562 (64.887 sec) +INFO:tensorflow:global_step/sec: 1.54122 +INFO:tensorflow:step = 47201, loss = 0.191254, precision = 0.992188 (64.884 sec) +INFO:tensorflow:global_step/sec: 1.5404 +INFO:tensorflow:step = 47301, loss = 0.202566, precision = 0.976562 (64.918 sec) +Saved checkpoint after 121 epoch(s) to data/resnet164/checkpoints/00121... +INFO:tensorflow:global_step/sec: 1.48556 +INFO:tensorflow:step = 47401, loss = 0.188768, precision = 0.984375 (67.315 sec) +INFO:tensorflow:global_step/sec: 1.54053 +INFO:tensorflow:step = 47501, loss = 0.182773, precision = 0.984375 (64.913 sec) +INFO:tensorflow:global_step/sec: 1.53916 +INFO:tensorflow:step = 47601, loss = 0.191302, precision = 0.984375 (64.971 sec) +INFO:tensorflow:global_step/sec: 1.5396 +INFO:tensorflow:step = 47701, loss = 0.198773, precision = 0.992188 (64.952 sec) +Saved checkpoint after 122 epoch(s) to data/resnet164/checkpoints/00122... +INFO:tensorflow:global_step/sec: 1.48791 +INFO:tensorflow:step = 47801, loss = 0.24725, precision = 0.976562 (67.209 sec) +INFO:tensorflow:global_step/sec: 1.53909 +INFO:tensorflow:step = 47901, loss = 0.225528, precision = 0.984375 (64.973 sec) +INFO:tensorflow:global_step/sec: 1.53751 +INFO:tensorflow:step = 48001, loss = 0.211138, precision = 0.976562 (65.040 sec) +Saved checkpoint after 123 epoch(s) to data/resnet164/checkpoints/00123... +INFO:tensorflow:global_step/sec: 1.48821 +INFO:tensorflow:step = 48101, loss = 0.212074, precision = 0.96875 (67.195 sec) +INFO:tensorflow:global_step/sec: 1.53927 +INFO:tensorflow:step = 48201, loss = 0.255868, precision = 0.960938 (64.966 sec) +INFO:tensorflow:global_step/sec: 1.53853 +INFO:tensorflow:step = 48301, loss = 0.20864, precision = 0.984375 (64.997 sec) +INFO:tensorflow:global_step/sec: 1.53862 +INFO:tensorflow:step = 48401, loss = 0.168352, precision = 1.0 (64.993 sec) +Saved checkpoint after 124 epoch(s) to data/resnet164/checkpoints/00124... +INFO:tensorflow:global_step/sec: 1.48848 +INFO:tensorflow:step = 48501, loss = 0.206871, precision = 0.984375 (67.183 sec) +INFO:tensorflow:global_step/sec: 1.53889 +INFO:tensorflow:step = 48601, loss = 0.169215, precision = 0.992188 (64.982 sec) +INFO:tensorflow:global_step/sec: 1.53845 +INFO:tensorflow:step = 48701, loss = 0.19047, precision = 0.976562 (65.001 sec) +INFO:tensorflow:global_step/sec: 1.53786 +INFO:tensorflow:step = 48801, loss = 0.176695, precision = 0.992188 (65.025 sec) +Saved checkpoint after 125 epoch(s) to data/resnet164/checkpoints/00125... +INFO:tensorflow:global_step/sec: 1.48697 +INFO:tensorflow:step = 48901, loss = 0.196252, precision = 0.984375 (67.251 sec) +INFO:tensorflow:global_step/sec: 1.53752 +INFO:tensorflow:step = 49001, loss = 0.215234, precision = 0.984375 (65.040 sec) +INFO:tensorflow:global_step/sec: 1.53788 +INFO:tensorflow:step = 49101, loss = 0.254512, precision = 0.960938 (65.025 sec) +INFO:tensorflow:global_step/sec: 1.5386 +INFO:tensorflow:step = 49201, loss = 0.219473, precision = 0.953125 (64.994 sec) +Saved checkpoint after 126 epoch(s) to data/resnet164/checkpoints/00126... +INFO:tensorflow:global_step/sec: 1.48788 +INFO:tensorflow:step = 49301, loss = 0.31456, precision = 0.953125 (67.210 sec) +INFO:tensorflow:global_step/sec: 1.53828 +INFO:tensorflow:step = 49401, loss = 0.201147, precision = 0.976562 (65.008 sec) +INFO:tensorflow:global_step/sec: 1.53784 +INFO:tensorflow:step = 49501, loss = 0.204158, precision = 0.976562 (65.026 sec) +INFO:tensorflow:global_step/sec: 1.53834 +INFO:tensorflow:step = 49601, loss = 0.171875, precision = 0.992188 (65.005 sec) +Saved checkpoint after 127 epoch(s) to data/resnet164/checkpoints/00127... +INFO:tensorflow:global_step/sec: 1.48765 +INFO:tensorflow:step = 49701, loss = 0.168124, precision = 1.0 (67.220 sec) +INFO:tensorflow:global_step/sec: 1.53825 +INFO:tensorflow:step = 49801, loss = 0.186093, precision = 0.992188 (65.009 sec) +INFO:tensorflow:global_step/sec: 1.53775 +INFO:tensorflow:step = 49901, loss = 0.170041, precision = 1.0 (65.030 sec) +INFO:tensorflow:global_step/sec: 1.53817 +INFO:tensorflow:step = 50001, loss = 0.211922, precision = 0.976562 (65.012 sec) +Saved checkpoint after 128 epoch(s) to data/resnet164/checkpoints/00128... +INFO:tensorflow:global_step/sec: 1.48806 +INFO:tensorflow:step = 50101, loss = 0.208299, precision = 0.976562 (67.202 sec) +INFO:tensorflow:global_step/sec: 1.5385 +INFO:tensorflow:step = 50201, loss = 0.177205, precision = 0.992188 (64.998 sec) +INFO:tensorflow:global_step/sec: 1.53832 +INFO:tensorflow:step = 50301, loss = 0.178166, precision = 0.984375 (65.006 sec) +INFO:tensorflow:global_step/sec: 1.53801 +INFO:tensorflow:step = 50401, loss = 0.168742, precision = 0.992188 (65.019 sec) +Saved checkpoint after 129 epoch(s) to data/resnet164/checkpoints/00129... +INFO:tensorflow:global_step/sec: 1.48606 +INFO:tensorflow:step = 50501, loss = 0.166231, precision = 0.992188 (67.292 sec) +INFO:tensorflow:global_step/sec: 1.53775 +INFO:tensorflow:step = 50601, loss = 0.157305, precision = 1.0 (65.030 sec) +INFO:tensorflow:global_step/sec: 1.53866 +INFO:tensorflow:step = 50701, loss = 0.193046, precision = 0.976562 (64.992 sec) +INFO:tensorflow:global_step/sec: 1.53812 +INFO:tensorflow:step = 50801, loss = 0.171203, precision = 0.992188 (65.014 sec) +Saved checkpoint after 130 epoch(s) to data/resnet164/checkpoints/00130... +INFO:tensorflow:global_step/sec: 1.48748 +INFO:tensorflow:step = 50901, loss = 0.186337, precision = 0.976562 (67.228 sec) +INFO:tensorflow:global_step/sec: 1.53774 +INFO:tensorflow:step = 51001, loss = 0.191591, precision = 0.992188 (65.030 sec) +INFO:tensorflow:global_step/sec: 1.53719 +INFO:tensorflow:step = 51101, loss = 0.206299, precision = 0.976562 (65.054 sec) +INFO:tensorflow:global_step/sec: 1.53864 +INFO:tensorflow:step = 51201, loss = 0.176519, precision = 1.0 (64.992 sec) +Saved checkpoint after 131 epoch(s) to data/resnet164/checkpoints/00131... +INFO:tensorflow:global_step/sec: 1.48303 +INFO:tensorflow:step = 51301, loss = 0.189343, precision = 0.984375 (67.430 sec) +INFO:tensorflow:global_step/sec: 1.5383 +INFO:tensorflow:step = 51401, loss = 0.161906, precision = 0.992188 (65.007 sec) +INFO:tensorflow:global_step/sec: 1.53881 +INFO:tensorflow:step = 51501, loss = 0.159862, precision = 0.992188 (64.985 sec) +INFO:tensorflow:global_step/sec: 1.53823 +INFO:tensorflow:step = 51601, loss = 0.176718, precision = 0.992188 (65.010 sec) +Saved checkpoint after 132 epoch(s) to data/resnet164/checkpoints/00132... +INFO:tensorflow:global_step/sec: 1.48826 +INFO:tensorflow:step = 51701, loss = 0.190228, precision = 0.984375 (67.193 sec) +INFO:tensorflow:global_step/sec: 1.53908 +INFO:tensorflow:step = 51801, loss = 0.262819, precision = 0.96875 (64.974 sec) +INFO:tensorflow:global_step/sec: 1.53821 +INFO:tensorflow:step = 51901, loss = 0.167837, precision = 0.992188 (65.011 sec) +INFO:tensorflow:global_step/sec: 1.53747 +INFO:tensorflow:step = 52001, loss = 0.170519, precision = 0.984375 (65.042 sec) +Saved checkpoint after 133 epoch(s) to data/resnet164/checkpoints/00133... +INFO:tensorflow:global_step/sec: 1.48818 +INFO:tensorflow:step = 52101, loss = 0.194466, precision = 0.976562 (67.196 sec) +INFO:tensorflow:global_step/sec: 1.5388 +INFO:tensorflow:step = 52201, loss = 0.173276, precision = 0.992188 (64.986 sec) +INFO:tensorflow:global_step/sec: 1.53859 +INFO:tensorflow:step = 52301, loss = 0.171532, precision = 0.992188 (64.995 sec) +Saved checkpoint after 134 epoch(s) to data/resnet164/checkpoints/00134... +INFO:tensorflow:global_step/sec: 1.48762 +INFO:tensorflow:step = 52401, loss = 0.192472, precision = 0.976562 (67.221 sec) +INFO:tensorflow:global_step/sec: 1.53766 +INFO:tensorflow:step = 52501, loss = 0.168744, precision = 0.992188 (65.034 sec) +INFO:tensorflow:global_step/sec: 1.53747 +INFO:tensorflow:step = 52601, loss = 0.171771, precision = 0.984375 (65.042 sec) +INFO:tensorflow:global_step/sec: 1.53854 +INFO:tensorflow:step = 52701, loss = 0.223454, precision = 0.96875 (64.996 sec) +Saved checkpoint after 135 epoch(s) to data/resnet164/checkpoints/00135... +INFO:tensorflow:global_step/sec: 1.48755 +INFO:tensorflow:step = 52801, loss = 0.161228, precision = 1.0 (67.225 sec) +INFO:tensorflow:global_step/sec: 1.53798 +INFO:tensorflow:step = 52901, loss = 0.184294, precision = 0.984375 (65.020 sec) +INFO:tensorflow:global_step/sec: 1.53794 +INFO:tensorflow:step = 53001, loss = 0.187066, precision = 0.976562 (65.022 sec) +INFO:tensorflow:global_step/sec: 1.53798 +INFO:tensorflow:step = 53101, loss = 0.191765, precision = 0.984375 (65.020 sec) +Saved checkpoint after 136 epoch(s) to data/resnet164/checkpoints/00136... +INFO:tensorflow:global_step/sec: 1.48814 +INFO:tensorflow:step = 53201, loss = 0.1646, precision = 0.992188 (67.198 sec) +INFO:tensorflow:global_step/sec: 1.53823 +INFO:tensorflow:step = 53301, loss = 0.177051, precision = 0.992188 (65.010 sec) +INFO:tensorflow:global_step/sec: 1.5383 +INFO:tensorflow:step = 53401, loss = 0.198561, precision = 0.976562 (65.007 sec) +INFO:tensorflow:global_step/sec: 1.53881 +INFO:tensorflow:step = 53501, loss = 0.152284, precision = 1.0 (64.985 sec) +Saved checkpoint after 137 epoch(s) to data/resnet164/checkpoints/00137... +INFO:tensorflow:global_step/sec: 1.48801 +INFO:tensorflow:step = 53601, loss = 0.150971, precision = 1.0 (67.204 sec) +INFO:tensorflow:global_step/sec: 1.53876 +INFO:tensorflow:step = 53701, loss = 0.145572, precision = 1.0 (64.987 sec) +INFO:tensorflow:global_step/sec: 1.53793 +INFO:tensorflow:step = 53801, loss = 0.157756, precision = 0.992188 (65.022 sec) +INFO:tensorflow:global_step/sec: 1.53856 +INFO:tensorflow:step = 53901, loss = 0.146404, precision = 1.0 (64.996 sec) +Saved checkpoint after 138 epoch(s) to data/resnet164/checkpoints/00138... +INFO:tensorflow:global_step/sec: 1.48836 +INFO:tensorflow:step = 54001, loss = 0.142319, precision = 1.0 (67.188 sec) +INFO:tensorflow:global_step/sec: 1.53906 +INFO:tensorflow:step = 54101, loss = 0.138653, precision = 1.0 (64.975 sec) +INFO:tensorflow:global_step/sec: 1.53836 +INFO:tensorflow:step = 54201, loss = 0.140467, precision = 1.0 (65.004 sec) +INFO:tensorflow:global_step/sec: 1.53835 +INFO:tensorflow:step = 54301, loss = 0.152594, precision = 0.992188 (65.005 sec) +Saved checkpoint after 139 epoch(s) to data/resnet164/checkpoints/00139... +INFO:tensorflow:global_step/sec: 1.48828 +INFO:tensorflow:step = 54401, loss = 0.150055, precision = 1.0 (67.192 sec) +INFO:tensorflow:global_step/sec: 1.53778 +INFO:tensorflow:step = 54501, loss = 0.14733, precision = 0.992188 (65.029 sec) +INFO:tensorflow:global_step/sec: 1.53814 +INFO:tensorflow:step = 54601, loss = 0.148478, precision = 1.0 (65.014 sec) +INFO:tensorflow:global_step/sec: 1.53911 +INFO:tensorflow:step = 54701, loss = 0.139671, precision = 1.0 (64.973 sec) +Saved checkpoint after 140 epoch(s) to data/resnet164/checkpoints/00140... +INFO:tensorflow:global_step/sec: 1.4879 +INFO:tensorflow:step = 54801, loss = 0.137278, precision = 1.0 (67.209 sec) +INFO:tensorflow:global_step/sec: 1.53821 +INFO:tensorflow:step = 54901, loss = 0.141581, precision = 1.0 (65.011 sec) +INFO:tensorflow:global_step/sec: 1.53835 +INFO:tensorflow:step = 55001, loss = 0.145117, precision = 1.0 (65.005 sec) +INFO:tensorflow:global_step/sec: 1.53826 +INFO:tensorflow:step = 55101, loss = 0.139431, precision = 1.0 (65.009 sec) +Saved checkpoint after 141 epoch(s) to data/resnet164/checkpoints/00141... +INFO:tensorflow:global_step/sec: 1.48305 +INFO:tensorflow:step = 55201, loss = 0.147848, precision = 1.0 (67.429 sec) +INFO:tensorflow:global_step/sec: 1.53811 +INFO:tensorflow:step = 55301, loss = 0.137561, precision = 1.0 (65.015 sec) +INFO:tensorflow:global_step/sec: 1.53806 +INFO:tensorflow:step = 55401, loss = 0.139827, precision = 1.0 (65.017 sec) +INFO:tensorflow:global_step/sec: 1.53787 +INFO:tensorflow:step = 55501, loss = 0.13618, precision = 1.0 (65.025 sec) +Saved checkpoint after 142 epoch(s) to data/resnet164/checkpoints/00142... +INFO:tensorflow:global_step/sec: 1.48827 +INFO:tensorflow:step = 55601, loss = 0.139227, precision = 1.0 (67.192 sec) +INFO:tensorflow:global_step/sec: 1.53851 +INFO:tensorflow:step = 55701, loss = 0.142242, precision = 1.0 (64.998 sec) +INFO:tensorflow:global_step/sec: 1.53891 +INFO:tensorflow:step = 55801, loss = 0.137513, precision = 1.0 (64.981 sec) +INFO:tensorflow:global_step/sec: 1.53809 +INFO:tensorflow:step = 55901, loss = 0.158402, precision = 0.992188 (65.016 sec) +Saved checkpoint after 143 epoch(s) to data/resnet164/checkpoints/00143... +INFO:tensorflow:global_step/sec: 1.48926 +INFO:tensorflow:step = 56001, loss = 0.137462, precision = 1.0 (67.148 sec) +INFO:tensorflow:global_step/sec: 1.53911 +INFO:tensorflow:step = 56101, loss = 0.150152, precision = 1.0 (64.973 sec) +INFO:tensorflow:global_step/sec: 1.53926 +INFO:tensorflow:step = 56201, loss = 0.148323, precision = 0.992188 (64.966 sec) +INFO:tensorflow:global_step/sec: 1.53854 +INFO:tensorflow:step = 56301, loss = 0.139611, precision = 1.0 (64.997 sec) +Saved checkpoint after 144 epoch(s) to data/resnet164/checkpoints/00144... +INFO:tensorflow:global_step/sec: 1.4885 +INFO:tensorflow:step = 56401, loss = 0.135566, precision = 1.0 (67.182 sec) +INFO:tensorflow:global_step/sec: 1.539 +INFO:tensorflow:step = 56501, loss = 0.1364, precision = 1.0 (64.977 sec) +INFO:tensorflow:global_step/sec: 1.53912 +INFO:tensorflow:step = 56601, loss = 0.137793, precision = 1.0 (64.972 sec) +Saved checkpoint after 145 epoch(s) to data/resnet164/checkpoints/00145... +INFO:tensorflow:global_step/sec: 1.48818 +INFO:tensorflow:step = 56701, loss = 0.136779, precision = 1.0 (67.196 sec) +INFO:tensorflow:global_step/sec: 1.54026 +INFO:tensorflow:step = 56801, loss = 0.139639, precision = 1.0 (64.924 sec) +INFO:tensorflow:global_step/sec: 1.53808 +INFO:tensorflow:step = 56901, loss = 0.138886, precision = 1.0 (65.016 sec) +INFO:tensorflow:global_step/sec: 1.53893 +INFO:tensorflow:step = 57001, loss = 0.137311, precision = 1.0 (64.980 sec) +Saved checkpoint after 146 epoch(s) to data/resnet164/checkpoints/00146... +INFO:tensorflow:global_step/sec: 1.48844 +INFO:tensorflow:step = 57101, loss = 0.137907, precision = 1.0 (67.184 sec) +INFO:tensorflow:global_step/sec: 1.53997 +INFO:tensorflow:step = 57201, loss = 0.141847, precision = 1.0 (64.936 sec) +INFO:tensorflow:global_step/sec: 1.53911 +INFO:tensorflow:step = 57301, loss = 0.151224, precision = 1.0 (64.973 sec) +INFO:tensorflow:global_step/sec: 1.53887 +INFO:tensorflow:step = 57401, loss = 0.138306, precision = 1.0 (64.983 sec) +Saved checkpoint after 147 epoch(s) to data/resnet164/checkpoints/00147... +INFO:tensorflow:global_step/sec: 1.48905 +INFO:tensorflow:step = 57501, loss = 0.135167, precision = 1.0 (67.157 sec) +INFO:tensorflow:global_step/sec: 1.53952 +INFO:tensorflow:step = 57601, loss = 0.134518, precision = 1.0 (64.955 sec) +INFO:tensorflow:global_step/sec: 1.53908 +INFO:tensorflow:step = 57701, loss = 0.16533, precision = 0.976562 (64.974 sec) +INFO:tensorflow:global_step/sec: 1.53946 +INFO:tensorflow:step = 57801, loss = 0.134852, precision = 1.0 (64.958 sec) +Saved checkpoint after 148 epoch(s) to data/resnet164/checkpoints/00148... +INFO:tensorflow:global_step/sec: 1.49032 +INFO:tensorflow:step = 57901, loss = 0.133127, precision = 1.0 (67.100 sec) +INFO:tensorflow:global_step/sec: 1.53948 +INFO:tensorflow:step = 58001, loss = 0.136448, precision = 1.0 (64.957 sec) +INFO:tensorflow:global_step/sec: 1.53907 +INFO:tensorflow:step = 58101, loss = 0.135039, precision = 1.0 (64.974 sec) +INFO:tensorflow:global_step/sec: 1.53876 +INFO:tensorflow:step = 58201, loss = 0.133162, precision = 1.0 (64.987 sec) +Saved checkpoint after 149 epoch(s) to data/resnet164/checkpoints/00149... +INFO:tensorflow:global_step/sec: 1.48838 +INFO:tensorflow:step = 58301, loss = 0.142333, precision = 1.0 (67.187 sec) +INFO:tensorflow:global_step/sec: 1.53933 +INFO:tensorflow:step = 58401, loss = 0.134281, precision = 1.0 (64.963 sec) +INFO:tensorflow:global_step/sec: 1.53872 +INFO:tensorflow:step = 58501, loss = 0.133638, precision = 1.0 (64.989 sec) +INFO:tensorflow:global_step/sec: 1.53923 +INFO:tensorflow:step = 58601, loss = 0.136008, precision = 1.0 (64.967 sec) +Saved checkpoint after 150 epoch(s) to data/resnet164/checkpoints/00150... +INFO:tensorflow:global_step/sec: 1.48699 +INFO:tensorflow:step = 58701, loss = 0.134927, precision = 1.0 (67.250 sec) +INFO:tensorflow:global_step/sec: 1.53933 +INFO:tensorflow:step = 58801, loss = 0.133564, precision = 1.0 (64.963 sec) +INFO:tensorflow:global_step/sec: 1.53887 +INFO:tensorflow:step = 58901, loss = 0.131281, precision = 1.0 (64.983 sec) +INFO:tensorflow:global_step/sec: 1.53853 +INFO:tensorflow:step = 59001, loss = 0.133589, precision = 1.0 (64.997 sec) +Saved checkpoint after 151 epoch(s) to data/resnet164/checkpoints/00151... +INFO:tensorflow:global_step/sec: 1.48702 +INFO:tensorflow:step = 59101, loss = 0.13193, precision = 1.0 (67.249 sec) +INFO:tensorflow:global_step/sec: 1.5392 +INFO:tensorflow:step = 59201, loss = 0.133622, precision = 1.0 (64.969 sec) +INFO:tensorflow:global_step/sec: 1.53911 +INFO:tensorflow:step = 59301, loss = 0.133641, precision = 1.0 (64.973 sec) +INFO:tensorflow:global_step/sec: 1.53813 +INFO:tensorflow:step = 59401, loss = 0.132041, precision = 1.0 (65.014 sec) +Saved checkpoint after 152 epoch(s) to data/resnet164/checkpoints/00152... +INFO:tensorflow:global_step/sec: 1.48212 +INFO:tensorflow:step = 59501, loss = 0.135449, precision = 1.0 (67.471 sec) +INFO:tensorflow:global_step/sec: 1.5388 +INFO:tensorflow:step = 59601, loss = 0.131657, precision = 1.0 (64.986 sec) +INFO:tensorflow:global_step/sec: 1.53811 +INFO:tensorflow:step = 59701, loss = 0.13365, precision = 1.0 (65.015 sec) +INFO:tensorflow:global_step/sec: 1.53777 +INFO:tensorflow:step = 59801, loss = 0.135519, precision = 1.0 (65.029 sec) +Saved checkpoint after 153 epoch(s) to data/resnet164/checkpoints/00153... +INFO:tensorflow:global_step/sec: 1.48711 +INFO:tensorflow:step = 59901, loss = 0.132837, precision = 1.0 (67.245 sec) +INFO:tensorflow:global_step/sec: 1.53853 +INFO:tensorflow:step = 60001, loss = 0.131177, precision = 1.0 (64.997 sec) +INFO:tensorflow:global_step/sec: 1.53805 +INFO:tensorflow:step = 60101, loss = 0.134473, precision = 1.0 (65.017 sec) +INFO:tensorflow:global_step/sec: 1.53867 +INFO:tensorflow:step = 60201, loss = 0.132343, precision = 1.0 (64.991 sec) +Saved checkpoint after 154 epoch(s) to data/resnet164/checkpoints/00154... +INFO:tensorflow:global_step/sec: 1.48719 +INFO:tensorflow:step = 60301, loss = 0.137987, precision = 0.992188 (67.241 sec) +INFO:tensorflow:global_step/sec: 1.53927 +INFO:tensorflow:step = 60401, loss = 0.1309, precision = 1.0 (64.966 sec) +INFO:tensorflow:global_step/sec: 1.53839 +INFO:tensorflow:step = 60501, loss = 0.13077, precision = 1.0 (65.003 sec) +INFO:tensorflow:global_step/sec: 1.53818 +INFO:tensorflow:step = 60601, loss = 0.129676, precision = 1.0 (65.012 sec) +Saved checkpoint after 155 epoch(s) to data/resnet164/checkpoints/00155... +INFO:tensorflow:global_step/sec: 1.48628 +INFO:tensorflow:step = 60701, loss = 0.134321, precision = 1.0 (67.282 sec) +INFO:tensorflow:global_step/sec: 1.53835 +INFO:tensorflow:step = 60801, loss = 0.132145, precision = 1.0 (65.005 sec) +INFO:tensorflow:global_step/sec: 1.53914 +INFO:tensorflow:step = 60901, loss = 0.128689, precision = 1.0 (64.971 sec) +Saved checkpoint after 156 epoch(s) to data/resnet164/checkpoints/00156... +INFO:tensorflow:global_step/sec: 1.48639 +INFO:tensorflow:step = 61001, loss = 0.132872, precision = 1.0 (67.277 sec) +INFO:tensorflow:global_step/sec: 1.53906 +INFO:tensorflow:step = 61101, loss = 0.137769, precision = 0.992188 (64.975 sec) +INFO:tensorflow:global_step/sec: 1.53875 +INFO:tensorflow:step = 61201, loss = 0.132783, precision = 1.0 (64.988 sec) +INFO:tensorflow:global_step/sec: 1.53795 +INFO:tensorflow:step = 61301, loss = 0.130126, precision = 1.0 (65.022 sec) +Saved checkpoint after 157 epoch(s) to data/resnet164/checkpoints/00157... +INFO:tensorflow:global_step/sec: 1.48641 +INFO:tensorflow:step = 61401, loss = 0.13144, precision = 1.0 (67.276 sec) +INFO:tensorflow:global_step/sec: 1.53905 +INFO:tensorflow:step = 61501, loss = 0.138985, precision = 1.0 (64.975 sec) +INFO:tensorflow:global_step/sec: 1.5372 +INFO:tensorflow:step = 61601, loss = 0.128053, precision = 1.0 (65.053 sec) +INFO:tensorflow:global_step/sec: 1.53777 +INFO:tensorflow:step = 61701, loss = 0.127998, precision = 1.0 (65.029 sec) +Saved checkpoint after 158 epoch(s) to data/resnet164/checkpoints/00158... +INFO:tensorflow:global_step/sec: 1.48674 +INFO:tensorflow:step = 61801, loss = 0.128371, precision = 1.0 (67.261 sec) +INFO:tensorflow:global_step/sec: 1.53858 +INFO:tensorflow:step = 61901, loss = 0.131399, precision = 1.0 (64.995 sec) +INFO:tensorflow:global_step/sec: 1.53835 +INFO:tensorflow:step = 62001, loss = 0.128016, precision = 1.0 (65.005 sec) +INFO:tensorflow:global_step/sec: 1.53798 +INFO:tensorflow:step = 62101, loss = 0.127999, precision = 1.0 (65.020 sec) +Saved checkpoint after 159 epoch(s) to data/resnet164/checkpoints/00159... +INFO:tensorflow:global_step/sec: 1.48641 +INFO:tensorflow:step = 62201, loss = 0.130226, precision = 1.0 (67.276 sec) +INFO:tensorflow:global_step/sec: 1.53888 +INFO:tensorflow:step = 62301, loss = 0.130155, precision = 1.0 (64.982 sec) +INFO:tensorflow:global_step/sec: 1.5379 +INFO:tensorflow:step = 62401, loss = 0.127652, precision = 1.0 (65.024 sec) +INFO:tensorflow:global_step/sec: 1.53759 +INFO:tensorflow:step = 62501, loss = 0.127892, precision = 1.0 (65.037 sec) +Saved checkpoint after 160 epoch(s) to data/resnet164/checkpoints/00160... +INFO:tensorflow:global_step/sec: 1.48678 +INFO:tensorflow:step = 62601, loss = 0.129296, precision = 1.0 (67.260 sec) +INFO:tensorflow:global_step/sec: 1.53861 +INFO:tensorflow:step = 62701, loss = 0.132905, precision = 1.0 (64.994 sec) +INFO:tensorflow:global_step/sec: 1.5376 +INFO:tensorflow:step = 62801, loss = 0.127103, precision = 1.0 (65.036 sec) +INFO:tensorflow:global_step/sec: 1.53886 +INFO:tensorflow:step = 62901, loss = 0.126885, precision = 1.0 (64.983 sec) +Saved checkpoint after 161 epoch(s) to data/resnet164/checkpoints/00161... +INFO:tensorflow:global_step/sec: 1.48664 +INFO:tensorflow:step = 63001, loss = 0.128375, precision = 1.0 (67.266 sec) +INFO:tensorflow:global_step/sec: 1.53831 +INFO:tensorflow:step = 63101, loss = 0.12891, precision = 1.0 (65.006 sec) +INFO:tensorflow:global_step/sec: 1.538 +INFO:tensorflow:step = 63201, loss = 0.125945, precision = 1.0 (65.019 sec) +INFO:tensorflow:global_step/sec: 1.53829 +INFO:tensorflow:step = 63301, loss = 0.125607, precision = 1.0 (65.007 sec) +Saved checkpoint after 162 epoch(s) to data/resnet164/checkpoints/00162... +INFO:tensorflow:global_step/sec: 1.48155 +INFO:tensorflow:step = 63401, loss = 0.126238, precision = 1.0 (67.497 sec) +INFO:tensorflow:global_step/sec: 1.53812 +INFO:tensorflow:step = 63501, loss = 0.126062, precision = 1.0 (65.014 sec) +INFO:tensorflow:global_step/sec: 1.53815 +INFO:tensorflow:step = 63601, loss = 0.127201, precision = 1.0 (65.013 sec) +INFO:tensorflow:global_step/sec: 1.53766 +INFO:tensorflow:step = 63701, loss = 0.126081, precision = 1.0 (65.034 sec) +Saved checkpoint after 163 epoch(s) to data/resnet164/checkpoints/00163... +INFO:tensorflow:global_step/sec: 1.48573 +INFO:tensorflow:step = 63801, loss = 0.126751, precision = 1.0 (67.307 sec) +INFO:tensorflow:global_step/sec: 1.53888 +INFO:tensorflow:step = 63901, loss = 0.130123, precision = 1.0 (64.982 sec) +INFO:tensorflow:global_step/sec: 1.53807 +INFO:tensorflow:step = 64001, loss = 0.126276, precision = 1.0 (65.017 sec) +INFO:tensorflow:global_step/sec: 1.53769 +INFO:tensorflow:step = 64101, loss = 0.127378, precision = 1.0 (65.033 sec) +Saved checkpoint after 164 epoch(s) to data/resnet164/checkpoints/00164... +INFO:tensorflow:global_step/sec: 1.48681 +INFO:tensorflow:step = 64201, loss = 0.124476, precision = 1.0 (67.258 sec) +INFO:tensorflow:global_step/sec: 1.53903 +INFO:tensorflow:step = 64301, loss = 0.12731, precision = 1.0 (64.976 sec) +INFO:tensorflow:global_step/sec: 1.53779 +INFO:tensorflow:step = 64401, loss = 0.126792, precision = 1.0 (65.029 sec) +INFO:tensorflow:global_step/sec: 1.53857 +INFO:tensorflow:step = 64501, loss = 0.125204, precision = 1.0 (64.996 sec) +Saved checkpoint after 165 epoch(s) to data/resnet164/checkpoints/00165... +INFO:tensorflow:global_step/sec: 1.48716 +INFO:tensorflow:step = 64601, loss = 0.124679, precision = 1.0 (67.242 sec) +INFO:tensorflow:global_step/sec: 1.53861 +INFO:tensorflow:step = 64701, loss = 0.125179, precision = 1.0 (64.994 sec) +INFO:tensorflow:global_step/sec: 1.53798 +INFO:tensorflow:step = 64801, loss = 0.124974, precision = 1.0 (65.020 sec) +INFO:tensorflow:global_step/sec: 1.53835 +INFO:tensorflow:step = 64901, loss = 0.125666, precision = 1.0 (65.005 sec) +Saved checkpoint after 166 epoch(s) to data/resnet164/checkpoints/00166... +INFO:tensorflow:global_step/sec: 1.4862 +INFO:tensorflow:step = 65001, loss = 0.124605, precision = 1.0 (67.286 sec) +INFO:tensorflow:global_step/sec: 1.53864 +INFO:tensorflow:step = 65101, loss = 0.123877, precision = 1.0 (64.992 sec) +INFO:tensorflow:global_step/sec: 1.53872 +INFO:tensorflow:step = 65201, loss = 0.127779, precision = 1.0 (64.989 sec) +Saved checkpoint after 167 epoch(s) to data/resnet164/checkpoints/00167... +INFO:tensorflow:global_step/sec: 1.48604 +INFO:tensorflow:step = 65301, loss = 0.126244, precision = 1.0 (67.293 sec) +INFO:tensorflow:global_step/sec: 1.53876 +INFO:tensorflow:step = 65401, loss = 0.123958, precision = 1.0 (64.987 sec) +INFO:tensorflow:global_step/sec: 1.53813 +INFO:tensorflow:step = 65501, loss = 0.126475, precision = 1.0 (65.014 sec) +INFO:tensorflow:global_step/sec: 1.53856 +INFO:tensorflow:step = 65601, loss = 0.124017, precision = 1.0 (64.996 sec) +Saved checkpoint after 168 epoch(s) to data/resnet164/checkpoints/00168... +INFO:tensorflow:global_step/sec: 1.48669 +INFO:tensorflow:step = 65701, loss = 0.123888, precision = 1.0 (67.264 sec) +INFO:tensorflow:global_step/sec: 1.53916 +INFO:tensorflow:step = 65801, loss = 0.124605, precision = 1.0 (64.970 sec) +INFO:tensorflow:global_step/sec: 1.53769 +INFO:tensorflow:step = 65901, loss = 0.123875, precision = 1.0 (65.033 sec) +INFO:tensorflow:global_step/sec: 1.53865 +INFO:tensorflow:step = 66001, loss = 0.124466, precision = 1.0 (64.992 sec) +Saved checkpoint after 169 epoch(s) to data/resnet164/checkpoints/00169... +INFO:tensorflow:global_step/sec: 1.48641 +INFO:tensorflow:step = 66101, loss = 0.126651, precision = 1.0 (67.276 sec) +INFO:tensorflow:global_step/sec: 1.53979 +INFO:tensorflow:step = 66201, loss = 0.122328, precision = 1.0 (64.944 sec) +INFO:tensorflow:global_step/sec: 1.53808 +INFO:tensorflow:step = 66301, loss = 0.12663, precision = 1.0 (65.016 sec) +INFO:tensorflow:global_step/sec: 1.53853 +INFO:tensorflow:step = 66401, loss = 0.122084, precision = 1.0 (64.997 sec) +Saved checkpoint after 170 epoch(s) to data/resnet164/checkpoints/00170... +INFO:tensorflow:global_step/sec: 1.48697 +INFO:tensorflow:step = 66501, loss = 0.122331, precision = 1.0 (67.251 sec) +INFO:tensorflow:global_step/sec: 1.53852 +INFO:tensorflow:step = 66601, loss = 0.125198, precision = 1.0 (64.998 sec) +INFO:tensorflow:global_step/sec: 1.53821 +INFO:tensorflow:step = 66701, loss = 0.12492, precision = 1.0 (65.011 sec) +INFO:tensorflow:global_step/sec: 1.53829 +INFO:tensorflow:step = 66801, loss = 0.121473, precision = 1.0 (65.007 sec) +Saved checkpoint after 171 epoch(s) to data/resnet164/checkpoints/00171... +INFO:tensorflow:global_step/sec: 1.48648 +INFO:tensorflow:step = 66901, loss = 0.124291, precision = 1.0 (67.273 sec) +INFO:tensorflow:global_step/sec: 1.53866 +INFO:tensorflow:step = 67001, loss = 0.123651, precision = 1.0 (64.992 sec) +INFO:tensorflow:global_step/sec: 1.53783 +INFO:tensorflow:step = 67101, loss = 0.12199, precision = 1.0 (65.027 sec) +INFO:tensorflow:global_step/sec: 1.53783 +INFO:tensorflow:step = 67201, loss = 0.121449, precision = 1.0 (65.027 sec) +Saved checkpoint after 172 epoch(s) to data/resnet164/checkpoints/00172... +INFO:tensorflow:global_step/sec: 1.4804 +INFO:tensorflow:step = 67301, loss = 0.120832, precision = 1.0 (67.549 sec) +INFO:tensorflow:global_step/sec: 1.53848 +INFO:tensorflow:step = 67401, loss = 0.121871, precision = 1.0 (64.999 sec) +INFO:tensorflow:global_step/sec: 1.53856 +INFO:tensorflow:step = 67501, loss = 0.120781, precision = 1.0 (64.996 sec) +INFO:tensorflow:global_step/sec: 1.53781 +INFO:tensorflow:step = 67601, loss = 0.121025, precision = 1.0 (65.027 sec) +Saved checkpoint after 173 epoch(s) to data/resnet164/checkpoints/00173... +INFO:tensorflow:global_step/sec: 1.4857 +INFO:tensorflow:step = 67701, loss = 0.121746, precision = 1.0 (67.309 sec) +INFO:tensorflow:global_step/sec: 1.53804 +INFO:tensorflow:step = 67801, loss = 0.120598, precision = 1.0 (65.017 sec) +INFO:tensorflow:global_step/sec: 1.53805 +INFO:tensorflow:step = 67901, loss = 0.135905, precision = 0.992188 (65.017 sec) +INFO:tensorflow:global_step/sec: 1.53766 +INFO:tensorflow:step = 68001, loss = 0.120879, precision = 1.0 (65.034 sec) +Saved checkpoint after 174 epoch(s) to data/resnet164/checkpoints/00174... +INFO:tensorflow:global_step/sec: 1.48593 +INFO:tensorflow:step = 68101, loss = 0.121888, precision = 1.0 (67.298 sec) +INFO:tensorflow:global_step/sec: 1.53771 +INFO:tensorflow:step = 68201, loss = 0.120469, precision = 1.0 (65.031 sec) +INFO:tensorflow:global_step/sec: 1.5375 +INFO:tensorflow:step = 68301, loss = 0.120527, precision = 1.0 (65.041 sec) +INFO:tensorflow:global_step/sec: 1.53836 +INFO:tensorflow:step = 68401, loss = 0.122612, precision = 1.0 (65.004 sec) +Saved checkpoint after 175 epoch(s) to data/resnet164/checkpoints/00175... +INFO:tensorflow:global_step/sec: 1.48595 +INFO:tensorflow:step = 68501, loss = 0.120928, precision = 1.0 (67.297 sec) +INFO:tensorflow:global_step/sec: 1.53764 +INFO:tensorflow:step = 68601, loss = 0.121234, precision = 1.0 (65.034 sec) +INFO:tensorflow:global_step/sec: 1.53785 +INFO:tensorflow:step = 68701, loss = 0.120552, precision = 1.0 (65.026 sec) +INFO:tensorflow:global_step/sec: 1.53754 +INFO:tensorflow:step = 68801, loss = 0.1195, precision = 1.0 (65.039 sec) +Saved checkpoint after 176 epoch(s) to data/resnet164/checkpoints/00176... +INFO:tensorflow:global_step/sec: 1.48544 +INFO:tensorflow:step = 68901, loss = 0.144473, precision = 0.992188 (67.320 sec) +INFO:tensorflow:global_step/sec: 1.5389 +INFO:tensorflow:step = 69001, loss = 0.11886, precision = 1.0 (64.981 sec) +INFO:tensorflow:global_step/sec: 1.53788 +INFO:tensorflow:step = 69101, loss = 0.123331, precision = 1.0 (65.025 sec) +INFO:tensorflow:global_step/sec: 1.53767 +INFO:tensorflow:step = 69201, loss = 0.119051, precision = 1.0 (65.034 sec) +Saved checkpoint after 177 epoch(s) to data/resnet164/checkpoints/00177... +INFO:tensorflow:global_step/sec: 1.48658 +INFO:tensorflow:step = 69301, loss = 0.118922, precision = 1.0 (67.269 sec) +INFO:tensorflow:global_step/sec: 1.53905 +INFO:tensorflow:step = 69401, loss = 0.120846, precision = 1.0 (64.975 sec) +INFO:tensorflow:global_step/sec: 1.53807 +INFO:tensorflow:step = 69501, loss = 0.119982, precision = 1.0 (65.017 sec) +Saved checkpoint after 178 epoch(s) to data/resnet164/checkpoints/00178... +INFO:tensorflow:global_step/sec: 1.48599 +INFO:tensorflow:step = 69601, loss = 0.118224, precision = 1.0 (67.295 sec) +INFO:tensorflow:global_step/sec: 1.53807 +INFO:tensorflow:step = 69701, loss = 0.11938, precision = 1.0 (65.017 sec) +INFO:tensorflow:global_step/sec: 1.53814 +INFO:tensorflow:step = 69801, loss = 0.118369, precision = 1.0 (65.014 sec) +INFO:tensorflow:global_step/sec: 1.53799 +INFO:tensorflow:step = 69901, loss = 0.11899, precision = 1.0 (65.020 sec) +Saved checkpoint after 179 epoch(s) to data/resnet164/checkpoints/00179... +INFO:tensorflow:global_step/sec: 1.48625 +INFO:tensorflow:step = 70001, loss = 0.120607, precision = 1.0 (67.284 sec) +INFO:tensorflow:global_step/sec: 1.53899 +INFO:tensorflow:step = 70101, loss = 0.118053, precision = 1.0 (64.978 sec) +INFO:tensorflow:global_step/sec: 1.53754 +INFO:tensorflow:step = 70201, loss = 0.117693, precision = 1.0 (65.039 sec) +INFO:tensorflow:global_step/sec: 1.53874 +INFO:tensorflow:step = 70301, loss = 0.122463, precision = 1.0 (64.988 sec) +Saved checkpoint after 180 epoch(s) to data/resnet164/checkpoints/00180... +INFO:tensorflow:global_step/sec: 1.4863 +INFO:tensorflow:step = 70401, loss = 0.121145, precision = 1.0 (67.281 sec) +INFO:tensorflow:global_step/sec: 1.53823 +INFO:tensorflow:step = 70501, loss = 0.118655, precision = 1.0 (65.010 sec) +INFO:tensorflow:global_step/sec: 1.53749 +INFO:tensorflow:step = 70601, loss = 0.117961, precision = 1.0 (65.041 sec) +INFO:tensorflow:global_step/sec: 1.53851 +INFO:tensorflow:step = 70701, loss = 0.117923, precision = 1.0 (64.998 sec) +Saved checkpoint after 181 epoch(s) to data/resnet164/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1k80_ec2/resnet20_train.log b/tensorflow/CIFAR10/logs/1k80_ec2/resnet20_train.log new file mode 100644 index 0000000..156772c --- /dev/null +++ b/tensorflow/CIFAR10/logs/1k80_ec2/resnet20_train.log @@ -0,0 +1,1728 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/269.03k params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/10.38b flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + +======================End of Report========================== +2017-07-30 07:24:35.819460: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero +2017-07-30 07:24:35.819959: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla K80 +major: 3 minor: 7 memoryClockRate (GHz) 0.8235 +pciBusID 0000:00:1e.0 +Total memory: 11.17GiB +Free memory: 11.11GiB +2017-07-30 07:24:35.819978: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-07-30 07:24:35.819984: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-07-30 07:24:35.819999: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:00:1e.0) +2017-07-30 07:24:36.740963: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 07:24:36.741009: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 4 visible devices +2017-07-30 07:24:36.742296: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x50d36a0 executing computations on platform Host. Devices: +2017-07-30 07:24:36.742312: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-07-30 07:24:36.742976: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 07:24:36.742995: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 4 visible devices +2017-07-30 07:24:36.744278: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x4ffcfb0 executing computations on platform CUDA. Devices: +2017-07-30 07:24:36.744292: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla K80, Compute Capability 3.7 +INFO:tensorflow:step = 1, loss = 2.70508, precision = 0.078125 +INFO:tensorflow:global_step/sec: 5.88805 +INFO:tensorflow:step = 101, loss = 2.04826, precision = 0.390625 (16.984 sec) +INFO:tensorflow:global_step/sec: 6.36517 +INFO:tensorflow:step = 201, loss = 1.8057, precision = 0.4375 (15.710 sec) +INFO:tensorflow:global_step/sec: 6.67586 +INFO:tensorflow:step = 301, loss = 1.691, precision = 0.539062 (14.979 sec) +total_params: 269034 +Saved checkpoint after 1 epoch(s) to data/resnet20/checkpoints/00001... +INFO:tensorflow:global_step/sec: 7.42299 +INFO:tensorflow:step = 401, loss = 2.11039, precision = 0.289062 (13.472 sec) +INFO:tensorflow:global_step/sec: 11.7645 +INFO:tensorflow:step = 501, loss = 1.58717, precision = 0.53125 (8.500 sec) +INFO:tensorflow:global_step/sec: 11.7423 +INFO:tensorflow:step = 601, loss = 1.40883, precision = 0.578125 (8.516 sec) +INFO:tensorflow:global_step/sec: 11.7318 +INFO:tensorflow:step = 701, loss = 1.46244, precision = 0.585938 (8.524 sec) +Saved checkpoint after 2 epoch(s) to data/resnet20/checkpoints/00002... +INFO:tensorflow:global_step/sec: 11.2939 +INFO:tensorflow:step = 801, loss = 1.37342, precision = 0.648438 (8.854 sec) +INFO:tensorflow:global_step/sec: 11.7117 +INFO:tensorflow:step = 901, loss = 1.24473, precision = 0.648438 (8.538 sec) +INFO:tensorflow:global_step/sec: 11.694 +INFO:tensorflow:step = 1001, loss = 1.33989, precision = 0.640625 (8.551 sec) +INFO:tensorflow:global_step/sec: 11.7048 +INFO:tensorflow:step = 1101, loss = 1.11862, precision = 0.679688 (8.544 sec) +Saved checkpoint after 3 epoch(s) to data/resnet20/checkpoints/00003... +INFO:tensorflow:global_step/sec: 11.3335 +INFO:tensorflow:step = 1201, loss = 1.10733, precision = 0.6875 (8.824 sec) +INFO:tensorflow:global_step/sec: 11.7146 +INFO:tensorflow:step = 1301, loss = 1.02881, precision = 0.726562 (8.536 sec) +INFO:tensorflow:global_step/sec: 11.7247 +INFO:tensorflow:step = 1401, loss = 1.03657, precision = 0.710938 (8.529 sec) +INFO:tensorflow:global_step/sec: 11.6708 +INFO:tensorflow:step = 1501, loss = 1.1092, precision = 0.703125 (8.568 sec) +Saved checkpoint after 4 epoch(s) to data/resnet20/checkpoints/00004... +INFO:tensorflow:global_step/sec: 11.3245 +INFO:tensorflow:step = 1601, loss = 0.908842, precision = 0.726562 (8.831 sec) +INFO:tensorflow:global_step/sec: 11.7263 +INFO:tensorflow:step = 1701, loss = 0.899166, precision = 0.734375 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7192 +INFO:tensorflow:step = 1801, loss = 0.92482, precision = 0.757812 (8.533 sec) +INFO:tensorflow:global_step/sec: 11.7138 +INFO:tensorflow:step = 1901, loss = 0.909458, precision = 0.78125 (8.537 sec) +Saved checkpoint after 5 epoch(s) to data/resnet20/checkpoints/00005... +INFO:tensorflow:global_step/sec: 11.3324 +INFO:tensorflow:step = 2001, loss = 0.912629, precision = 0.765625 (8.824 sec) +INFO:tensorflow:global_step/sec: 11.7394 +INFO:tensorflow:step = 2101, loss = 0.972707, precision = 0.757812 (8.518 sec) +INFO:tensorflow:global_step/sec: 11.6742 +INFO:tensorflow:step = 2201, loss = 0.760366, precision = 0.828125 (8.566 sec) +INFO:tensorflow:global_step/sec: 11.7019 +INFO:tensorflow:step = 2301, loss = 0.892651, precision = 0.757812 (8.546 sec) +Saved checkpoint after 6 epoch(s) to data/resnet20/checkpoints/00006... +INFO:tensorflow:global_step/sec: 11.3425 +INFO:tensorflow:step = 2401, loss = 0.826753, precision = 0.804688 (8.816 sec) +INFO:tensorflow:global_step/sec: 11.7316 +INFO:tensorflow:step = 2501, loss = 0.806918, precision = 0.804688 (8.524 sec) +INFO:tensorflow:global_step/sec: 11.6899 +INFO:tensorflow:step = 2601, loss = 0.879, precision = 0.78125 (8.554 sec) +INFO:tensorflow:global_step/sec: 11.7039 +INFO:tensorflow:step = 2701, loss = 0.784762, precision = 0.789062 (8.544 sec) +Saved checkpoint after 7 epoch(s) to data/resnet20/checkpoints/00007... +INFO:tensorflow:global_step/sec: 11.3541 +INFO:tensorflow:step = 2801, loss = 0.807531, precision = 0.8125 (8.807 sec) +INFO:tensorflow:global_step/sec: 11.6978 +INFO:tensorflow:step = 2901, loss = 0.823573, precision = 0.78125 (8.548 sec) +INFO:tensorflow:global_step/sec: 11.7017 +INFO:tensorflow:step = 3001, loss = 0.786019, precision = 0.765625 (8.546 sec) +INFO:tensorflow:global_step/sec: 11.6904 +INFO:tensorflow:step = 3101, loss = 0.756386, precision = 0.8125 (8.554 sec) +Saved checkpoint after 8 epoch(s) to data/resnet20/checkpoints/00008... +INFO:tensorflow:global_step/sec: 11.3358 +INFO:tensorflow:step = 3201, loss = 0.702099, precision = 0.8125 (8.822 sec) +INFO:tensorflow:global_step/sec: 11.7081 +INFO:tensorflow:step = 3301, loss = 0.755381, precision = 0.820312 (8.541 sec) +INFO:tensorflow:global_step/sec: 11.7122 +INFO:tensorflow:step = 3401, loss = 0.732768, precision = 0.804688 (8.538 sec) +INFO:tensorflow:global_step/sec: 11.6989 +INFO:tensorflow:step = 3501, loss = 0.784596, precision = 0.789062 (8.548 sec) +Saved checkpoint after 9 epoch(s) to data/resnet20/checkpoints/00009... +INFO:tensorflow:global_step/sec: 11.3552 +INFO:tensorflow:step = 3601, loss = 0.638817, precision = 0.875 (8.807 sec) +INFO:tensorflow:global_step/sec: 11.711 +INFO:tensorflow:step = 3701, loss = 0.871964, precision = 0.757812 (8.539 sec) +INFO:tensorflow:global_step/sec: 11.7179 +INFO:tensorflow:step = 3801, loss = 0.681936, precision = 0.835938 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7011 +INFO:tensorflow:step = 3901, loss = 0.631837, precision = 0.882812 (8.546 sec) +Saved checkpoint after 10 epoch(s) to data/resnet20/checkpoints/00010... +INFO:tensorflow:global_step/sec: 11.3277 +INFO:tensorflow:step = 4001, loss = 0.748999, precision = 0.804688 (8.828 sec) +INFO:tensorflow:global_step/sec: 11.7121 +INFO:tensorflow:step = 4101, loss = 0.703764, precision = 0.828125 (8.538 sec) +INFO:tensorflow:global_step/sec: 11.7255 +INFO:tensorflow:step = 4201, loss = 0.772281, precision = 0.828125 (8.528 sec) +Saved checkpoint after 11 epoch(s) to data/resnet20/checkpoints/00011... +INFO:tensorflow:global_step/sec: 11.3771 +INFO:tensorflow:step = 4301, loss = 0.723411, precision = 0.804688 (8.790 sec) +INFO:tensorflow:global_step/sec: 11.7105 +INFO:tensorflow:step = 4401, loss = 0.759331, precision = 0.820312 (8.539 sec) +INFO:tensorflow:global_step/sec: 11.7275 +INFO:tensorflow:step = 4501, loss = 0.823486, precision = 0.789062 (8.527 sec) +INFO:tensorflow:global_step/sec: 11.6885 +INFO:tensorflow:step = 4601, loss = 0.691031, precision = 0.835938 (8.555 sec) +Saved checkpoint after 12 epoch(s) to data/resnet20/checkpoints/00012... +INFO:tensorflow:global_step/sec: 11.3299 +INFO:tensorflow:step = 4701, loss = 0.743962, precision = 0.84375 (8.826 sec) +INFO:tensorflow:global_step/sec: 11.717 +INFO:tensorflow:step = 4801, loss = 0.802274, precision = 0.804688 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7315 +INFO:tensorflow:step = 4901, loss = 0.752785, precision = 0.828125 (8.524 sec) +INFO:tensorflow:global_step/sec: 11.7127 +INFO:tensorflow:step = 5001, loss = 0.744115, precision = 0.828125 (8.538 sec) +Saved checkpoint after 13 epoch(s) to data/resnet20/checkpoints/00013... +INFO:tensorflow:global_step/sec: 11.3509 +INFO:tensorflow:step = 5101, loss = 0.669418, precision = 0.859375 (8.810 sec) +INFO:tensorflow:global_step/sec: 11.7381 +INFO:tensorflow:step = 5201, loss = 0.785193, precision = 0.835938 (8.519 sec) +INFO:tensorflow:global_step/sec: 11.7378 +INFO:tensorflow:step = 5301, loss = 0.672324, precision = 0.859375 (8.519 sec) +INFO:tensorflow:global_step/sec: 11.7548 +INFO:tensorflow:step = 5401, loss = 0.634646, precision = 0.882812 (8.507 sec) +Saved checkpoint after 14 epoch(s) to data/resnet20/checkpoints/00014... +INFO:tensorflow:global_step/sec: 11.3209 +INFO:tensorflow:step = 5501, loss = 0.619361, precision = 0.875 (8.833 sec) +INFO:tensorflow:global_step/sec: 11.7388 +INFO:tensorflow:step = 5601, loss = 0.7144, precision = 0.851562 (8.519 sec) +INFO:tensorflow:global_step/sec: 11.7006 +INFO:tensorflow:step = 5701, loss = 0.722199, precision = 0.804688 (8.547 sec) +INFO:tensorflow:global_step/sec: 11.7052 +INFO:tensorflow:step = 5801, loss = 0.643883, precision = 0.882812 (8.543 sec) +Saved checkpoint after 15 epoch(s) to data/resnet20/checkpoints/00015... +INFO:tensorflow:global_step/sec: 11.3481 +INFO:tensorflow:step = 5901, loss = 0.742739, precision = 0.8125 (8.812 sec) +INFO:tensorflow:global_step/sec: 11.7427 +INFO:tensorflow:step = 6001, loss = 0.681583, precision = 0.8125 (8.516 sec) +INFO:tensorflow:global_step/sec: 11.7176 +INFO:tensorflow:step = 6101, loss = 0.82453, precision = 0.804688 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7332 +INFO:tensorflow:step = 6201, loss = 0.766456, precision = 0.8125 (8.523 sec) +Saved checkpoint after 16 epoch(s) to data/resnet20/checkpoints/00016... +INFO:tensorflow:global_step/sec: 11.3848 +INFO:tensorflow:step = 6301, loss = 0.827716, precision = 0.820312 (8.784 sec) +INFO:tensorflow:global_step/sec: 11.7019 +INFO:tensorflow:step = 6401, loss = 0.797942, precision = 0.828125 (8.546 sec) +INFO:tensorflow:global_step/sec: 11.6983 +INFO:tensorflow:step = 6501, loss = 0.750959, precision = 0.8125 (8.548 sec) +INFO:tensorflow:global_step/sec: 11.7144 +INFO:tensorflow:step = 6601, loss = 0.729449, precision = 0.804688 (8.536 sec) +Saved checkpoint after 17 epoch(s) to data/resnet20/checkpoints/00017... +INFO:tensorflow:global_step/sec: 11.3573 +INFO:tensorflow:step = 6701, loss = 0.683554, precision = 0.859375 (8.805 sec) +INFO:tensorflow:global_step/sec: 11.7183 +INFO:tensorflow:step = 6801, loss = 0.760272, precision = 0.835938 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7178 +INFO:tensorflow:step = 6901, loss = 0.656102, precision = 0.859375 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7312 +INFO:tensorflow:step = 7001, loss = 0.70816, precision = 0.859375 (8.524 sec) +Saved checkpoint after 18 epoch(s) to data/resnet20/checkpoints/00018... +INFO:tensorflow:global_step/sec: 11.3185 +INFO:tensorflow:step = 7101, loss = 0.590043, precision = 0.875 (8.835 sec) +INFO:tensorflow:global_step/sec: 11.7446 +INFO:tensorflow:step = 7201, loss = 0.691524, precision = 0.804688 (8.514 sec) +INFO:tensorflow:global_step/sec: 11.7294 +INFO:tensorflow:step = 7301, loss = 0.692819, precision = 0.859375 (8.526 sec) +INFO:tensorflow:global_step/sec: 11.7223 +INFO:tensorflow:step = 7401, loss = 0.631987, precision = 0.898438 (8.531 sec) +Saved checkpoint after 19 epoch(s) to data/resnet20/checkpoints/00019... +INFO:tensorflow:global_step/sec: 11.376 +INFO:tensorflow:step = 7501, loss = 0.790159, precision = 0.835938 (8.791 sec) +INFO:tensorflow:global_step/sec: 11.7262 +INFO:tensorflow:step = 7601, loss = 0.623453, precision = 0.835938 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7331 +INFO:tensorflow:step = 7701, loss = 0.767728, precision = 0.828125 (8.523 sec) +INFO:tensorflow:global_step/sec: 11.7195 +INFO:tensorflow:step = 7801, loss = 0.681587, precision = 0.851562 (8.533 sec) +Saved checkpoint after 20 epoch(s) to data/resnet20/checkpoints/00020... +INFO:tensorflow:global_step/sec: 11.2662 +INFO:tensorflow:step = 7901, loss = 0.614336, precision = 0.875 (8.876 sec) +INFO:tensorflow:global_step/sec: 11.7127 +INFO:tensorflow:step = 8001, loss = 0.637372, precision = 0.859375 (8.538 sec) +INFO:tensorflow:global_step/sec: 11.7139 +INFO:tensorflow:step = 8101, loss = 0.665357, precision = 0.859375 (8.537 sec) +INFO:tensorflow:global_step/sec: 11.7209 +INFO:tensorflow:step = 8201, loss = 0.716565, precision = 0.804688 (8.532 sec) +Saved checkpoint after 21 epoch(s) to data/resnet20/checkpoints/00021... +INFO:tensorflow:global_step/sec: 11.3012 +INFO:tensorflow:step = 8301, loss = 0.783658, precision = 0.859375 (8.849 sec) +INFO:tensorflow:global_step/sec: 11.7278 +INFO:tensorflow:step = 8401, loss = 0.726745, precision = 0.796875 (8.527 sec) +INFO:tensorflow:global_step/sec: 11.7345 +INFO:tensorflow:step = 8501, loss = 0.708003, precision = 0.84375 (8.522 sec) +INFO:tensorflow:global_step/sec: 11.7405 +INFO:tensorflow:step = 8601, loss = 0.63671, precision = 0.882812 (8.518 sec) +Saved checkpoint after 22 epoch(s) to data/resnet20/checkpoints/00022... +INFO:tensorflow:global_step/sec: 11.3339 +INFO:tensorflow:step = 8701, loss = 0.668923, precision = 0.867188 (8.823 sec) +INFO:tensorflow:global_step/sec: 11.7294 +INFO:tensorflow:step = 8801, loss = 0.745158, precision = 0.84375 (8.526 sec) +INFO:tensorflow:global_step/sec: 11.7374 +INFO:tensorflow:step = 8901, loss = 0.695335, precision = 0.84375 (8.520 sec) +Saved checkpoint after 23 epoch(s) to data/resnet20/checkpoints/00023... +INFO:tensorflow:global_step/sec: 11.3456 +INFO:tensorflow:step = 9001, loss = 0.756385, precision = 0.8125 (8.814 sec) +INFO:tensorflow:global_step/sec: 11.7258 +INFO:tensorflow:step = 9101, loss = 0.595737, precision = 0.875 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7617 +INFO:tensorflow:step = 9201, loss = 0.757607, precision = 0.820312 (8.502 sec) +INFO:tensorflow:global_step/sec: 11.7352 +INFO:tensorflow:step = 9301, loss = 0.716109, precision = 0.820312 (8.521 sec) +Saved checkpoint after 24 epoch(s) to data/resnet20/checkpoints/00024... +INFO:tensorflow:global_step/sec: 11.3429 +INFO:tensorflow:step = 9401, loss = 0.695088, precision = 0.84375 (8.816 sec) +INFO:tensorflow:global_step/sec: 11.7294 +INFO:tensorflow:step = 9501, loss = 0.612953, precision = 0.867188 (8.525 sec) +INFO:tensorflow:global_step/sec: 11.7172 +INFO:tensorflow:step = 9601, loss = 0.662777, precision = 0.859375 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7067 +INFO:tensorflow:step = 9701, loss = 0.708215, precision = 0.8125 (8.542 sec) +Saved checkpoint after 25 epoch(s) to data/resnet20/checkpoints/00025... +INFO:tensorflow:global_step/sec: 11.3387 +INFO:tensorflow:step = 9801, loss = 0.631295, precision = 0.890625 (8.819 sec) +INFO:tensorflow:global_step/sec: 11.7504 +INFO:tensorflow:step = 9901, loss = 0.697828, precision = 0.851562 (8.510 sec) +INFO:tensorflow:global_step/sec: 11.7196 +INFO:tensorflow:step = 10001, loss = 0.703215, precision = 0.875 (8.533 sec) +INFO:tensorflow:global_step/sec: 11.7285 +INFO:tensorflow:step = 10101, loss = 0.625616, precision = 0.898438 (8.526 sec) +Saved checkpoint after 26 epoch(s) to data/resnet20/checkpoints/00026... +INFO:tensorflow:global_step/sec: 11.3426 +INFO:tensorflow:step = 10201, loss = 0.708378, precision = 0.84375 (8.817 sec) +INFO:tensorflow:global_step/sec: 11.7498 +INFO:tensorflow:step = 10301, loss = 0.6045, precision = 0.898438 (8.511 sec) +INFO:tensorflow:global_step/sec: 11.7582 +INFO:tensorflow:step = 10401, loss = 0.601678, precision = 0.90625 (8.505 sec) +INFO:tensorflow:global_step/sec: 11.7256 +INFO:tensorflow:step = 10501, loss = 0.636338, precision = 0.84375 (8.528 sec) +Saved checkpoint after 27 epoch(s) to data/resnet20/checkpoints/00027... +INFO:tensorflow:global_step/sec: 11.3385 +INFO:tensorflow:step = 10601, loss = 0.64685, precision = 0.867188 (8.820 sec) +INFO:tensorflow:global_step/sec: 11.764 +INFO:tensorflow:step = 10701, loss = 0.614454, precision = 0.882812 (8.500 sec) +INFO:tensorflow:global_step/sec: 11.7442 +INFO:tensorflow:step = 10801, loss = 0.67378, precision = 0.84375 (8.515 sec) +INFO:tensorflow:global_step/sec: 11.7415 +INFO:tensorflow:step = 10901, loss = 0.646939, precision = 0.882812 (8.517 sec) +Saved checkpoint after 28 epoch(s) to data/resnet20/checkpoints/00028... +INFO:tensorflow:global_step/sec: 11.3667 +INFO:tensorflow:step = 11001, loss = 0.852764, precision = 0.765625 (8.798 sec) +INFO:tensorflow:global_step/sec: 11.7354 +INFO:tensorflow:step = 11101, loss = 0.74999, precision = 0.851562 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7479 +INFO:tensorflow:step = 11201, loss = 0.612951, precision = 0.882812 (8.512 sec) +INFO:tensorflow:global_step/sec: 11.7335 +INFO:tensorflow:step = 11301, loss = 0.669274, precision = 0.851562 (8.523 sec) +Saved checkpoint after 29 epoch(s) to data/resnet20/checkpoints/00029... +INFO:tensorflow:global_step/sec: 11.3741 +INFO:tensorflow:step = 11401, loss = 0.594972, precision = 0.875 (8.792 sec) +INFO:tensorflow:global_step/sec: 11.7519 +INFO:tensorflow:step = 11501, loss = 0.737729, precision = 0.851562 (8.509 sec) +INFO:tensorflow:global_step/sec: 11.7334 +INFO:tensorflow:step = 11601, loss = 0.728466, precision = 0.828125 (8.523 sec) +INFO:tensorflow:global_step/sec: 11.7588 +INFO:tensorflow:step = 11701, loss = 0.790069, precision = 0.835938 (8.504 sec) +Saved checkpoint after 30 epoch(s) to data/resnet20/checkpoints/00030... +INFO:tensorflow:global_step/sec: 11.3815 +INFO:tensorflow:step = 11801, loss = 0.868352, precision = 0.8125 (8.786 sec) +INFO:tensorflow:global_step/sec: 11.7539 +INFO:tensorflow:step = 11901, loss = 0.722305, precision = 0.851562 (8.508 sec) +INFO:tensorflow:global_step/sec: 11.722 +INFO:tensorflow:step = 12001, loss = 0.678611, precision = 0.835938 (8.531 sec) +INFO:tensorflow:global_step/sec: 11.7316 +INFO:tensorflow:step = 12101, loss = 0.644927, precision = 0.859375 (8.524 sec) +Saved checkpoint after 31 epoch(s) to data/resnet20/checkpoints/00031... +INFO:tensorflow:global_step/sec: 11.3735 +INFO:tensorflow:step = 12201, loss = 0.631129, precision = 0.875 (8.792 sec) +INFO:tensorflow:global_step/sec: 11.7347 +INFO:tensorflow:step = 12301, loss = 0.666451, precision = 0.875 (8.522 sec) +INFO:tensorflow:global_step/sec: 11.7066 +INFO:tensorflow:step = 12401, loss = 0.782566, precision = 0.835938 (8.542 sec) +INFO:tensorflow:global_step/sec: 11.7437 +INFO:tensorflow:step = 12501, loss = 0.820418, precision = 0.8125 (8.515 sec) +Saved checkpoint after 32 epoch(s) to data/resnet20/checkpoints/00032... +INFO:tensorflow:global_step/sec: 11.37 +INFO:tensorflow:step = 12601, loss = 0.750212, precision = 0.828125 (8.795 sec) +INFO:tensorflow:global_step/sec: 11.7568 +INFO:tensorflow:step = 12701, loss = 0.613324, precision = 0.867188 (8.506 sec) +INFO:tensorflow:global_step/sec: 11.753 +INFO:tensorflow:step = 12801, loss = 0.72308, precision = 0.851562 (8.509 sec) +INFO:tensorflow:global_step/sec: 11.7256 +INFO:tensorflow:step = 12901, loss = 0.936138, precision = 0.78125 (8.528 sec) +Saved checkpoint after 33 epoch(s) to data/resnet20/checkpoints/00033... +INFO:tensorflow:global_step/sec: 11.3818 +INFO:tensorflow:step = 13001, loss = 0.601315, precision = 0.898438 (8.786 sec) +INFO:tensorflow:global_step/sec: 11.7421 +INFO:tensorflow:step = 13101, loss = 0.730527, precision = 0.859375 (8.516 sec) +INFO:tensorflow:global_step/sec: 11.7642 +INFO:tensorflow:step = 13201, loss = 0.841537, precision = 0.828125 (8.500 sec) +Saved checkpoint after 34 epoch(s) to data/resnet20/checkpoints/00034... +INFO:tensorflow:global_step/sec: 11.3865 +INFO:tensorflow:step = 13301, loss = 0.659189, precision = 0.859375 (8.782 sec) +INFO:tensorflow:global_step/sec: 11.7635 +INFO:tensorflow:step = 13401, loss = 0.609686, precision = 0.875 (8.501 sec) +INFO:tensorflow:global_step/sec: 11.7199 +INFO:tensorflow:step = 13501, loss = 0.695796, precision = 0.835938 (8.532 sec) +INFO:tensorflow:global_step/sec: 11.7198 +INFO:tensorflow:step = 13601, loss = 0.592402, precision = 0.898438 (8.533 sec) +Saved checkpoint after 35 epoch(s) to data/resnet20/checkpoints/00035... +INFO:tensorflow:global_step/sec: 11.3406 +INFO:tensorflow:step = 13701, loss = 0.650642, precision = 0.898438 (8.818 sec) +INFO:tensorflow:global_step/sec: 11.7354 +INFO:tensorflow:step = 13801, loss = 0.776742, precision = 0.820312 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7503 +INFO:tensorflow:step = 13901, loss = 0.766548, precision = 0.8125 (8.510 sec) +INFO:tensorflow:global_step/sec: 11.7037 +INFO:tensorflow:step = 14001, loss = 0.61887, precision = 0.84375 (8.544 sec) +Saved checkpoint after 36 epoch(s) to data/resnet20/checkpoints/00036... +INFO:tensorflow:global_step/sec: 11.3692 +INFO:tensorflow:step = 14101, loss = 0.742694, precision = 0.859375 (8.796 sec) +INFO:tensorflow:global_step/sec: 11.7174 +INFO:tensorflow:step = 14201, loss = 0.583732, precision = 0.851562 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7233 +INFO:tensorflow:step = 14301, loss = 0.666401, precision = 0.867188 (8.530 sec) +INFO:tensorflow:global_step/sec: 11.7408 +INFO:tensorflow:step = 14401, loss = 0.599256, precision = 0.882812 (8.517 sec) +Saved checkpoint after 37 epoch(s) to data/resnet20/checkpoints/00037... +INFO:tensorflow:global_step/sec: 11.3387 +INFO:tensorflow:step = 14501, loss = 0.649326, precision = 0.875 (8.819 sec) +INFO:tensorflow:global_step/sec: 11.7081 +INFO:tensorflow:step = 14601, loss = 0.846672, precision = 0.789062 (8.541 sec) +INFO:tensorflow:global_step/sec: 11.722 +INFO:tensorflow:step = 14701, loss = 0.74026, precision = 0.84375 (8.531 sec) +INFO:tensorflow:global_step/sec: 11.7494 +INFO:tensorflow:step = 14801, loss = 0.557708, precision = 0.898438 (8.511 sec) +Saved checkpoint after 38 epoch(s) to data/resnet20/checkpoints/00038... +INFO:tensorflow:global_step/sec: 11.3501 +INFO:tensorflow:step = 14901, loss = 0.658411, precision = 0.875 (8.811 sec) +INFO:tensorflow:global_step/sec: 11.7095 +INFO:tensorflow:step = 15001, loss = 0.740123, precision = 0.820312 (8.540 sec) +INFO:tensorflow:global_step/sec: 11.7447 +INFO:tensorflow:step = 15101, loss = 0.806236, precision = 0.828125 (8.514 sec) +INFO:tensorflow:global_step/sec: 11.7539 +INFO:tensorflow:step = 15201, loss = 0.722408, precision = 0.875 (8.508 sec) +Saved checkpoint after 39 epoch(s) to data/resnet20/checkpoints/00039... +INFO:tensorflow:global_step/sec: 11.4306 +INFO:tensorflow:step = 15301, loss = 0.752467, precision = 0.851562 (8.749 sec) +INFO:tensorflow:global_step/sec: 11.7679 +INFO:tensorflow:step = 15401, loss = 0.748417, precision = 0.828125 (8.498 sec) +INFO:tensorflow:global_step/sec: 11.81 +INFO:tensorflow:step = 15501, loss = 0.783364, precision = 0.835938 (8.467 sec) +INFO:tensorflow:global_step/sec: 11.773 +INFO:tensorflow:step = 15601, loss = 0.59894, precision = 0.898438 (8.494 sec) +Saved checkpoint after 40 epoch(s) to data/resnet20/checkpoints/00040... +INFO:tensorflow:global_step/sec: 11.3921 +INFO:tensorflow:step = 15701, loss = 0.603407, precision = 0.859375 (8.778 sec) +INFO:tensorflow:global_step/sec: 11.781 +INFO:tensorflow:step = 15801, loss = 0.697215, precision = 0.828125 (8.488 sec) +INFO:tensorflow:global_step/sec: 11.8009 +INFO:tensorflow:step = 15901, loss = 0.719477, precision = 0.859375 (8.474 sec) +INFO:tensorflow:global_step/sec: 11.7825 +INFO:tensorflow:step = 16001, loss = 0.749194, precision = 0.84375 (8.487 sec) +Saved checkpoint after 41 epoch(s) to data/resnet20/checkpoints/00041... +INFO:tensorflow:global_step/sec: 11.4178 +INFO:tensorflow:step = 16101, loss = 0.799902, precision = 0.84375 (8.758 sec) +INFO:tensorflow:global_step/sec: 11.7797 +INFO:tensorflow:step = 16201, loss = 0.669, precision = 0.84375 (8.489 sec) +INFO:tensorflow:global_step/sec: 11.7983 +INFO:tensorflow:step = 16301, loss = 0.764165, precision = 0.828125 (8.476 sec) +INFO:tensorflow:global_step/sec: 11.7716 +INFO:tensorflow:step = 16401, loss = 0.694825, precision = 0.820312 (8.495 sec) +Saved checkpoint after 42 epoch(s) to data/resnet20/checkpoints/00042... +INFO:tensorflow:global_step/sec: 11.4321 +INFO:tensorflow:step = 16501, loss = 0.713898, precision = 0.867188 (8.747 sec) +INFO:tensorflow:global_step/sec: 11.8047 +INFO:tensorflow:step = 16601, loss = 0.679987, precision = 0.828125 (8.471 sec) +INFO:tensorflow:global_step/sec: 11.7569 +INFO:tensorflow:step = 16701, loss = 0.658346, precision = 0.851562 (8.506 sec) +INFO:tensorflow:global_step/sec: 11.763 +INFO:tensorflow:step = 16801, loss = 0.654045, precision = 0.898438 (8.501 sec) +Saved checkpoint after 43 epoch(s) to data/resnet20/checkpoints/00043... +INFO:tensorflow:global_step/sec: 11.4371 +INFO:tensorflow:step = 16901, loss = 0.734812, precision = 0.8125 (8.744 sec) +INFO:tensorflow:global_step/sec: 11.8034 +INFO:tensorflow:step = 17001, loss = 0.594816, precision = 0.914062 (8.472 sec) +INFO:tensorflow:global_step/sec: 11.7896 +INFO:tensorflow:step = 17101, loss = 0.736645, precision = 0.859375 (8.482 sec) +INFO:tensorflow:global_step/sec: 11.7535 +INFO:tensorflow:step = 17201, loss = 0.498745, precision = 0.9375 (8.508 sec) +Saved checkpoint after 44 epoch(s) to data/resnet20/checkpoints/00044... +INFO:tensorflow:global_step/sec: 11.363 +INFO:tensorflow:step = 17301, loss = 0.761695, precision = 0.84375 (8.801 sec) +INFO:tensorflow:global_step/sec: 11.8111 +INFO:tensorflow:step = 17401, loss = 0.72327, precision = 0.851562 (8.466 sec) +INFO:tensorflow:global_step/sec: 11.7855 +INFO:tensorflow:step = 17501, loss = 0.733098, precision = 0.835938 (8.485 sec) +Saved checkpoint after 45 epoch(s) to data/resnet20/checkpoints/00045... +INFO:tensorflow:global_step/sec: 11.4325 +INFO:tensorflow:step = 17601, loss = 0.615371, precision = 0.882812 (8.747 sec) +INFO:tensorflow:global_step/sec: 11.797 +INFO:tensorflow:step = 17701, loss = 0.510974, precision = 0.90625 (8.477 sec) +INFO:tensorflow:global_step/sec: 11.8017 +INFO:tensorflow:step = 17801, loss = 0.8302, precision = 0.835938 (8.473 sec) +INFO:tensorflow:global_step/sec: 11.7932 +INFO:tensorflow:step = 17901, loss = 0.818529, precision = 0.8125 (8.479 sec) +Saved checkpoint after 46 epoch(s) to data/resnet20/checkpoints/00046... +INFO:tensorflow:global_step/sec: 11.3714 +INFO:tensorflow:step = 18001, loss = 0.646355, precision = 0.882812 (8.794 sec) +INFO:tensorflow:global_step/sec: 11.741 +INFO:tensorflow:step = 18101, loss = 0.663365, precision = 0.867188 (8.517 sec) +INFO:tensorflow:global_step/sec: 11.7822 +INFO:tensorflow:step = 18201, loss = 0.779272, precision = 0.828125 (8.487 sec) +INFO:tensorflow:global_step/sec: 11.7503 +INFO:tensorflow:step = 18301, loss = 0.806716, precision = 0.765625 (8.510 sec) +Saved checkpoint after 47 epoch(s) to data/resnet20/checkpoints/00047... +INFO:tensorflow:global_step/sec: 11.3876 +INFO:tensorflow:step = 18401, loss = 0.682621, precision = 0.859375 (8.782 sec) +INFO:tensorflow:global_step/sec: 11.7417 +INFO:tensorflow:step = 18501, loss = 0.70905, precision = 0.867188 (8.517 sec) +INFO:tensorflow:global_step/sec: 11.7362 +INFO:tensorflow:step = 18601, loss = 0.56185, precision = 0.890625 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7696 +INFO:tensorflow:step = 18701, loss = 0.686936, precision = 0.851562 (8.496 sec) +Saved checkpoint after 48 epoch(s) to data/resnet20/checkpoints/00048... +INFO:tensorflow:global_step/sec: 11.3529 +INFO:tensorflow:step = 18801, loss = 0.622421, precision = 0.882812 (8.809 sec) +INFO:tensorflow:global_step/sec: 11.7496 +INFO:tensorflow:step = 18901, loss = 0.721585, precision = 0.84375 (8.511 sec) +INFO:tensorflow:global_step/sec: 11.759 +INFO:tensorflow:step = 19001, loss = 0.911403, precision = 0.796875 (8.504 sec) +INFO:tensorflow:global_step/sec: 11.7474 +INFO:tensorflow:step = 19101, loss = 0.54249, precision = 0.882812 (8.513 sec) +Saved checkpoint after 49 epoch(s) to data/resnet20/checkpoints/00049... +INFO:tensorflow:global_step/sec: 11.3667 +INFO:tensorflow:step = 19201, loss = 0.656957, precision = 0.859375 (8.798 sec) +INFO:tensorflow:global_step/sec: 11.7324 +INFO:tensorflow:step = 19301, loss = 0.663862, precision = 0.851562 (8.523 sec) +INFO:tensorflow:global_step/sec: 11.7627 +INFO:tensorflow:step = 19401, loss = 0.630376, precision = 0.898438 (8.501 sec) +INFO:tensorflow:global_step/sec: 11.7358 +INFO:tensorflow:step = 19501, loss = 0.817732, precision = 0.789062 (8.521 sec) +Saved checkpoint after 50 epoch(s) to data/resnet20/checkpoints/00050... +INFO:tensorflow:global_step/sec: 11.3841 +INFO:tensorflow:step = 19601, loss = 0.643818, precision = 0.875 (8.784 sec) +INFO:tensorflow:global_step/sec: 11.7085 +INFO:tensorflow:step = 19701, loss = 0.651223, precision = 0.875 (8.541 sec) +INFO:tensorflow:global_step/sec: 11.7492 +INFO:tensorflow:step = 19801, loss = 0.813597, precision = 0.835938 (8.511 sec) +INFO:tensorflow:global_step/sec: 11.7797 +INFO:tensorflow:step = 19901, loss = 0.954683, precision = 0.75 (8.489 sec) +Saved checkpoint after 51 epoch(s) to data/resnet20/checkpoints/00051... +INFO:tensorflow:global_step/sec: 11.3709 +INFO:tensorflow:step = 20001, loss = 0.672424, precision = 0.875 (8.794 sec) +INFO:tensorflow:global_step/sec: 11.7473 +INFO:tensorflow:step = 20101, loss = 0.705952, precision = 0.859375 (8.513 sec) +INFO:tensorflow:global_step/sec: 11.7763 +INFO:tensorflow:step = 20201, loss = 0.693874, precision = 0.867188 (8.492 sec) +INFO:tensorflow:global_step/sec: 11.7388 +INFO:tensorflow:step = 20301, loss = 0.61187, precision = 0.882812 (8.519 sec) +Saved checkpoint after 52 epoch(s) to data/resnet20/checkpoints/00052... +INFO:tensorflow:global_step/sec: 11.3751 +INFO:tensorflow:step = 20401, loss = 0.708973, precision = 0.851562 (8.791 sec) +INFO:tensorflow:global_step/sec: 11.7766 +INFO:tensorflow:step = 20501, loss = 0.809586, precision = 0.8125 (8.491 sec) +INFO:tensorflow:global_step/sec: 11.735 +INFO:tensorflow:step = 20601, loss = 0.592288, precision = 0.90625 (8.522 sec) +INFO:tensorflow:global_step/sec: 11.7344 +INFO:tensorflow:step = 20701, loss = 0.636789, precision = 0.875 (8.522 sec) +Saved checkpoint after 53 epoch(s) to data/resnet20/checkpoints/00053... +INFO:tensorflow:global_step/sec: 11.3811 +INFO:tensorflow:step = 20801, loss = 0.624729, precision = 0.851562 (8.787 sec) +INFO:tensorflow:global_step/sec: 11.7518 +INFO:tensorflow:step = 20901, loss = 0.745583, precision = 0.820312 (8.509 sec) +INFO:tensorflow:global_step/sec: 11.747 +INFO:tensorflow:step = 21001, loss = 0.630527, precision = 0.875 (8.513 sec) +INFO:tensorflow:global_step/sec: 11.7444 +INFO:tensorflow:step = 21101, loss = 0.613153, precision = 0.882812 (8.515 sec) +Saved checkpoint after 54 epoch(s) to data/resnet20/checkpoints/00054... +INFO:tensorflow:global_step/sec: 11.3826 +INFO:tensorflow:step = 21201, loss = 0.736509, precision = 0.8125 (8.785 sec) +INFO:tensorflow:global_step/sec: 11.7543 +INFO:tensorflow:step = 21301, loss = 0.641803, precision = 0.859375 (8.507 sec) +INFO:tensorflow:global_step/sec: 11.773 +INFO:tensorflow:step = 21401, loss = 0.690811, precision = 0.828125 (8.494 sec) +INFO:tensorflow:global_step/sec: 11.7471 +INFO:tensorflow:step = 21501, loss = 0.60983, precision = 0.867188 (8.513 sec) +Saved checkpoint after 55 epoch(s) to data/resnet20/checkpoints/00055... +INFO:tensorflow:global_step/sec: 11.3921 +INFO:tensorflow:step = 21601, loss = 0.650531, precision = 0.882812 (8.778 sec) +INFO:tensorflow:global_step/sec: 11.7491 +INFO:tensorflow:step = 21701, loss = 0.652628, precision = 0.851562 (8.511 sec) +INFO:tensorflow:global_step/sec: 11.7663 +INFO:tensorflow:step = 21801, loss = 0.651307, precision = 0.851562 (8.499 sec) +Saved checkpoint after 56 epoch(s) to data/resnet20/checkpoints/00056... +INFO:tensorflow:global_step/sec: 11.3687 +INFO:tensorflow:step = 21901, loss = 0.718917, precision = 0.84375 (8.796 sec) +INFO:tensorflow:global_step/sec: 11.7567 +INFO:tensorflow:step = 22001, loss = 0.644002, precision = 0.84375 (8.506 sec) +INFO:tensorflow:global_step/sec: 11.741 +INFO:tensorflow:step = 22101, loss = 0.842601, precision = 0.8125 (8.517 sec) +INFO:tensorflow:global_step/sec: 11.7691 +INFO:tensorflow:step = 22201, loss = 0.692468, precision = 0.851562 (8.497 sec) +Saved checkpoint after 57 epoch(s) to data/resnet20/checkpoints/00057... +INFO:tensorflow:global_step/sec: 11.3917 +INFO:tensorflow:step = 22301, loss = 0.64956, precision = 0.84375 (8.778 sec) +INFO:tensorflow:global_step/sec: 11.7677 +INFO:tensorflow:step = 22401, loss = 0.653789, precision = 0.890625 (8.498 sec) +INFO:tensorflow:global_step/sec: 11.7246 +INFO:tensorflow:step = 22501, loss = 0.664125, precision = 0.820312 (8.529 sec) +INFO:tensorflow:global_step/sec: 11.721 +INFO:tensorflow:step = 22601, loss = 0.7519, precision = 0.8125 (8.532 sec) +Saved checkpoint after 58 epoch(s) to data/resnet20/checkpoints/00058... +INFO:tensorflow:global_step/sec: 11.2907 +INFO:tensorflow:step = 22701, loss = 0.534679, precision = 0.914062 (8.857 sec) +INFO:tensorflow:global_step/sec: 11.7474 +INFO:tensorflow:step = 22801, loss = 0.675578, precision = 0.84375 (8.512 sec) +INFO:tensorflow:global_step/sec: 11.7418 +INFO:tensorflow:step = 22901, loss = 0.613244, precision = 0.867188 (8.517 sec) +INFO:tensorflow:global_step/sec: 11.7212 +INFO:tensorflow:step = 23001, loss = 0.672935, precision = 0.859375 (8.532 sec) +Saved checkpoint after 59 epoch(s) to data/resnet20/checkpoints/00059... +INFO:tensorflow:global_step/sec: 11.3655 +INFO:tensorflow:step = 23101, loss = 0.72545, precision = 0.84375 (8.799 sec) +INFO:tensorflow:global_step/sec: 11.7339 +INFO:tensorflow:step = 23201, loss = 0.838076, precision = 0.804688 (8.522 sec) +INFO:tensorflow:global_step/sec: 11.718 +INFO:tensorflow:step = 23301, loss = 0.653233, precision = 0.90625 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7065 +INFO:tensorflow:step = 23401, loss = 0.642315, precision = 0.875 (8.542 sec) +Saved checkpoint after 60 epoch(s) to data/resnet20/checkpoints/00060... +INFO:tensorflow:global_step/sec: 11.3817 +INFO:tensorflow:step = 23501, loss = 0.669002, precision = 0.875 (8.786 sec) +INFO:tensorflow:global_step/sec: 11.724 +INFO:tensorflow:step = 23601, loss = 0.889696, precision = 0.804688 (8.529 sec) +INFO:tensorflow:global_step/sec: 11.7504 +INFO:tensorflow:step = 23701, loss = 0.711124, precision = 0.84375 (8.510 sec) +INFO:tensorflow:global_step/sec: 11.7194 +INFO:tensorflow:step = 23801, loss = 0.642775, precision = 0.875 (8.533 sec) +Saved checkpoint after 61 epoch(s) to data/resnet20/checkpoints/00061... +INFO:tensorflow:global_step/sec: 11.3617 +INFO:tensorflow:step = 23901, loss = 0.631487, precision = 0.875 (8.802 sec) +INFO:tensorflow:global_step/sec: 11.7703 +INFO:tensorflow:step = 24001, loss = 0.580223, precision = 0.867188 (8.496 sec) +INFO:tensorflow:global_step/sec: 11.768 +INFO:tensorflow:step = 24101, loss = 0.693966, precision = 0.859375 (8.498 sec) +INFO:tensorflow:global_step/sec: 11.7671 +INFO:tensorflow:step = 24201, loss = 0.623286, precision = 0.882812 (8.498 sec) +Saved checkpoint after 62 epoch(s) to data/resnet20/checkpoints/00062... +INFO:tensorflow:global_step/sec: 11.3949 +INFO:tensorflow:step = 24301, loss = 0.625237, precision = 0.890625 (8.776 sec) +INFO:tensorflow:global_step/sec: 11.7648 +INFO:tensorflow:step = 24401, loss = 0.50348, precision = 0.898438 (8.500 sec) +INFO:tensorflow:global_step/sec: 11.7557 +INFO:tensorflow:step = 24501, loss = 0.542066, precision = 0.890625 (8.507 sec) +INFO:tensorflow:global_step/sec: 11.7564 +INFO:tensorflow:step = 24601, loss = 0.751176, precision = 0.820312 (8.506 sec) +Saved checkpoint after 63 epoch(s) to data/resnet20/checkpoints/00063... +INFO:tensorflow:global_step/sec: 11.3698 +INFO:tensorflow:step = 24701, loss = 0.674776, precision = 0.867188 (8.795 sec) +INFO:tensorflow:global_step/sec: 11.7494 +INFO:tensorflow:step = 24801, loss = 0.587697, precision = 0.875 (8.511 sec) +INFO:tensorflow:global_step/sec: 11.7579 +INFO:tensorflow:step = 24901, loss = 0.497716, precision = 0.929688 (8.505 sec) +INFO:tensorflow:global_step/sec: 11.7407 +INFO:tensorflow:step = 25001, loss = 0.806323, precision = 0.820312 (8.517 sec) +Saved checkpoint after 64 epoch(s) to data/resnet20/checkpoints/00064... +INFO:tensorflow:global_step/sec: 11.3617 +INFO:tensorflow:step = 25101, loss = 0.581504, precision = 0.898438 (8.802 sec) +INFO:tensorflow:global_step/sec: 11.7358 +INFO:tensorflow:step = 25201, loss = 0.615544, precision = 0.898438 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7299 +INFO:tensorflow:step = 25301, loss = 0.541206, precision = 0.890625 (8.525 sec) +INFO:tensorflow:global_step/sec: 11.7487 +INFO:tensorflow:step = 25401, loss = 0.663572, precision = 0.859375 (8.512 sec) +Saved checkpoint after 65 epoch(s) to data/resnet20/checkpoints/00065... +INFO:tensorflow:global_step/sec: 11.4102 +INFO:tensorflow:step = 25501, loss = 0.692837, precision = 0.828125 (8.764 sec) +INFO:tensorflow:global_step/sec: 11.7411 +INFO:tensorflow:step = 25601, loss = 0.634725, precision = 0.90625 (8.517 sec) +INFO:tensorflow:global_step/sec: 11.7416 +INFO:tensorflow:step = 25701, loss = 0.650736, precision = 0.859375 (8.517 sec) +INFO:tensorflow:global_step/sec: 11.7565 +INFO:tensorflow:step = 25801, loss = 0.718005, precision = 0.859375 (8.506 sec) +Saved checkpoint after 66 epoch(s) to data/resnet20/checkpoints/00066... +INFO:tensorflow:global_step/sec: 11.363 +INFO:tensorflow:step = 25901, loss = 0.684085, precision = 0.875 (8.801 sec) +INFO:tensorflow:global_step/sec: 11.7011 +INFO:tensorflow:step = 26001, loss = 0.652581, precision = 0.890625 (8.546 sec) +INFO:tensorflow:global_step/sec: 11.7182 +INFO:tensorflow:step = 26101, loss = 0.632729, precision = 0.875 (8.534 sec) +Saved checkpoint after 67 epoch(s) to data/resnet20/checkpoints/00067... +INFO:tensorflow:global_step/sec: 11.3755 +INFO:tensorflow:step = 26201, loss = 0.671332, precision = 0.867188 (8.791 sec) +INFO:tensorflow:global_step/sec: 11.725 +INFO:tensorflow:step = 26301, loss = 0.678349, precision = 0.859375 (8.529 sec) +INFO:tensorflow:global_step/sec: 11.7294 +INFO:tensorflow:step = 26401, loss = 0.593178, precision = 0.898438 (8.526 sec) +INFO:tensorflow:global_step/sec: 11.7461 +INFO:tensorflow:step = 26501, loss = 0.714126, precision = 0.851562 (8.513 sec) +Saved checkpoint after 68 epoch(s) to data/resnet20/checkpoints/00068... +INFO:tensorflow:global_step/sec: 11.3588 +INFO:tensorflow:step = 26601, loss = 0.593408, precision = 0.890625 (8.804 sec) +INFO:tensorflow:global_step/sec: 11.7226 +INFO:tensorflow:step = 26701, loss = 0.616006, precision = 0.875 (8.530 sec) +INFO:tensorflow:global_step/sec: 11.7597 +INFO:tensorflow:step = 26801, loss = 0.665084, precision = 0.859375 (8.504 sec) +INFO:tensorflow:global_step/sec: 11.7562 +INFO:tensorflow:step = 26901, loss = 0.747736, precision = 0.820312 (8.506 sec) +Saved checkpoint after 69 epoch(s) to data/resnet20/checkpoints/00069... +INFO:tensorflow:global_step/sec: 11.3886 +INFO:tensorflow:step = 27001, loss = 0.655399, precision = 0.875 (8.781 sec) +INFO:tensorflow:global_step/sec: 11.7417 +INFO:tensorflow:step = 27101, loss = 0.720524, precision = 0.796875 (8.517 sec) +INFO:tensorflow:global_step/sec: 11.7256 +INFO:tensorflow:step = 27201, loss = 0.585069, precision = 0.875 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7576 +INFO:tensorflow:step = 27301, loss = 0.708719, precision = 0.835938 (8.505 sec) +Saved checkpoint after 70 epoch(s) to data/resnet20/checkpoints/00070... +INFO:tensorflow:global_step/sec: 11.3712 +INFO:tensorflow:step = 27401, loss = 0.673218, precision = 0.859375 (8.794 sec) +INFO:tensorflow:global_step/sec: 11.733 +INFO:tensorflow:step = 27501, loss = 0.617135, precision = 0.890625 (8.523 sec) +INFO:tensorflow:global_step/sec: 11.7165 +INFO:tensorflow:step = 27601, loss = 0.640994, precision = 0.867188 (8.535 sec) +INFO:tensorflow:global_step/sec: 11.7218 +INFO:tensorflow:step = 27701, loss = 0.745403, precision = 0.835938 (8.531 sec) +Saved checkpoint after 71 epoch(s) to data/resnet20/checkpoints/00071... +INFO:tensorflow:global_step/sec: 11.4148 +INFO:tensorflow:step = 27801, loss = 0.718813, precision = 0.859375 (8.761 sec) +INFO:tensorflow:global_step/sec: 11.7056 +INFO:tensorflow:step = 27901, loss = 0.732367, precision = 0.851562 (8.543 sec) +INFO:tensorflow:global_step/sec: 11.6965 +INFO:tensorflow:step = 28001, loss = 0.626121, precision = 0.898438 (8.550 sec) +INFO:tensorflow:global_step/sec: 11.7191 +INFO:tensorflow:step = 28101, loss = 0.564342, precision = 0.882812 (8.533 sec) +Saved checkpoint after 72 epoch(s) to data/resnet20/checkpoints/00072... +INFO:tensorflow:global_step/sec: 11.3871 +INFO:tensorflow:step = 28201, loss = 0.644737, precision = 0.914062 (8.782 sec) +INFO:tensorflow:global_step/sec: 11.73 +INFO:tensorflow:step = 28301, loss = 0.629661, precision = 0.898438 (8.525 sec) +INFO:tensorflow:global_step/sec: 11.7419 +INFO:tensorflow:step = 28401, loss = 0.573489, precision = 0.882812 (8.517 sec) +INFO:tensorflow:global_step/sec: 11.7198 +INFO:tensorflow:step = 28501, loss = 0.681408, precision = 0.859375 (8.533 sec) +Saved checkpoint after 73 epoch(s) to data/resnet20/checkpoints/00073... +INFO:tensorflow:global_step/sec: 11.3813 +INFO:tensorflow:step = 28601, loss = 0.723006, precision = 0.820312 (8.786 sec) +INFO:tensorflow:global_step/sec: 11.7266 +INFO:tensorflow:step = 28701, loss = 0.572249, precision = 0.882812 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7123 +INFO:tensorflow:step = 28801, loss = 0.700794, precision = 0.875 (8.538 sec) +INFO:tensorflow:global_step/sec: 11.7502 +INFO:tensorflow:step = 28901, loss = 0.573588, precision = 0.890625 (8.511 sec) +Saved checkpoint after 74 epoch(s) to data/resnet20/checkpoints/00074... +INFO:tensorflow:global_step/sec: 11.3844 +INFO:tensorflow:step = 29001, loss = 0.678141, precision = 0.851562 (8.784 sec) +INFO:tensorflow:global_step/sec: 11.7085 +INFO:tensorflow:step = 29101, loss = 0.61079, precision = 0.890625 (8.541 sec) +INFO:tensorflow:global_step/sec: 11.7263 +INFO:tensorflow:step = 29201, loss = 0.656154, precision = 0.84375 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7112 +INFO:tensorflow:step = 29301, loss = 0.541608, precision = 0.914062 (8.539 sec) +Saved checkpoint after 75 epoch(s) to data/resnet20/checkpoints/00075... +INFO:tensorflow:global_step/sec: 11.3517 +INFO:tensorflow:step = 29401, loss = 0.703256, precision = 0.851562 (8.809 sec) +INFO:tensorflow:global_step/sec: 11.695 +INFO:tensorflow:step = 29501, loss = 0.667922, precision = 0.890625 (8.551 sec) +INFO:tensorflow:global_step/sec: 11.7306 +INFO:tensorflow:step = 29601, loss = 0.620663, precision = 0.867188 (8.525 sec) +INFO:tensorflow:global_step/sec: 11.7368 +INFO:tensorflow:step = 29701, loss = 0.633371, precision = 0.875 (8.520 sec) +Saved checkpoint after 76 epoch(s) to data/resnet20/checkpoints/00076... +INFO:tensorflow:global_step/sec: 11.3611 +INFO:tensorflow:step = 29801, loss = 0.711831, precision = 0.84375 (8.802 sec) +INFO:tensorflow:global_step/sec: 11.7239 +INFO:tensorflow:step = 29901, loss = 0.664434, precision = 0.867188 (8.530 sec) +INFO:tensorflow:global_step/sec: 11.7075 +INFO:tensorflow:step = 30001, loss = 0.644476, precision = 0.859375 (8.542 sec) +INFO:tensorflow:global_step/sec: 11.7144 +INFO:tensorflow:step = 30101, loss = 0.657201, precision = 0.859375 (8.537 sec) +Saved checkpoint after 77 epoch(s) to data/resnet20/checkpoints/00077... +INFO:tensorflow:global_step/sec: 11.3143 +INFO:tensorflow:step = 30201, loss = 0.636536, precision = 0.875 (8.838 sec) +INFO:tensorflow:global_step/sec: 11.7454 +INFO:tensorflow:step = 30301, loss = 0.646299, precision = 0.875 (8.514 sec) +INFO:tensorflow:global_step/sec: 11.7249 +INFO:tensorflow:step = 30401, loss = 0.617016, precision = 0.867188 (8.529 sec) +Saved checkpoint after 78 epoch(s) to data/resnet20/checkpoints/00078... +INFO:tensorflow:global_step/sec: 11.3667 +INFO:tensorflow:step = 30501, loss = 0.604687, precision = 0.859375 (8.798 sec) +INFO:tensorflow:global_step/sec: 11.7191 +INFO:tensorflow:step = 30601, loss = 0.702591, precision = 0.859375 (8.533 sec) +INFO:tensorflow:global_step/sec: 11.7613 +INFO:tensorflow:step = 30701, loss = 0.684973, precision = 0.835938 (8.502 sec) +INFO:tensorflow:global_step/sec: 11.7721 +INFO:tensorflow:step = 30801, loss = 0.537486, precision = 0.921875 (8.495 sec) +Saved checkpoint after 79 epoch(s) to data/resnet20/checkpoints/00079... +INFO:tensorflow:global_step/sec: 11.4309 +INFO:tensorflow:step = 30901, loss = 0.665746, precision = 0.851562 (8.748 sec) +INFO:tensorflow:global_step/sec: 11.8262 +INFO:tensorflow:step = 31001, loss = 0.644188, precision = 0.859375 (8.456 sec) +INFO:tensorflow:global_step/sec: 11.8148 +INFO:tensorflow:step = 31101, loss = 0.621896, precision = 0.882812 (8.464 sec) +INFO:tensorflow:global_step/sec: 11.7905 +INFO:tensorflow:step = 31201, loss = 0.710463, precision = 0.820312 (8.481 sec) +Saved checkpoint after 80 epoch(s) to data/resnet20/checkpoints/00080... +INFO:tensorflow:global_step/sec: 11.415 +INFO:tensorflow:step = 31301, loss = 0.61317, precision = 0.890625 (8.761 sec) +INFO:tensorflow:global_step/sec: 11.8093 +INFO:tensorflow:step = 31401, loss = 0.584789, precision = 0.882812 (8.468 sec) +INFO:tensorflow:global_step/sec: 11.7945 +INFO:tensorflow:step = 31501, loss = 0.612058, precision = 0.90625 (8.479 sec) +INFO:tensorflow:global_step/sec: 11.7769 +INFO:tensorflow:step = 31601, loss = 0.690184, precision = 0.867188 (8.491 sec) +Saved checkpoint after 81 epoch(s) to data/resnet20/checkpoints/00081... +INFO:tensorflow:global_step/sec: 11.4181 +INFO:tensorflow:step = 31701, loss = 0.566949, precision = 0.890625 (8.758 sec) +INFO:tensorflow:global_step/sec: 11.7652 +INFO:tensorflow:step = 31801, loss = 0.652673, precision = 0.867188 (8.500 sec) +INFO:tensorflow:global_step/sec: 11.7897 +INFO:tensorflow:step = 31901, loss = 0.658082, precision = 0.859375 (8.482 sec) +INFO:tensorflow:global_step/sec: 11.8076 +INFO:tensorflow:step = 32001, loss = 0.623206, precision = 0.921875 (8.469 sec) +Saved checkpoint after 82 epoch(s) to data/resnet20/checkpoints/00082... +INFO:tensorflow:global_step/sec: 11.4206 +INFO:tensorflow:step = 32101, loss = 0.718051, precision = 0.859375 (8.756 sec) +INFO:tensorflow:global_step/sec: 11.7627 +INFO:tensorflow:step = 32201, loss = 0.679342, precision = 0.859375 (8.501 sec) +INFO:tensorflow:global_step/sec: 11.771 +INFO:tensorflow:step = 32301, loss = 0.664909, precision = 0.890625 (8.495 sec) +INFO:tensorflow:global_step/sec: 11.7498 +INFO:tensorflow:step = 32401, loss = 0.500235, precision = 0.9375 (8.511 sec) +Saved checkpoint after 83 epoch(s) to data/resnet20/checkpoints/00083... +INFO:tensorflow:global_step/sec: 11.4003 +INFO:tensorflow:step = 32501, loss = 0.628109, precision = 0.859375 (8.772 sec) +INFO:tensorflow:global_step/sec: 11.7641 +INFO:tensorflow:step = 32601, loss = 0.641592, precision = 0.898438 (8.500 sec) +INFO:tensorflow:global_step/sec: 11.7771 +INFO:tensorflow:step = 32701, loss = 0.645573, precision = 0.890625 (8.491 sec) +INFO:tensorflow:global_step/sec: 11.7913 +INFO:tensorflow:step = 32801, loss = 0.752238, precision = 0.828125 (8.481 sec) +Saved checkpoint after 84 epoch(s) to data/resnet20/checkpoints/00084... +INFO:tensorflow:global_step/sec: 11.3902 +INFO:tensorflow:step = 32901, loss = 0.598605, precision = 0.851562 (8.780 sec) +INFO:tensorflow:global_step/sec: 11.7715 +INFO:tensorflow:step = 33001, loss = 0.600392, precision = 0.882812 (8.495 sec) +INFO:tensorflow:global_step/sec: 11.7961 +INFO:tensorflow:step = 33101, loss = 0.532559, precision = 0.914062 (8.477 sec) +INFO:tensorflow:global_step/sec: 11.7613 +INFO:tensorflow:step = 33201, loss = 0.514052, precision = 0.921875 (8.502 sec) +Saved checkpoint after 85 epoch(s) to data/resnet20/checkpoints/00085... +INFO:tensorflow:global_step/sec: 11.4473 +INFO:tensorflow:step = 33301, loss = 0.757357, precision = 0.890625 (8.736 sec) +INFO:tensorflow:global_step/sec: 11.7832 +INFO:tensorflow:step = 33401, loss = 0.611102, precision = 0.90625 (8.487 sec) +INFO:tensorflow:global_step/sec: 11.7611 +INFO:tensorflow:step = 33501, loss = 0.584605, precision = 0.90625 (8.503 sec) +INFO:tensorflow:global_step/sec: 11.7681 +INFO:tensorflow:step = 33601, loss = 0.713821, precision = 0.835938 (8.498 sec) +Saved checkpoint after 86 epoch(s) to data/resnet20/checkpoints/00086... +INFO:tensorflow:global_step/sec: 11.4225 +INFO:tensorflow:step = 33701, loss = 0.632823, precision = 0.890625 (8.755 sec) +INFO:tensorflow:global_step/sec: 11.7783 +INFO:tensorflow:step = 33801, loss = 0.614499, precision = 0.90625 (8.490 sec) +INFO:tensorflow:global_step/sec: 11.8147 +INFO:tensorflow:step = 33901, loss = 0.784932, precision = 0.820312 (8.464 sec) +INFO:tensorflow:global_step/sec: 11.747 +INFO:tensorflow:step = 34001, loss = 0.707596, precision = 0.859375 (8.513 sec) +Saved checkpoint after 87 epoch(s) to data/resnet20/checkpoints/00087... +INFO:tensorflow:global_step/sec: 11.4236 +INFO:tensorflow:step = 34101, loss = 0.70584, precision = 0.828125 (8.754 sec) +INFO:tensorflow:global_step/sec: 11.7718 +INFO:tensorflow:step = 34201, loss = 0.803372, precision = 0.828125 (8.495 sec) +INFO:tensorflow:global_step/sec: 11.7684 +INFO:tensorflow:step = 34301, loss = 0.710663, precision = 0.875 (8.497 sec) +INFO:tensorflow:global_step/sec: 11.7777 +INFO:tensorflow:step = 34401, loss = 0.781232, precision = 0.84375 (8.491 sec) +Saved checkpoint after 88 epoch(s) to data/resnet20/checkpoints/00088... +INFO:tensorflow:global_step/sec: 11.4163 +INFO:tensorflow:step = 34501, loss = 0.663166, precision = 0.867188 (8.760 sec) +INFO:tensorflow:global_step/sec: 11.768 +INFO:tensorflow:step = 34601, loss = 0.621432, precision = 0.867188 (8.497 sec) +INFO:tensorflow:global_step/sec: 11.7717 +INFO:tensorflow:step = 34701, loss = 0.618137, precision = 0.882812 (8.495 sec) +Saved checkpoint after 89 epoch(s) to data/resnet20/checkpoints/00089... +INFO:tensorflow:global_step/sec: 11.4183 +INFO:tensorflow:step = 34801, loss = 0.501134, precision = 0.945312 (8.758 sec) +INFO:tensorflow:global_step/sec: 11.7828 +INFO:tensorflow:step = 34901, loss = 0.835004, precision = 0.828125 (8.487 sec) +INFO:tensorflow:global_step/sec: 11.7676 +INFO:tensorflow:step = 35001, loss = 0.573131, precision = 0.898438 (8.498 sec) +INFO:tensorflow:global_step/sec: 11.7966 +INFO:tensorflow:step = 35101, loss = 0.733829, precision = 0.828125 (8.477 sec) +Saved checkpoint after 90 epoch(s) to data/resnet20/checkpoints/00090... +INFO:tensorflow:global_step/sec: 11.3998 +INFO:tensorflow:step = 35201, loss = 0.587008, precision = 0.875 (8.772 sec) +INFO:tensorflow:global_step/sec: 11.7813 +INFO:tensorflow:step = 35301, loss = 0.714862, precision = 0.859375 (8.488 sec) +INFO:tensorflow:global_step/sec: 11.7754 +INFO:tensorflow:step = 35401, loss = 0.686365, precision = 0.835938 (8.492 sec) +INFO:tensorflow:global_step/sec: 11.7369 +INFO:tensorflow:step = 35501, loss = 0.728621, precision = 0.859375 (8.520 sec) +Saved checkpoint after 91 epoch(s) to data/resnet20/checkpoints/00091... +INFO:tensorflow:global_step/sec: 11.3465 +INFO:tensorflow:step = 35601, loss = 0.6946, precision = 0.867188 (8.813 sec) +INFO:tensorflow:global_step/sec: 11.7303 +INFO:tensorflow:step = 35701, loss = 0.583826, precision = 0.867188 (8.525 sec) +INFO:tensorflow:global_step/sec: 11.739 +INFO:tensorflow:step = 35801, loss = 0.565612, precision = 0.921875 (8.519 sec) +INFO:tensorflow:global_step/sec: 11.7479 +INFO:tensorflow:step = 35901, loss = 0.467111, precision = 0.9375 (8.512 sec) +Saved checkpoint after 92 epoch(s) to data/resnet20/checkpoints/00092... +INFO:tensorflow:global_step/sec: 11.3478 +INFO:tensorflow:step = 36001, loss = 0.559272, precision = 0.90625 (8.813 sec) +INFO:tensorflow:global_step/sec: 11.7232 +INFO:tensorflow:step = 36101, loss = 0.410905, precision = 0.960938 (8.530 sec) +INFO:tensorflow:global_step/sec: 11.7259 +INFO:tensorflow:step = 36201, loss = 0.483528, precision = 0.929688 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7317 +INFO:tensorflow:step = 36301, loss = 0.516215, precision = 0.921875 (8.524 sec) +Saved checkpoint after 93 epoch(s) to data/resnet20/checkpoints/00093... +INFO:tensorflow:global_step/sec: 11.3508 +INFO:tensorflow:step = 36401, loss = 0.42175, precision = 0.960938 (8.810 sec) +INFO:tensorflow:global_step/sec: 11.7035 +INFO:tensorflow:step = 36501, loss = 0.449019, precision = 0.929688 (8.544 sec) +INFO:tensorflow:global_step/sec: 11.7353 +INFO:tensorflow:step = 36601, loss = 0.485014, precision = 0.929688 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7125 +INFO:tensorflow:step = 36701, loss = 0.490125, precision = 0.914062 (8.538 sec) +Saved checkpoint after 94 epoch(s) to data/resnet20/checkpoints/00094... +INFO:tensorflow:global_step/sec: 11.3508 +INFO:tensorflow:step = 36801, loss = 0.41527, precision = 0.9375 (8.810 sec) +INFO:tensorflow:global_step/sec: 11.7232 +INFO:tensorflow:step = 36901, loss = 0.487922, precision = 0.929688 (8.530 sec) +INFO:tensorflow:global_step/sec: 11.7244 +INFO:tensorflow:step = 37001, loss = 0.369052, precision = 0.953125 (8.529 sec) +INFO:tensorflow:global_step/sec: 11.7205 +INFO:tensorflow:step = 37101, loss = 0.431151, precision = 0.9375 (8.532 sec) +Saved checkpoint after 95 epoch(s) to data/resnet20/checkpoints/00095... +INFO:tensorflow:global_step/sec: 11.3596 +INFO:tensorflow:step = 37201, loss = 0.446431, precision = 0.929688 (8.803 sec) +INFO:tensorflow:global_step/sec: 11.7353 +INFO:tensorflow:step = 37301, loss = 0.38334, precision = 0.960938 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7259 +INFO:tensorflow:step = 37401, loss = 0.392837, precision = 0.953125 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7125 +INFO:tensorflow:step = 37501, loss = 0.369679, precision = 0.953125 (8.538 sec) +Saved checkpoint after 96 epoch(s) to data/resnet20/checkpoints/00096... +INFO:tensorflow:global_step/sec: 11.3027 +INFO:tensorflow:step = 37601, loss = 0.389414, precision = 0.945312 (8.847 sec) +INFO:tensorflow:global_step/sec: 11.6996 +INFO:tensorflow:step = 37701, loss = 0.359307, precision = 0.960938 (8.547 sec) +INFO:tensorflow:global_step/sec: 11.7192 +INFO:tensorflow:step = 37801, loss = 0.366819, precision = 0.953125 (8.533 sec) +INFO:tensorflow:global_step/sec: 11.7227 +INFO:tensorflow:step = 37901, loss = 0.336138, precision = 0.96875 (8.530 sec) +Saved checkpoint after 97 epoch(s) to data/resnet20/checkpoints/00097... +INFO:tensorflow:global_step/sec: 11.376 +INFO:tensorflow:step = 38001, loss = 0.407235, precision = 0.953125 (8.791 sec) +INFO:tensorflow:global_step/sec: 11.7422 +INFO:tensorflow:step = 38101, loss = 0.429224, precision = 0.921875 (8.516 sec) +INFO:tensorflow:global_step/sec: 11.7226 +INFO:tensorflow:step = 38201, loss = 0.368784, precision = 0.953125 (8.531 sec) +INFO:tensorflow:global_step/sec: 11.729 +INFO:tensorflow:step = 38301, loss = 0.291788, precision = 0.976562 (8.526 sec) +Saved checkpoint after 98 epoch(s) to data/resnet20/checkpoints/00098... +INFO:tensorflow:global_step/sec: 11.3575 +INFO:tensorflow:step = 38401, loss = 0.347599, precision = 0.953125 (8.805 sec) +INFO:tensorflow:global_step/sec: 11.696 +INFO:tensorflow:step = 38501, loss = 0.341276, precision = 0.953125 (8.550 sec) +INFO:tensorflow:global_step/sec: 11.6926 +INFO:tensorflow:step = 38601, loss = 0.394527, precision = 0.945312 (8.552 sec) +INFO:tensorflow:global_step/sec: 11.725 +INFO:tensorflow:step = 38701, loss = 0.334309, precision = 0.945312 (8.529 sec) +Saved checkpoint after 99 epoch(s) to data/resnet20/checkpoints/00099... +INFO:tensorflow:global_step/sec: 11.3443 +INFO:tensorflow:step = 38801, loss = 0.397664, precision = 0.9375 (8.815 sec) +INFO:tensorflow:global_step/sec: 11.7285 +INFO:tensorflow:step = 38901, loss = 0.3538, precision = 0.929688 (8.526 sec) +INFO:tensorflow:global_step/sec: 11.7464 +INFO:tensorflow:step = 39001, loss = 0.472091, precision = 0.914062 (8.513 sec) +Saved checkpoint after 100 epoch(s) to data/resnet20/checkpoints/00100... +INFO:tensorflow:global_step/sec: 11.3532 +INFO:tensorflow:step = 39101, loss = 0.359779, precision = 0.96875 (8.808 sec) +INFO:tensorflow:global_step/sec: 11.6953 +INFO:tensorflow:step = 39201, loss = 0.328236, precision = 0.96875 (8.550 sec) +INFO:tensorflow:global_step/sec: 11.6827 +INFO:tensorflow:step = 39301, loss = 0.387104, precision = 0.921875 (8.560 sec) +INFO:tensorflow:global_step/sec: 11.7208 +INFO:tensorflow:step = 39401, loss = 0.349211, precision = 0.945312 (8.532 sec) +Saved checkpoint after 101 epoch(s) to data/resnet20/checkpoints/00101... +INFO:tensorflow:global_step/sec: 11.353 +INFO:tensorflow:step = 39501, loss = 0.290806, precision = 0.976562 (8.808 sec) +INFO:tensorflow:global_step/sec: 11.7188 +INFO:tensorflow:step = 39601, loss = 0.375319, precision = 0.929688 (8.533 sec) +INFO:tensorflow:global_step/sec: 11.6957 +INFO:tensorflow:step = 39701, loss = 0.436322, precision = 0.914062 (8.550 sec) +INFO:tensorflow:global_step/sec: 11.7254 +INFO:tensorflow:step = 39801, loss = 0.324318, precision = 0.976562 (8.529 sec) +Saved checkpoint after 102 epoch(s) to data/resnet20/checkpoints/00102... +INFO:tensorflow:global_step/sec: 11.316 +INFO:tensorflow:step = 39901, loss = 0.311889, precision = 0.960938 (8.837 sec) +INFO:tensorflow:global_step/sec: 11.7317 +INFO:tensorflow:step = 40001, loss = 0.312399, precision = 0.960938 (8.524 sec) +INFO:tensorflow:global_step/sec: 11.6891 +INFO:tensorflow:step = 40101, loss = 0.374433, precision = 0.929688 (8.555 sec) +INFO:tensorflow:global_step/sec: 11.7016 +INFO:tensorflow:step = 40201, loss = 0.354826, precision = 0.96875 (8.546 sec) +Saved checkpoint after 103 epoch(s) to data/resnet20/checkpoints/00103... +INFO:tensorflow:global_step/sec: 11.3561 +INFO:tensorflow:step = 40301, loss = 0.348335, precision = 0.929688 (8.806 sec) +INFO:tensorflow:global_step/sec: 11.711 +INFO:tensorflow:step = 40401, loss = 0.310208, precision = 0.960938 (8.539 sec) +INFO:tensorflow:global_step/sec: 11.7233 +INFO:tensorflow:step = 40501, loss = 0.384397, precision = 0.945312 (8.530 sec) +INFO:tensorflow:global_step/sec: 11.706 +INFO:tensorflow:step = 40601, loss = 0.301855, precision = 0.953125 (8.543 sec) +Saved checkpoint after 104 epoch(s) to data/resnet20/checkpoints/00104... +INFO:tensorflow:global_step/sec: 11.3734 +INFO:tensorflow:step = 40701, loss = 0.316286, precision = 0.945312 (8.793 sec) +INFO:tensorflow:global_step/sec: 11.6913 +INFO:tensorflow:step = 40801, loss = 0.424704, precision = 0.90625 (8.553 sec) +INFO:tensorflow:global_step/sec: 11.7212 +INFO:tensorflow:step = 40901, loss = 0.280967, precision = 0.96875 (8.532 sec) +INFO:tensorflow:global_step/sec: 11.7284 +INFO:tensorflow:step = 41001, loss = 0.412834, precision = 0.90625 (8.526 sec) +Saved checkpoint after 105 epoch(s) to data/resnet20/checkpoints/00105... +INFO:tensorflow:global_step/sec: 11.3398 +INFO:tensorflow:step = 41101, loss = 0.258783, precision = 0.984375 (8.819 sec) +INFO:tensorflow:global_step/sec: 11.6999 +INFO:tensorflow:step = 41201, loss = 0.296273, precision = 0.96875 (8.547 sec) +INFO:tensorflow:global_step/sec: 11.6868 +INFO:tensorflow:step = 41301, loss = 0.320602, precision = 0.960938 (8.557 sec) +INFO:tensorflow:global_step/sec: 11.7209 +INFO:tensorflow:step = 41401, loss = 0.311651, precision = 0.960938 (8.532 sec) +Saved checkpoint after 106 epoch(s) to data/resnet20/checkpoints/00106... +INFO:tensorflow:global_step/sec: 11.3928 +INFO:tensorflow:step = 41501, loss = 0.338739, precision = 0.945312 (8.778 sec) +INFO:tensorflow:global_step/sec: 11.7355 +INFO:tensorflow:step = 41601, loss = 0.270487, precision = 0.976562 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7516 +INFO:tensorflow:step = 41701, loss = 0.260884, precision = 0.976562 (8.509 sec) +INFO:tensorflow:global_step/sec: 11.7165 +INFO:tensorflow:step = 41801, loss = 0.32726, precision = 0.945312 (8.535 sec) +Saved checkpoint after 107 epoch(s) to data/resnet20/checkpoints/00107... +INFO:tensorflow:global_step/sec: 11.3752 +INFO:tensorflow:step = 41901, loss = 0.276497, precision = 0.953125 (8.791 sec) +INFO:tensorflow:global_step/sec: 11.6975 +INFO:tensorflow:step = 42001, loss = 0.307742, precision = 0.960938 (8.549 sec) +INFO:tensorflow:global_step/sec: 11.7134 +INFO:tensorflow:step = 42101, loss = 0.316932, precision = 0.953125 (8.537 sec) +INFO:tensorflow:global_step/sec: 11.731 +INFO:tensorflow:step = 42201, loss = 0.264994, precision = 0.96875 (8.524 sec) +Saved checkpoint after 108 epoch(s) to data/resnet20/checkpoints/00108... +INFO:tensorflow:global_step/sec: 11.363 +INFO:tensorflow:step = 42301, loss = 0.260125, precision = 0.984375 (8.801 sec) +INFO:tensorflow:global_step/sec: 11.7305 +INFO:tensorflow:step = 42401, loss = 0.302111, precision = 0.960938 (8.525 sec) +INFO:tensorflow:global_step/sec: 11.7354 +INFO:tensorflow:step = 42501, loss = 0.252859, precision = 0.96875 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7179 +INFO:tensorflow:step = 42601, loss = 0.330509, precision = 0.929688 (8.534 sec) +Saved checkpoint after 109 epoch(s) to data/resnet20/checkpoints/00109... +INFO:tensorflow:global_step/sec: 11.3451 +INFO:tensorflow:step = 42701, loss = 0.316143, precision = 0.960938 (8.814 sec) +INFO:tensorflow:global_step/sec: 11.6801 +INFO:tensorflow:step = 42801, loss = 0.26409, precision = 0.976562 (8.561 sec) +INFO:tensorflow:global_step/sec: 11.7314 +INFO:tensorflow:step = 42901, loss = 0.27969, precision = 0.96875 (8.524 sec) +INFO:tensorflow:global_step/sec: 11.7305 +INFO:tensorflow:step = 43001, loss = 0.311006, precision = 0.960938 (8.525 sec) +Saved checkpoint after 110 epoch(s) to data/resnet20/checkpoints/00110... +INFO:tensorflow:global_step/sec: 11.3534 +INFO:tensorflow:step = 43101, loss = 0.28034, precision = 0.96875 (8.808 sec) +INFO:tensorflow:global_step/sec: 11.7303 +INFO:tensorflow:step = 43201, loss = 0.321418, precision = 0.929688 (8.525 sec) +INFO:tensorflow:global_step/sec: 11.7317 +INFO:tensorflow:step = 43301, loss = 0.263075, precision = 0.976562 (8.524 sec) +Saved checkpoint after 111 epoch(s) to data/resnet20/checkpoints/00111... +INFO:tensorflow:global_step/sec: 11.3341 +INFO:tensorflow:step = 43401, loss = 0.367721, precision = 0.945312 (8.823 sec) +INFO:tensorflow:global_step/sec: 11.762 +INFO:tensorflow:step = 43501, loss = 0.263577, precision = 0.96875 (8.502 sec) +INFO:tensorflow:global_step/sec: 11.7347 +INFO:tensorflow:step = 43601, loss = 0.312059, precision = 0.945312 (8.522 sec) +INFO:tensorflow:global_step/sec: 11.7167 +INFO:tensorflow:step = 43701, loss = 0.341055, precision = 0.9375 (8.535 sec) +Saved checkpoint after 112 epoch(s) to data/resnet20/checkpoints/00112... +INFO:tensorflow:global_step/sec: 11.3767 +INFO:tensorflow:step = 43801, loss = 0.266373, precision = 0.953125 (8.790 sec) +INFO:tensorflow:global_step/sec: 11.7055 +INFO:tensorflow:step = 43901, loss = 0.225136, precision = 0.984375 (8.543 sec) +INFO:tensorflow:global_step/sec: 11.7102 +INFO:tensorflow:step = 44001, loss = 0.28005, precision = 0.96875 (8.540 sec) +INFO:tensorflow:global_step/sec: 11.7387 +INFO:tensorflow:step = 44101, loss = 0.281532, precision = 0.953125 (8.519 sec) +Saved checkpoint after 113 epoch(s) to data/resnet20/checkpoints/00113... +INFO:tensorflow:global_step/sec: 11.376 +INFO:tensorflow:step = 44201, loss = 0.253728, precision = 0.96875 (8.791 sec) +INFO:tensorflow:global_step/sec: 11.7448 +INFO:tensorflow:step = 44301, loss = 0.372716, precision = 0.90625 (8.514 sec) +INFO:tensorflow:global_step/sec: 11.7592 +INFO:tensorflow:step = 44401, loss = 0.28431, precision = 0.945312 (8.504 sec) +INFO:tensorflow:global_step/sec: 11.7225 +INFO:tensorflow:step = 44501, loss = 0.274036, precision = 0.953125 (8.530 sec) +Saved checkpoint after 114 epoch(s) to data/resnet20/checkpoints/00114... +INFO:tensorflow:global_step/sec: 11.3198 +INFO:tensorflow:step = 44601, loss = 0.262256, precision = 0.96875 (8.834 sec) +INFO:tensorflow:global_step/sec: 11.697 +INFO:tensorflow:step = 44701, loss = 0.27341, precision = 0.960938 (8.549 sec) +INFO:tensorflow:global_step/sec: 11.7222 +INFO:tensorflow:step = 44801, loss = 0.324852, precision = 0.945312 (8.531 sec) +INFO:tensorflow:global_step/sec: 11.7294 +INFO:tensorflow:step = 44901, loss = 0.227386, precision = 0.984375 (8.526 sec) +Saved checkpoint after 115 epoch(s) to data/resnet20/checkpoints/00115... +INFO:tensorflow:global_step/sec: 11.3596 +INFO:tensorflow:step = 45001, loss = 0.268061, precision = 0.960938 (8.803 sec) +INFO:tensorflow:global_step/sec: 11.7257 +INFO:tensorflow:step = 45101, loss = 0.260544, precision = 0.953125 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7173 +INFO:tensorflow:step = 45201, loss = 0.308793, precision = 0.953125 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7092 +INFO:tensorflow:step = 45301, loss = 0.241105, precision = 0.953125 (8.540 sec) +Saved checkpoint after 116 epoch(s) to data/resnet20/checkpoints/00116... +INFO:tensorflow:global_step/sec: 11.3559 +INFO:tensorflow:step = 45401, loss = 0.336211, precision = 0.9375 (8.806 sec) +INFO:tensorflow:global_step/sec: 11.7068 +INFO:tensorflow:step = 45501, loss = 0.29008, precision = 0.953125 (8.542 sec) +INFO:tensorflow:global_step/sec: 11.7337 +INFO:tensorflow:step = 45601, loss = 0.314208, precision = 0.9375 (8.522 sec) +INFO:tensorflow:global_step/sec: 11.7259 +INFO:tensorflow:step = 45701, loss = 0.263565, precision = 0.976562 (8.528 sec) +Saved checkpoint after 117 epoch(s) to data/resnet20/checkpoints/00117... +INFO:tensorflow:global_step/sec: 11.3294 +INFO:tensorflow:step = 45801, loss = 0.277256, precision = 0.96875 (8.827 sec) +INFO:tensorflow:global_step/sec: 11.7104 +INFO:tensorflow:step = 45901, loss = 0.279065, precision = 0.953125 (8.539 sec) +INFO:tensorflow:global_step/sec: 11.706 +INFO:tensorflow:step = 46001, loss = 0.323067, precision = 0.953125 (8.543 sec) +INFO:tensorflow:global_step/sec: 11.7192 +INFO:tensorflow:step = 46101, loss = 0.228363, precision = 0.976562 (8.533 sec) +Saved checkpoint after 118 epoch(s) to data/resnet20/checkpoints/00118... +INFO:tensorflow:global_step/sec: 11.3247 +INFO:tensorflow:step = 46201, loss = 0.259237, precision = 0.96875 (8.830 sec) +INFO:tensorflow:global_step/sec: 11.6976 +INFO:tensorflow:step = 46301, loss = 0.221416, precision = 0.96875 (8.549 sec) +INFO:tensorflow:global_step/sec: 11.7146 +INFO:tensorflow:step = 46401, loss = 0.197768, precision = 0.984375 (8.536 sec) +INFO:tensorflow:global_step/sec: 11.753 +INFO:tensorflow:step = 46501, loss = 0.254456, precision = 0.976562 (8.508 sec) +Saved checkpoint after 119 epoch(s) to data/resnet20/checkpoints/00119... +INFO:tensorflow:global_step/sec: 11.347 +INFO:tensorflow:step = 46601, loss = 0.185528, precision = 1.0 (8.813 sec) +INFO:tensorflow:global_step/sec: 11.7255 +INFO:tensorflow:step = 46701, loss = 0.215334, precision = 0.992188 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7194 +INFO:tensorflow:step = 46801, loss = 0.34868, precision = 0.929688 (8.533 sec) +INFO:tensorflow:global_step/sec: 11.719 +INFO:tensorflow:step = 46901, loss = 0.22661, precision = 0.984375 (8.533 sec) +Saved checkpoint after 120 epoch(s) to data/resnet20/checkpoints/00120... +INFO:tensorflow:global_step/sec: 11.3507 +INFO:tensorflow:step = 47001, loss = 0.203606, precision = 0.976562 (8.810 sec) +INFO:tensorflow:global_step/sec: 11.7339 +INFO:tensorflow:step = 47101, loss = 0.223079, precision = 0.976562 (8.522 sec) +INFO:tensorflow:global_step/sec: 11.7019 +INFO:tensorflow:step = 47201, loss = 0.348375, precision = 0.921875 (8.546 sec) +INFO:tensorflow:global_step/sec: 11.7092 +INFO:tensorflow:step = 47301, loss = 0.250247, precision = 0.960938 (8.540 sec) +Saved checkpoint after 121 epoch(s) to data/resnet20/checkpoints/00121... +INFO:tensorflow:global_step/sec: 11.3834 +INFO:tensorflow:step = 47401, loss = 0.282641, precision = 0.960938 (8.785 sec) +INFO:tensorflow:global_step/sec: 11.7104 +INFO:tensorflow:step = 47501, loss = 0.260449, precision = 0.960938 (8.539 sec) +INFO:tensorflow:global_step/sec: 11.734 +INFO:tensorflow:step = 47601, loss = 0.238708, precision = 0.96875 (8.522 sec) +INFO:tensorflow:global_step/sec: 11.771 +INFO:tensorflow:step = 47701, loss = 0.231121, precision = 0.96875 (8.495 sec) +Saved checkpoint after 122 epoch(s) to data/resnet20/checkpoints/00122... +INFO:tensorflow:global_step/sec: 11.4077 +INFO:tensorflow:step = 47801, loss = 0.207762, precision = 0.992188 (8.766 sec) +INFO:tensorflow:global_step/sec: 11.7439 +INFO:tensorflow:step = 47901, loss = 0.263494, precision = 0.960938 (8.515 sec) +INFO:tensorflow:global_step/sec: 11.7643 +INFO:tensorflow:step = 48001, loss = 0.293659, precision = 0.9375 (8.500 sec) +Saved checkpoint after 123 epoch(s) to data/resnet20/checkpoints/00123... +INFO:tensorflow:global_step/sec: 11.4199 +INFO:tensorflow:step = 48101, loss = 0.295899, precision = 0.945312 (8.757 sec) +INFO:tensorflow:global_step/sec: 11.7912 +INFO:tensorflow:step = 48201, loss = 0.254308, precision = 0.953125 (8.481 sec) +INFO:tensorflow:global_step/sec: 11.7525 +INFO:tensorflow:step = 48301, loss = 0.240668, precision = 0.976562 (8.509 sec) +INFO:tensorflow:global_step/sec: 11.772 +INFO:tensorflow:step = 48401, loss = 0.261077, precision = 0.960938 (8.495 sec) +Saved checkpoint after 124 epoch(s) to data/resnet20/checkpoints/00124... +INFO:tensorflow:global_step/sec: 11.4095 +INFO:tensorflow:step = 48501, loss = 0.235336, precision = 0.96875 (8.765 sec) +INFO:tensorflow:global_step/sec: 11.7843 +INFO:tensorflow:step = 48601, loss = 0.249457, precision = 0.9375 (8.486 sec) +INFO:tensorflow:global_step/sec: 11.7695 +INFO:tensorflow:step = 48701, loss = 0.296576, precision = 0.945312 (8.496 sec) +INFO:tensorflow:global_step/sec: 11.7704 +INFO:tensorflow:step = 48801, loss = 0.291966, precision = 0.921875 (8.496 sec) +Saved checkpoint after 125 epoch(s) to data/resnet20/checkpoints/00125... +INFO:tensorflow:global_step/sec: 11.4024 +INFO:tensorflow:step = 48901, loss = 0.273443, precision = 0.953125 (8.770 sec) +INFO:tensorflow:global_step/sec: 11.7682 +INFO:tensorflow:step = 49001, loss = 0.253701, precision = 0.945312 (8.497 sec) +INFO:tensorflow:global_step/sec: 11.7563 +INFO:tensorflow:step = 49101, loss = 0.244842, precision = 0.960938 (8.506 sec) +INFO:tensorflow:global_step/sec: 11.7806 +INFO:tensorflow:step = 49201, loss = 0.262684, precision = 0.945312 (8.489 sec) +Saved checkpoint after 126 epoch(s) to data/resnet20/checkpoints/00126... +INFO:tensorflow:global_step/sec: 11.4121 +INFO:tensorflow:step = 49301, loss = 0.201517, precision = 0.96875 (8.763 sec) +INFO:tensorflow:global_step/sec: 11.7671 +INFO:tensorflow:step = 49401, loss = 0.20452, precision = 0.96875 (8.498 sec) +INFO:tensorflow:global_step/sec: 11.7937 +INFO:tensorflow:step = 49501, loss = 0.246459, precision = 0.953125 (8.479 sec) +INFO:tensorflow:global_step/sec: 11.7619 +INFO:tensorflow:step = 49601, loss = 0.281348, precision = 0.953125 (8.502 sec) +Saved checkpoint after 127 epoch(s) to data/resnet20/checkpoints/00127... +INFO:tensorflow:global_step/sec: 11.42 +INFO:tensorflow:step = 49701, loss = 0.222034, precision = 0.96875 (8.757 sec) +INFO:tensorflow:global_step/sec: 11.7784 +INFO:tensorflow:step = 49801, loss = 0.295962, precision = 0.953125 (8.490 sec) +INFO:tensorflow:global_step/sec: 11.7848 +INFO:tensorflow:step = 49901, loss = 0.223525, precision = 0.953125 (8.485 sec) +INFO:tensorflow:global_step/sec: 11.7854 +INFO:tensorflow:step = 50001, loss = 0.205811, precision = 0.976562 (8.485 sec) +Saved checkpoint after 128 epoch(s) to data/resnet20/checkpoints/00128... +INFO:tensorflow:global_step/sec: 11.4268 +INFO:tensorflow:step = 50101, loss = 0.287217, precision = 0.921875 (8.752 sec) +INFO:tensorflow:global_step/sec: 11.7714 +INFO:tensorflow:step = 50201, loss = 0.262448, precision = 0.9375 (8.495 sec) +INFO:tensorflow:global_step/sec: 11.7911 +INFO:tensorflow:step = 50301, loss = 0.219577, precision = 0.976562 (8.481 sec) +INFO:tensorflow:global_step/sec: 11.7566 +INFO:tensorflow:step = 50401, loss = 0.18016, precision = 0.984375 (8.506 sec) +Saved checkpoint after 129 epoch(s) to data/resnet20/checkpoints/00129... +INFO:tensorflow:global_step/sec: 11.4242 +INFO:tensorflow:step = 50501, loss = 0.231703, precision = 0.976562 (8.753 sec) +INFO:tensorflow:global_step/sec: 11.7974 +INFO:tensorflow:step = 50601, loss = 0.248554, precision = 0.960938 (8.476 sec) +INFO:tensorflow:global_step/sec: 11.8098 +INFO:tensorflow:step = 50701, loss = 0.254936, precision = 0.953125 (8.468 sec) +INFO:tensorflow:global_step/sec: 11.7674 +INFO:tensorflow:step = 50801, loss = 0.209441, precision = 0.976562 (8.498 sec) +Saved checkpoint after 130 epoch(s) to data/resnet20/checkpoints/00130... +INFO:tensorflow:global_step/sec: 11.432 +INFO:tensorflow:step = 50901, loss = 0.236172, precision = 0.945312 (8.747 sec) +INFO:tensorflow:global_step/sec: 11.7972 +INFO:tensorflow:step = 51001, loss = 0.23897, precision = 0.984375 (8.477 sec) +INFO:tensorflow:global_step/sec: 11.7835 +INFO:tensorflow:step = 51101, loss = 0.277957, precision = 0.945312 (8.486 sec) +INFO:tensorflow:global_step/sec: 11.7209 +INFO:tensorflow:step = 51201, loss = 0.15981, precision = 1.0 (8.532 sec) +Saved checkpoint after 131 epoch(s) to data/resnet20/checkpoints/00131... +INFO:tensorflow:global_step/sec: 11.375 +INFO:tensorflow:step = 51301, loss = 0.214944, precision = 0.984375 (8.791 sec) +INFO:tensorflow:global_step/sec: 11.7015 +INFO:tensorflow:step = 51401, loss = 0.258056, precision = 0.96875 (8.546 sec) +INFO:tensorflow:global_step/sec: 11.7187 +INFO:tensorflow:step = 51501, loss = 0.25668, precision = 0.960938 (8.533 sec) +INFO:tensorflow:global_step/sec: 11.7344 +INFO:tensorflow:step = 51601, loss = 0.27316, precision = 0.953125 (8.522 sec) +Saved checkpoint after 132 epoch(s) to data/resnet20/checkpoints/00132... +INFO:tensorflow:global_step/sec: 11.378 +INFO:tensorflow:step = 51701, loss = 0.256908, precision = 0.96875 (8.789 sec) +INFO:tensorflow:global_step/sec: 11.7376 +INFO:tensorflow:step = 51801, loss = 0.209732, precision = 0.96875 (8.520 sec) +INFO:tensorflow:global_step/sec: 11.7382 +INFO:tensorflow:step = 51901, loss = 0.284373, precision = 0.945312 (8.519 sec) +INFO:tensorflow:global_step/sec: 11.7418 +INFO:tensorflow:step = 52001, loss = 0.237605, precision = 0.960938 (8.517 sec) +Saved checkpoint after 133 epoch(s) to data/resnet20/checkpoints/00133... +INFO:tensorflow:global_step/sec: 11.3801 +INFO:tensorflow:step = 52101, loss = 0.192831, precision = 0.984375 (8.787 sec) +INFO:tensorflow:global_step/sec: 11.7476 +INFO:tensorflow:step = 52201, loss = 0.316533, precision = 0.921875 (8.512 sec) +INFO:tensorflow:global_step/sec: 11.7534 +INFO:tensorflow:step = 52301, loss = 0.240637, precision = 0.953125 (8.508 sec) +Saved checkpoint after 134 epoch(s) to data/resnet20/checkpoints/00134... +INFO:tensorflow:global_step/sec: 11.2936 +INFO:tensorflow:step = 52401, loss = 0.249036, precision = 0.953125 (8.855 sec) +INFO:tensorflow:global_step/sec: 11.7048 +INFO:tensorflow:step = 52501, loss = 0.234178, precision = 0.953125 (8.544 sec) +INFO:tensorflow:global_step/sec: 11.7649 +INFO:tensorflow:step = 52601, loss = 0.348396, precision = 0.9375 (8.500 sec) +INFO:tensorflow:global_step/sec: 11.6999 +INFO:tensorflow:step = 52701, loss = 0.218946, precision = 0.992188 (8.547 sec) +Saved checkpoint after 135 epoch(s) to data/resnet20/checkpoints/00135... +INFO:tensorflow:global_step/sec: 11.396 +INFO:tensorflow:step = 52801, loss = 0.225902, precision = 0.953125 (8.775 sec) +INFO:tensorflow:global_step/sec: 11.7355 +INFO:tensorflow:step = 52901, loss = 0.195264, precision = 0.976562 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7154 +INFO:tensorflow:step = 53001, loss = 0.320138, precision = 0.929688 (8.536 sec) +INFO:tensorflow:global_step/sec: 11.703 +INFO:tensorflow:step = 53101, loss = 0.282111, precision = 0.9375 (8.545 sec) +Saved checkpoint after 136 epoch(s) to data/resnet20/checkpoints/00136... +INFO:tensorflow:global_step/sec: 11.3801 +INFO:tensorflow:step = 53201, loss = 0.228458, precision = 0.960938 (8.787 sec) +INFO:tensorflow:global_step/sec: 11.7344 +INFO:tensorflow:step = 53301, loss = 0.205084, precision = 0.976562 (8.522 sec) +INFO:tensorflow:global_step/sec: 11.7434 +INFO:tensorflow:step = 53401, loss = 0.181292, precision = 0.992188 (8.515 sec) +INFO:tensorflow:global_step/sec: 11.7603 +INFO:tensorflow:step = 53501, loss = 0.177953, precision = 0.984375 (8.503 sec) +Saved checkpoint after 137 epoch(s) to data/resnet20/checkpoints/00137... +INFO:tensorflow:global_step/sec: 11.349 +INFO:tensorflow:step = 53601, loss = 0.182064, precision = 0.992188 (8.811 sec) +INFO:tensorflow:global_step/sec: 11.7275 +INFO:tensorflow:step = 53701, loss = 0.174345, precision = 1.0 (8.527 sec) +INFO:tensorflow:global_step/sec: 11.7389 +INFO:tensorflow:step = 53801, loss = 0.235924, precision = 0.976562 (8.519 sec) +INFO:tensorflow:global_step/sec: 11.7523 +INFO:tensorflow:step = 53901, loss = 0.187082, precision = 0.992188 (8.509 sec) +Saved checkpoint after 138 epoch(s) to data/resnet20/checkpoints/00138... +INFO:tensorflow:global_step/sec: 11.3472 +INFO:tensorflow:step = 54001, loss = 0.190877, precision = 0.984375 (8.813 sec) +INFO:tensorflow:global_step/sec: 11.77 +INFO:tensorflow:step = 54101, loss = 0.238874, precision = 0.945312 (8.496 sec) +INFO:tensorflow:global_step/sec: 11.7645 +INFO:tensorflow:step = 54201, loss = 0.181074, precision = 0.992188 (8.500 sec) +INFO:tensorflow:global_step/sec: 11.7488 +INFO:tensorflow:step = 54301, loss = 0.196151, precision = 0.976562 (8.512 sec) +Saved checkpoint after 139 epoch(s) to data/resnet20/checkpoints/00139... +INFO:tensorflow:global_step/sec: 11.3843 +INFO:tensorflow:step = 54401, loss = 0.166859, precision = 0.992188 (8.784 sec) +INFO:tensorflow:global_step/sec: 11.7443 +INFO:tensorflow:step = 54501, loss = 0.194751, precision = 0.984375 (8.515 sec) +INFO:tensorflow:global_step/sec: 11.7612 +INFO:tensorflow:step = 54601, loss = 0.195982, precision = 0.984375 (8.503 sec) +INFO:tensorflow:global_step/sec: 11.7171 +INFO:tensorflow:step = 54701, loss = 0.19015, precision = 0.984375 (8.534 sec) +Saved checkpoint after 140 epoch(s) to data/resnet20/checkpoints/00140... +INFO:tensorflow:global_step/sec: 11.3805 +INFO:tensorflow:step = 54801, loss = 0.199294, precision = 0.976562 (8.787 sec) +INFO:tensorflow:global_step/sec: 11.7396 +INFO:tensorflow:step = 54901, loss = 0.160679, precision = 0.992188 (8.518 sec) +INFO:tensorflow:global_step/sec: 11.7476 +INFO:tensorflow:step = 55001, loss = 0.182093, precision = 0.984375 (8.512 sec) +INFO:tensorflow:global_step/sec: 11.762 +INFO:tensorflow:step = 55101, loss = 0.18581, precision = 0.984375 (8.502 sec) +Saved checkpoint after 141 epoch(s) to data/resnet20/checkpoints/00141... +INFO:tensorflow:global_step/sec: 11.3759 +INFO:tensorflow:step = 55201, loss = 0.167101, precision = 0.984375 (8.791 sec) +INFO:tensorflow:global_step/sec: 11.7261 +INFO:tensorflow:step = 55301, loss = 0.143966, precision = 1.0 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7258 +INFO:tensorflow:step = 55401, loss = 0.168159, precision = 0.984375 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7322 +INFO:tensorflow:step = 55501, loss = 0.172443, precision = 1.0 (8.524 sec) +Saved checkpoint after 142 epoch(s) to data/resnet20/checkpoints/00142... +INFO:tensorflow:global_step/sec: 11.3552 +INFO:tensorflow:step = 55601, loss = 0.189267, precision = 0.976562 (8.807 sec) +INFO:tensorflow:global_step/sec: 11.7175 +INFO:tensorflow:step = 55701, loss = 0.153787, precision = 1.0 (8.534 sec) +INFO:tensorflow:global_step/sec: 11.7215 +INFO:tensorflow:step = 55801, loss = 0.208708, precision = 0.960938 (8.531 sec) +INFO:tensorflow:global_step/sec: 11.7284 +INFO:tensorflow:step = 55901, loss = 0.16253, precision = 1.0 (8.526 sec) +Saved checkpoint after 143 epoch(s) to data/resnet20/checkpoints/00143... +INFO:tensorflow:global_step/sec: 11.3754 +INFO:tensorflow:step = 56001, loss = 0.172091, precision = 0.992188 (8.791 sec) +INFO:tensorflow:global_step/sec: 11.7219 +INFO:tensorflow:step = 56101, loss = 0.15343, precision = 1.0 (8.531 sec) +INFO:tensorflow:global_step/sec: 11.7322 +INFO:tensorflow:step = 56201, loss = 0.18424, precision = 0.976562 (8.524 sec) +INFO:tensorflow:global_step/sec: 11.728 +INFO:tensorflow:step = 56301, loss = 0.17396, precision = 0.984375 (8.527 sec) +Saved checkpoint after 144 epoch(s) to data/resnet20/checkpoints/00144... +INFO:tensorflow:global_step/sec: 11.3963 +INFO:tensorflow:step = 56401, loss = 0.182061, precision = 0.976562 (8.775 sec) +INFO:tensorflow:global_step/sec: 11.7294 +INFO:tensorflow:step = 56501, loss = 0.153887, precision = 1.0 (8.525 sec) +INFO:tensorflow:global_step/sec: 11.7518 +INFO:tensorflow:step = 56601, loss = 0.215572, precision = 0.976562 (8.509 sec) +Saved checkpoint after 145 epoch(s) to data/resnet20/checkpoints/00145... +INFO:tensorflow:global_step/sec: 11.3398 +INFO:tensorflow:step = 56701, loss = 0.177923, precision = 0.984375 (8.819 sec) +INFO:tensorflow:global_step/sec: 11.7575 +INFO:tensorflow:step = 56801, loss = 0.157397, precision = 0.992188 (8.505 sec) +INFO:tensorflow:global_step/sec: 11.7261 +INFO:tensorflow:step = 56901, loss = 0.172902, precision = 0.984375 (8.528 sec) +INFO:tensorflow:global_step/sec: 11.7427 +INFO:tensorflow:step = 57001, loss = 0.154847, precision = 1.0 (8.516 sec) +Saved checkpoint after 146 epoch(s) to data/resnet20/checkpoints/00146... +INFO:tensorflow:global_step/sec: 11.3382 +INFO:tensorflow:step = 57101, loss = 0.195557, precision = 0.976562 (8.820 sec) +INFO:tensorflow:global_step/sec: 11.7019 +INFO:tensorflow:step = 57201, loss = 0.194951, precision = 0.984375 (8.546 sec) +INFO:tensorflow:global_step/sec: 11.7581 +INFO:tensorflow:step = 57301, loss = 0.160796, precision = 0.992188 (8.505 sec) +INFO:tensorflow:global_step/sec: 11.7548 +INFO:tensorflow:step = 57401, loss = 0.162322, precision = 0.992188 (8.507 sec) +Saved checkpoint after 147 epoch(s) to data/resnet20/checkpoints/00147... +INFO:tensorflow:global_step/sec: 11.3398 +INFO:tensorflow:step = 57501, loss = 0.154496, precision = 0.992188 (8.819 sec) +INFO:tensorflow:global_step/sec: 11.7725 +INFO:tensorflow:step = 57601, loss = 0.160154, precision = 1.0 (8.494 sec) +INFO:tensorflow:global_step/sec: 11.7476 +INFO:tensorflow:step = 57701, loss = 0.179448, precision = 0.976562 (8.512 sec) +INFO:tensorflow:global_step/sec: 11.7561 +INFO:tensorflow:step = 57801, loss = 0.17915, precision = 0.984375 (8.506 sec) +Saved checkpoint after 148 epoch(s) to data/resnet20/checkpoints/00148... +INFO:tensorflow:global_step/sec: 11.3481 +INFO:tensorflow:step = 57901, loss = 0.18201, precision = 0.984375 (8.812 sec) +INFO:tensorflow:global_step/sec: 11.7126 +INFO:tensorflow:step = 58001, loss = 0.17547, precision = 0.992188 (8.538 sec) +INFO:tensorflow:global_step/sec: 11.7453 +INFO:tensorflow:step = 58101, loss = 0.173973, precision = 0.984375 (8.514 sec) +INFO:tensorflow:global_step/sec: 11.7268 +INFO:tensorflow:step = 58201, loss = 0.142316, precision = 0.992188 (8.528 sec) +Saved checkpoint after 149 epoch(s) to data/resnet20/checkpoints/00149... +INFO:tensorflow:global_step/sec: 11.4055 +INFO:tensorflow:step = 58301, loss = 0.167178, precision = 0.992188 (8.768 sec) +INFO:tensorflow:global_step/sec: 11.7207 +INFO:tensorflow:step = 58401, loss = 0.177025, precision = 0.992188 (8.532 sec) +INFO:tensorflow:global_step/sec: 11.7243 +INFO:tensorflow:step = 58501, loss = 0.228331, precision = 0.96875 (8.529 sec) +INFO:tensorflow:global_step/sec: 11.7398 +INFO:tensorflow:step = 58601, loss = 0.162634, precision = 0.984375 (8.518 sec) +Saved checkpoint after 150 epoch(s) to data/resnet20/checkpoints/00150... +INFO:tensorflow:global_step/sec: 11.3558 +INFO:tensorflow:step = 58701, loss = 0.185215, precision = 0.984375 (8.806 sec) +INFO:tensorflow:global_step/sec: 11.7357 +INFO:tensorflow:step = 58801, loss = 0.204545, precision = 0.984375 (8.521 sec) +INFO:tensorflow:global_step/sec: 11.7552 +INFO:tensorflow:step = 58901, loss = 0.169194, precision = 0.976562 (8.507 sec) +INFO:tensorflow:global_step/sec: 11.6848 +INFO:tensorflow:step = 59001, loss = 0.158384, precision = 0.992188 (8.558 sec) +Saved checkpoint after 151 epoch(s) to data/resnet20/checkpoints/00151... +INFO:tensorflow:global_step/sec: 11.3303 +INFO:tensorflow:step = 59101, loss = 0.164977, precision = 0.992188 (8.826 sec) +INFO:tensorflow:global_step/sec: 11.7641 +INFO:tensorflow:step = 59201, loss = 0.163896, precision = 0.992188 (8.500 sec) +INFO:tensorflow:global_step/sec: 11.7544 +INFO:tensorflow:step = 59301, loss = 0.178378, precision = 0.984375 (8.507 sec) +INFO:tensorflow:global_step/sec: 11.7306 +INFO:tensorflow:step = 59401, loss = 0.156797, precision = 0.992188 (8.525 sec) +Saved checkpoint after 152 epoch(s) to data/resnet20/checkpoints/00152... +INFO:tensorflow:global_step/sec: 11.3675 +INFO:tensorflow:step = 59501, loss = 0.191142, precision = 0.984375 (8.797 sec) +INFO:tensorflow:global_step/sec: 11.7315 +INFO:tensorflow:step = 59601, loss = 0.176069, precision = 0.992188 (8.524 sec) +INFO:tensorflow:global_step/sec: 11.7491 +INFO:tensorflow:step = 59701, loss = 0.144689, precision = 1.0 (8.511 sec) +INFO:tensorflow:global_step/sec: 11.7636 +INFO:tensorflow:step = 59801, loss = 0.157503, precision = 0.992188 (8.501 sec) +Saved checkpoint after 153 epoch(s) to data/resnet20/checkpoints/00153... +INFO:tensorflow:global_step/sec: 11.3654 +INFO:tensorflow:step = 59901, loss = 0.146779, precision = 1.0 (8.799 sec) +INFO:tensorflow:global_step/sec: 11.7531 +INFO:tensorflow:step = 60001, loss = 0.142394, precision = 1.0 (8.508 sec) +INFO:tensorflow:global_step/sec: 11.7454 +INFO:tensorflow:step = 60101, loss = 0.173243, precision = 0.984375 (8.514 sec) +INFO:tensorflow:global_step/sec: 11.7816 +INFO:tensorflow:step = 60201, loss = 0.165184, precision = 0.992188 (8.488 sec) +Saved checkpoint after 154 epoch(s) to data/resnet20/checkpoints/00154... +INFO:tensorflow:global_step/sec: 11.417 +INFO:tensorflow:step = 60301, loss = 0.143409, precision = 1.0 (8.759 sec) +INFO:tensorflow:global_step/sec: 11.7448 +INFO:tensorflow:step = 60401, loss = 0.158065, precision = 1.0 (8.514 sec) +INFO:tensorflow:global_step/sec: 11.7724 +INFO:tensorflow:step = 60501, loss = 0.149408, precision = 1.0 (8.494 sec) +INFO:tensorflow:global_step/sec: 11.7602 +INFO:tensorflow:step = 60601, loss = 0.174324, precision = 0.984375 (8.503 sec) +Saved checkpoint after 155 epoch(s) to data/resnet20/checkpoints/00155... +INFO:tensorflow:global_step/sec: 11.3685 +INFO:tensorflow:step = 60701, loss = 0.174432, precision = 0.984375 (8.796 sec) +INFO:tensorflow:global_step/sec: 11.8026 +INFO:tensorflow:step = 60801, loss = 0.153166, precision = 0.992188 (8.473 sec) +INFO:tensorflow:global_step/sec: 11.7675 +INFO:tensorflow:step = 60901, loss = 0.214362, precision = 0.976562 (8.498 sec) +Saved checkpoint after 156 epoch(s) to data/resnet20/checkpoints/00156... +INFO:tensorflow:global_step/sec: 11.4091 +INFO:tensorflow:step = 61001, loss = 0.159395, precision = 0.984375 (8.765 sec) +INFO:tensorflow:global_step/sec: 11.7953 +INFO:tensorflow:step = 61101, loss = 0.142897, precision = 1.0 (8.478 sec) +INFO:tensorflow:global_step/sec: 11.7755 +INFO:tensorflow:step = 61201, loss = 0.17391, precision = 0.984375 (8.492 sec) +INFO:tensorflow:global_step/sec: 11.7723 +INFO:tensorflow:step = 61301, loss = 0.163715, precision = 0.992188 (8.495 sec) +Saved checkpoint after 157 epoch(s) to data/resnet20/checkpoints/00157... +INFO:tensorflow:global_step/sec: 11.4076 +INFO:tensorflow:step = 61401, loss = 0.201083, precision = 0.960938 (8.766 sec) +INFO:tensorflow:global_step/sec: 11.7742 +INFO:tensorflow:step = 61501, loss = 0.141672, precision = 1.0 (8.493 sec) +INFO:tensorflow:global_step/sec: 11.7477 +INFO:tensorflow:step = 61601, loss = 0.168665, precision = 0.992188 (8.512 sec) +INFO:tensorflow:global_step/sec: 11.7757 +INFO:tensorflow:step = 61701, loss = 0.164015, precision = 0.984375 (8.492 sec) +Saved checkpoint after 158 epoch(s) to data/resnet20/checkpoints/00158... +INFO:tensorflow:global_step/sec: 11.3686 +INFO:tensorflow:step = 61801, loss = 0.143982, precision = 0.992188 (8.796 sec) +INFO:tensorflow:global_step/sec: 11.7728 +INFO:tensorflow:step = 61901, loss = 0.142362, precision = 1.0 (8.494 sec) +INFO:tensorflow:global_step/sec: 11.7857 +INFO:tensorflow:step = 62001, loss = 0.14413, precision = 1.0 (8.485 sec) +INFO:tensorflow:global_step/sec: 11.7723 +INFO:tensorflow:step = 62101, loss = 0.152303, precision = 0.992188 (8.495 sec) +Saved checkpoint after 159 epoch(s) to data/resnet20/checkpoints/00159... +INFO:tensorflow:global_step/sec: 11.3705 +INFO:tensorflow:step = 62201, loss = 0.139535, precision = 1.0 (8.795 sec) +INFO:tensorflow:global_step/sec: 11.7639 +INFO:tensorflow:step = 62301, loss = 0.173239, precision = 0.984375 (8.501 sec) +INFO:tensorflow:global_step/sec: 11.7807 +INFO:tensorflow:step = 62401, loss = 0.145492, precision = 0.992188 (8.488 sec) +INFO:tensorflow:global_step/sec: 11.7864 +INFO:tensorflow:step = 62501, loss = 0.138233, precision = 1.0 (8.484 sec) +Saved checkpoint after 160 epoch(s) to data/resnet20/checkpoints/00160... +INFO:tensorflow:global_step/sec: 11.4036 +INFO:tensorflow:step = 62601, loss = 0.150536, precision = 1.0 (8.769 sec) +INFO:tensorflow:global_step/sec: 11.7561 +INFO:tensorflow:step = 62701, loss = 0.140962, precision = 1.0 (8.506 sec) +INFO:tensorflow:global_step/sec: 11.8051 +INFO:tensorflow:step = 62801, loss = 0.161211, precision = 0.992188 (8.471 sec) +INFO:tensorflow:global_step/sec: 11.7775 +INFO:tensorflow:step = 62901, loss = 0.147847, precision = 1.0 (8.491 sec) +Saved checkpoint after 161 epoch(s) to data/resnet20/checkpoints/00161... +INFO:tensorflow:global_step/sec: 11.3991 +INFO:tensorflow:step = 63001, loss = 0.13165, precision = 1.0 (8.773 sec) +INFO:tensorflow:global_step/sec: 11.7703 +INFO:tensorflow:step = 63101, loss = 0.143196, precision = 1.0 (8.496 sec) +INFO:tensorflow:global_step/sec: 11.7311 +INFO:tensorflow:step = 63201, loss = 0.179342, precision = 0.976562 (8.524 sec) +INFO:tensorflow:global_step/sec: 11.7581 +INFO:tensorflow:step = 63301, loss = 0.143759, precision = 0.992188 (8.505 sec) +Saved checkpoint after 162 epoch(s) to data/resnet20/checkpoints/00162... +INFO:tensorflow:global_step/sec: 11.4346 +INFO:tensorflow:step = 63401, loss = 0.158606, precision = 0.984375 (8.745 sec) +INFO:tensorflow:global_step/sec: 11.7821 +INFO:tensorflow:step = 63501, loss = 0.149791, precision = 1.0 (8.487 sec) +INFO:tensorflow:global_step/sec: 11.7537 +INFO:tensorflow:step = 63601, loss = 0.140088, precision = 0.992188 (8.508 sec) +INFO:tensorflow:global_step/sec: 11.757 +INFO:tensorflow:step = 63701, loss = 0.160562, precision = 0.984375 (8.506 sec) +Saved checkpoint after 163 epoch(s) to data/resnet20/checkpoints/00163... +INFO:tensorflow:global_step/sec: 11.4108 +INFO:tensorflow:step = 63801, loss = 0.14661, precision = 1.0 (8.764 sec) +INFO:tensorflow:global_step/sec: 11.7954 +INFO:tensorflow:step = 63901, loss = 0.148125, precision = 1.0 (8.478 sec) +INFO:tensorflow:global_step/sec: 11.8111 +INFO:tensorflow:step = 64001, loss = 0.156999, precision = 0.992188 (8.467 sec) +INFO:tensorflow:global_step/sec: 11.7734 +INFO:tensorflow:step = 64101, loss = 0.13902, precision = 1.0 (8.494 sec) +Saved checkpoint after 164 epoch(s) to data/resnet20/checkpoints/00164... +INFO:tensorflow:global_step/sec: 11.398 +INFO:tensorflow:step = 64201, loss = 0.134961, precision = 1.0 (8.774 sec) +INFO:tensorflow:global_step/sec: 11.7778 +INFO:tensorflow:step = 64301, loss = 0.157034, precision = 0.992188 (8.491 sec) +INFO:tensorflow:global_step/sec: 11.7857 +INFO:tensorflow:step = 64401, loss = 0.15347, precision = 0.984375 (8.485 sec) +INFO:tensorflow:global_step/sec: 11.776 +INFO:tensorflow:step = 64501, loss = 0.164666, precision = 0.992188 (8.492 sec) +Saved checkpoint after 165 epoch(s) to data/resnet20/checkpoints/00165... +INFO:tensorflow:global_step/sec: 11.427 +INFO:tensorflow:step = 64601, loss = 0.149636, precision = 0.992188 (8.751 sec) +INFO:tensorflow:global_step/sec: 11.7805 +INFO:tensorflow:step = 64701, loss = 0.160546, precision = 0.992188 (8.489 sec) +INFO:tensorflow:global_step/sec: 11.7751 +INFO:tensorflow:step = 64801, loss = 0.158733, precision = 0.992188 (8.492 sec) +INFO:tensorflow:global_step/sec: 11.7869 +INFO:tensorflow:step = 64901, loss = 0.142774, precision = 0.992188 (8.484 sec) +Saved checkpoint after 166 epoch(s) to data/resnet20/checkpoints/00166... +INFO:tensorflow:global_step/sec: 11.4174 +INFO:tensorflow:step = 65001, loss = 0.157286, precision = 0.984375 (8.759 sec) +INFO:tensorflow:global_step/sec: 11.7779 +INFO:tensorflow:step = 65101, loss = 0.181134, precision = 0.992188 (8.490 sec) +INFO:tensorflow:global_step/sec: 11.7717 +INFO:tensorflow:step = 65201, loss = 0.142549, precision = 0.992188 (8.495 sec) +Saved checkpoint after 167 epoch(s) to data/resnet20/checkpoints/00167... +INFO:tensorflow:global_step/sec: 11.4072 +INFO:tensorflow:step = 65301, loss = 0.137823, precision = 1.0 (8.766 sec) +INFO:tensorflow:global_step/sec: 11.793 +INFO:tensorflow:step = 65401, loss = 0.140586, precision = 1.0 (8.479 sec) +INFO:tensorflow:global_step/sec: 11.7826 +INFO:tensorflow:step = 65501, loss = 0.136549, precision = 0.992188 (8.487 sec) +INFO:tensorflow:global_step/sec: 11.749 +INFO:tensorflow:step = 65601, loss = 0.1571, precision = 0.992188 (8.511 sec) +Saved checkpoint after 168 epoch(s) to data/resnet20/checkpoints/00168... +INFO:tensorflow:global_step/sec: 11.4174 +INFO:tensorflow:step = 65701, loss = 0.151404, precision = 1.0 (8.759 sec) +INFO:tensorflow:global_step/sec: 11.7829 +INFO:tensorflow:step = 65801, loss = 0.137183, precision = 1.0 (8.487 sec) +INFO:tensorflow:global_step/sec: 11.7649 +INFO:tensorflow:step = 65901, loss = 0.147244, precision = 0.992188 (8.500 sec) +INFO:tensorflow:global_step/sec: 11.7877 +INFO:tensorflow:step = 66001, loss = 0.143572, precision = 0.992188 (8.485 sec) +Saved checkpoint after 169 epoch(s) to data/resnet20/checkpoints/00169... +INFO:tensorflow:global_step/sec: 11.4234 +INFO:tensorflow:step = 66101, loss = 0.133476, precision = 1.0 (8.753 sec) +INFO:tensorflow:global_step/sec: 11.7926 +INFO:tensorflow:step = 66201, loss = 0.17417, precision = 0.984375 (8.480 sec) +INFO:tensorflow:global_step/sec: 11.7604 +INFO:tensorflow:step = 66301, loss = 0.163416, precision = 0.976562 (8.503 sec) +INFO:tensorflow:global_step/sec: 11.796 +INFO:tensorflow:step = 66401, loss = 0.153192, precision = 0.992188 (8.477 sec) +Saved checkpoint after 170 epoch(s) to data/resnet20/checkpoints/00170... +INFO:tensorflow:global_step/sec: 11.4123 +INFO:tensorflow:step = 66501, loss = 0.136582, precision = 1.0 (8.763 sec) +INFO:tensorflow:global_step/sec: 11.7865 +INFO:tensorflow:step = 66601, loss = 0.145612, precision = 0.992188 (8.484 sec) +INFO:tensorflow:global_step/sec: 11.7689 +INFO:tensorflow:step = 66701, loss = 0.130615, precision = 1.0 (8.497 sec) +INFO:tensorflow:global_step/sec: 11.7868 +INFO:tensorflow:step = 66801, loss = 0.156203, precision = 0.984375 (8.484 sec) +Saved checkpoint after 171 epoch(s) to data/resnet20/checkpoints/00171... +INFO:tensorflow:global_step/sec: 11.3557 +INFO:tensorflow:step = 66901, loss = 0.137771, precision = 0.992188 (8.806 sec) +INFO:tensorflow:global_step/sec: 11.8152 +INFO:tensorflow:step = 67001, loss = 0.143168, precision = 0.992188 (8.464 sec) +INFO:tensorflow:global_step/sec: 11.7888 +INFO:tensorflow:step = 67101, loss = 0.15764, precision = 1.0 (8.483 sec) +INFO:tensorflow:global_step/sec: 11.7922 +INFO:tensorflow:step = 67201, loss = 0.155264, precision = 0.984375 (8.480 sec) +Saved checkpoint after 172 epoch(s) to data/resnet20/checkpoints/00172... +INFO:tensorflow:global_step/sec: 11.4118 +INFO:tensorflow:step = 67301, loss = 0.139652, precision = 1.0 (8.763 sec) +INFO:tensorflow:global_step/sec: 11.7957 +INFO:tensorflow:step = 67401, loss = 0.132751, precision = 1.0 (8.478 sec) +INFO:tensorflow:global_step/sec: 11.7697 +INFO:tensorflow:step = 67501, loss = 0.130827, precision = 1.0 (8.496 sec) +INFO:tensorflow:global_step/sec: 11.7734 +INFO:tensorflow:step = 67601, loss = 0.132972, precision = 1.0 (8.494 sec) +Saved checkpoint after 173 epoch(s) to data/resnet20/checkpoints/00173... +INFO:tensorflow:global_step/sec: 11.4011 +INFO:tensorflow:step = 67701, loss = 0.145731, precision = 0.992188 (8.771 sec) +INFO:tensorflow:global_step/sec: 11.8118 +INFO:tensorflow:step = 67801, loss = 0.151751, precision = 1.0 (8.466 sec) +INFO:tensorflow:global_step/sec: 11.7935 +INFO:tensorflow:step = 67901, loss = 0.16692, precision = 0.984375 (8.479 sec) +INFO:tensorflow:global_step/sec: 11.788 +INFO:tensorflow:step = 68001, loss = 0.140024, precision = 1.0 (8.483 sec) +Saved checkpoint after 174 epoch(s) to data/resnet20/checkpoints/00174... +INFO:tensorflow:global_step/sec: 11.428 +INFO:tensorflow:step = 68101, loss = 0.134724, precision = 1.0 (8.750 sec) +INFO:tensorflow:global_step/sec: 11.7892 +INFO:tensorflow:step = 68201, loss = 0.140175, precision = 0.992188 (8.482 sec) +INFO:tensorflow:global_step/sec: 11.798 +INFO:tensorflow:step = 68301, loss = 0.152174, precision = 0.992188 (8.476 sec) +INFO:tensorflow:global_step/sec: 11.7944 +INFO:tensorflow:step = 68401, loss = 0.132268, precision = 0.992188 (8.479 sec) +Saved checkpoint after 175 epoch(s) to data/resnet20/checkpoints/00175... +INFO:tensorflow:global_step/sec: 11.4336 +INFO:tensorflow:step = 68501, loss = 0.141294, precision = 0.992188 (8.746 sec) +INFO:tensorflow:global_step/sec: 11.779 +INFO:tensorflow:step = 68601, loss = 0.19421, precision = 0.976562 (8.490 sec) +INFO:tensorflow:global_step/sec: 11.7542 +INFO:tensorflow:step = 68701, loss = 0.156549, precision = 0.992188 (8.508 sec) +INFO:tensorflow:global_step/sec: 11.7845 +INFO:tensorflow:step = 68801, loss = 0.148589, precision = 0.992188 (8.486 sec) +Saved checkpoint after 176 epoch(s) to data/resnet20/checkpoints/00176... +INFO:tensorflow:global_step/sec: 11.3967 +INFO:tensorflow:step = 68901, loss = 0.135187, precision = 1.0 (8.774 sec) +INFO:tensorflow:global_step/sec: 11.7704 +INFO:tensorflow:step = 69001, loss = 0.157692, precision = 0.992188 (8.496 sec) +INFO:tensorflow:global_step/sec: 11.7818 +INFO:tensorflow:step = 69101, loss = 0.140618, precision = 0.992188 (8.488 sec) +INFO:tensorflow:global_step/sec: 11.7613 +INFO:tensorflow:step = 69201, loss = 0.156204, precision = 0.992188 (8.502 sec) +Saved checkpoint after 177 epoch(s) to data/resnet20/checkpoints/00177... +INFO:tensorflow:global_step/sec: 11.3834 +INFO:tensorflow:step = 69301, loss = 0.140532, precision = 1.0 (8.785 sec) +INFO:tensorflow:global_step/sec: 11.798 +INFO:tensorflow:step = 69401, loss = 0.137465, precision = 0.992188 (8.476 sec) +INFO:tensorflow:global_step/sec: 11.7968 +INFO:tensorflow:step = 69501, loss = 0.142194, precision = 1.0 (8.477 sec) +Saved checkpoint after 178 epoch(s) to data/resnet20/checkpoints/00178... +INFO:tensorflow:global_step/sec: 11.4004 +INFO:tensorflow:step = 69601, loss = 0.132042, precision = 1.0 (8.772 sec) +INFO:tensorflow:global_step/sec: 11.8085 +INFO:tensorflow:step = 69701, loss = 0.148674, precision = 1.0 (8.468 sec) +INFO:tensorflow:global_step/sec: 11.8098 +INFO:tensorflow:step = 69801, loss = 0.136395, precision = 1.0 (8.468 sec) +INFO:tensorflow:global_step/sec: 11.7787 +INFO:tensorflow:step = 69901, loss = 0.121135, precision = 1.0 (8.490 sec) +Saved checkpoint after 179 epoch(s) to data/resnet20/checkpoints/00179... +INFO:tensorflow:global_step/sec: 11.3554 +INFO:tensorflow:step = 70001, loss = 0.148856, precision = 0.992188 (8.806 sec) +INFO:tensorflow:global_step/sec: 11.7595 +INFO:tensorflow:step = 70101, loss = 0.134836, precision = 1.0 (8.504 sec) +INFO:tensorflow:global_step/sec: 11.7711 +INFO:tensorflow:step = 70201, loss = 0.152757, precision = 0.992188 (8.496 sec) +INFO:tensorflow:global_step/sec: 11.7784 +INFO:tensorflow:step = 70301, loss = 0.136302, precision = 1.0 (8.490 sec) +Saved checkpoint after 180 epoch(s) to data/resnet20/checkpoints/00180... +INFO:tensorflow:global_step/sec: 11.3826 +INFO:tensorflow:step = 70401, loss = 0.165441, precision = 0.976562 (8.785 sec) +INFO:tensorflow:global_step/sec: 11.7805 +INFO:tensorflow:step = 70501, loss = 0.143318, precision = 1.0 (8.489 sec) +INFO:tensorflow:global_step/sec: 11.7695 +INFO:tensorflow:step = 70601, loss = 0.141137, precision = 0.992188 (8.497 sec) +INFO:tensorflow:global_step/sec: 11.7951 +INFO:tensorflow:step = 70701, loss = 0.144377, precision = 1.0 (8.478 sec) +Saved checkpoint after 181 epoch(s) to data/resnet20/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1k80_ec2/resnet56_train.log b/tensorflow/CIFAR10/logs/1k80_ec2/resnet56_train.log new file mode 100644 index 0000000..dc3f31f --- /dev/null +++ b/tensorflow/CIFAR10/logs/1k80_ec2/resnet56_train.log @@ -0,0 +1,1836 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/850.99k params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_3/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_4/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_5/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_6/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_7/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_8/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_3/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_4/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_5/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_6/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_7/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_8/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_3/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_4/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_5/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_6/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_7/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_8/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/32.12b flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + +======================End of Report========================== +2017-07-30 07:27:23.975413: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero +2017-07-30 07:27:23.975905: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla K80 +major: 3 minor: 7 memoryClockRate (GHz) 0.8235 +pciBusID 0000:00:1e.0 +Total memory: 11.17GiB +Free memory: 11.11GiB +2017-07-30 07:27:23.975923: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-07-30 07:27:23.975929: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-07-30 07:27:23.975944: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:00:1e.0) +2017-07-30 07:27:25.069779: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 07:27:25.069817: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 4 visible devices +2017-07-30 07:27:25.071053: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x5db3730 executing computations on platform Host. Devices: +2017-07-30 07:27:25.071068: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-07-30 07:27:25.071761: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 07:27:25.071780: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 4 visible devices +2017-07-30 07:27:25.072157: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x5429c40 executing computations on platform CUDA. Devices: +2017-07-30 07:27:25.072170: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla K80, Compute Capability 3.7 +INFO:tensorflow:step = 1, loss = 3.32622, precision = 0.164062 +INFO:tensorflow:global_step/sec: 4.42477 +INFO:tensorflow:step = 101, loss = 2.66449, precision = 0.359375 (22.601 sec) +INFO:tensorflow:global_step/sec: 4.50669 +INFO:tensorflow:step = 201, loss = 2.51271, precision = 0.429688 (22.189 sec) +INFO:tensorflow:global_step/sec: 4.48674 +INFO:tensorflow:step = 301, loss = 2.45513, precision = 0.492188 (22.288 sec) +total_params: 850986 +Saved checkpoint after 1 epoch(s) to data/resnet56/checkpoints/00001... +INFO:tensorflow:global_step/sec: 4.30091 +INFO:tensorflow:step = 401, loss = 2.7181, precision = 0.351562 (23.251 sec) +INFO:tensorflow:global_step/sec: 4.46573 +INFO:tensorflow:step = 501, loss = 2.27679, precision = 0.484375 (22.393 sec) +INFO:tensorflow:global_step/sec: 4.45313 +INFO:tensorflow:step = 601, loss = 1.92367, precision = 0.5625 (22.456 sec) +INFO:tensorflow:global_step/sec: 4.45914 +INFO:tensorflow:step = 701, loss = 1.80959, precision = 0.585938 (22.426 sec) +Saved checkpoint after 2 epoch(s) to data/resnet56/checkpoints/00002... +INFO:tensorflow:global_step/sec: 4.31751 +INFO:tensorflow:step = 801, loss = 1.68848, precision = 0.703125 (23.162 sec) +INFO:tensorflow:global_step/sec: 4.45559 +INFO:tensorflow:step = 901, loss = 1.7864, precision = 0.617188 (22.444 sec) +INFO:tensorflow:global_step/sec: 4.4527 +INFO:tensorflow:step = 1001, loss = 1.42206, precision = 0.726562 (22.458 sec) +INFO:tensorflow:global_step/sec: 4.45323 +INFO:tensorflow:step = 1101, loss = 1.36998, precision = 0.773438 (22.456 sec) +Saved checkpoint after 3 epoch(s) to data/resnet56/checkpoints/00003... +INFO:tensorflow:global_step/sec: 4.32169 +INFO:tensorflow:step = 1201, loss = 1.36128, precision = 0.726562 (23.139 sec) +INFO:tensorflow:global_step/sec: 4.46064 +INFO:tensorflow:step = 1301, loss = 1.26582, precision = 0.765625 (22.418 sec) +INFO:tensorflow:global_step/sec: 4.4551 +INFO:tensorflow:step = 1401, loss = 1.2677, precision = 0.734375 (22.446 sec) +INFO:tensorflow:global_step/sec: 4.46028 +INFO:tensorflow:step = 1501, loss = 1.168, precision = 0.773438 (22.420 sec) +Saved checkpoint after 4 epoch(s) to data/resnet56/checkpoints/00004... +INFO:tensorflow:global_step/sec: 4.32043 +INFO:tensorflow:step = 1601, loss = 1.14813, precision = 0.726562 (23.146 sec) +INFO:tensorflow:global_step/sec: 4.45377 +INFO:tensorflow:step = 1701, loss = 1.07151, precision = 0.8125 (22.453 sec) +INFO:tensorflow:global_step/sec: 4.45454 +INFO:tensorflow:step = 1801, loss = 1.10903, precision = 0.726562 (22.449 sec) +INFO:tensorflow:global_step/sec: 4.45908 +INFO:tensorflow:step = 1901, loss = 0.947002, precision = 0.796875 (22.426 sec) +Saved checkpoint after 5 epoch(s) to data/resnet56/checkpoints/00005... +INFO:tensorflow:global_step/sec: 4.31843 +INFO:tensorflow:step = 2001, loss = 0.955393, precision = 0.835938 (23.157 sec) +INFO:tensorflow:global_step/sec: 4.45866 +INFO:tensorflow:step = 2101, loss = 0.923389, precision = 0.84375 (22.428 sec) +INFO:tensorflow:global_step/sec: 4.46334 +INFO:tensorflow:step = 2201, loss = 0.987496, precision = 0.804688 (22.405 sec) +INFO:tensorflow:global_step/sec: 4.45613 +INFO:tensorflow:step = 2301, loss = 1.15975, precision = 0.726562 (22.441 sec) +Saved checkpoint after 6 epoch(s) to data/resnet56/checkpoints/00006... +INFO:tensorflow:global_step/sec: 4.32005 +INFO:tensorflow:step = 2401, loss = 0.879887, precision = 0.78125 (23.148 sec) +INFO:tensorflow:global_step/sec: 4.45803 +INFO:tensorflow:step = 2501, loss = 0.911229, precision = 0.789062 (22.431 sec) +INFO:tensorflow:global_step/sec: 4.45678 +INFO:tensorflow:step = 2601, loss = 0.69046, precision = 0.898438 (22.438 sec) +INFO:tensorflow:global_step/sec: 4.46156 +INFO:tensorflow:step = 2701, loss = 0.817805, precision = 0.78125 (22.414 sec) +Saved checkpoint after 7 epoch(s) to data/resnet56/checkpoints/00007... +INFO:tensorflow:global_step/sec: 4.31689 +INFO:tensorflow:step = 2801, loss = 1.07426, precision = 0.671875 (23.165 sec) +INFO:tensorflow:global_step/sec: 4.45627 +INFO:tensorflow:step = 2901, loss = 0.74294, precision = 0.84375 (22.440 sec) +INFO:tensorflow:global_step/sec: 4.46161 +INFO:tensorflow:step = 3001, loss = 0.868435, precision = 0.765625 (22.413 sec) +INFO:tensorflow:global_step/sec: 4.45813 +INFO:tensorflow:step = 3101, loss = 0.927438, precision = 0.804688 (22.431 sec) +Saved checkpoint after 8 epoch(s) to data/resnet56/checkpoints/00008... +INFO:tensorflow:global_step/sec: 4.30774 +INFO:tensorflow:step = 3201, loss = 0.68573, precision = 0.859375 (23.214 sec) +INFO:tensorflow:global_step/sec: 4.45963 +INFO:tensorflow:step = 3301, loss = 0.711697, precision = 0.867188 (22.423 sec) +INFO:tensorflow:global_step/sec: 4.46172 +INFO:tensorflow:step = 3401, loss = 0.710988, precision = 0.851562 (22.413 sec) +INFO:tensorflow:global_step/sec: 4.457 +INFO:tensorflow:step = 3501, loss = 0.822261, precision = 0.796875 (22.437 sec) +Saved checkpoint after 9 epoch(s) to data/resnet56/checkpoints/00009... +INFO:tensorflow:global_step/sec: 4.32022 +INFO:tensorflow:step = 3601, loss = 0.872244, precision = 0.78125 (23.147 sec) +INFO:tensorflow:global_step/sec: 4.45715 +INFO:tensorflow:step = 3701, loss = 0.723737, precision = 0.828125 (22.436 sec) +INFO:tensorflow:global_step/sec: 4.45691 +INFO:tensorflow:step = 3801, loss = 0.779144, precision = 0.835938 (22.437 sec) +INFO:tensorflow:global_step/sec: 4.4588 +INFO:tensorflow:step = 3901, loss = 0.96097, precision = 0.75 (22.428 sec) +Saved checkpoint after 10 epoch(s) to data/resnet56/checkpoints/00010... +INFO:tensorflow:global_step/sec: 4.31852 +INFO:tensorflow:step = 4001, loss = 0.699941, precision = 0.859375 (23.156 sec) +INFO:tensorflow:global_step/sec: 4.45907 +INFO:tensorflow:step = 4101, loss = 0.700696, precision = 0.8125 (22.426 sec) +INFO:tensorflow:global_step/sec: 4.45425 +INFO:tensorflow:step = 4201, loss = 0.756563, precision = 0.820312 (22.450 sec) +Saved checkpoint after 11 epoch(s) to data/resnet56/checkpoints/00011... +INFO:tensorflow:global_step/sec: 4.31916 +INFO:tensorflow:step = 4301, loss = 0.842191, precision = 0.820312 (23.153 sec) +INFO:tensorflow:global_step/sec: 4.45956 +INFO:tensorflow:step = 4401, loss = 0.72696, precision = 0.851562 (22.424 sec) +INFO:tensorflow:global_step/sec: 4.46301 +INFO:tensorflow:step = 4501, loss = 0.712493, precision = 0.875 (22.406 sec) +INFO:tensorflow:global_step/sec: 4.45681 +INFO:tensorflow:step = 4601, loss = 0.667135, precision = 0.851562 (22.438 sec) +Saved checkpoint after 12 epoch(s) to data/resnet56/checkpoints/00012... +INFO:tensorflow:global_step/sec: 4.31457 +INFO:tensorflow:step = 4701, loss = 0.629464, precision = 0.882812 (23.177 sec) +INFO:tensorflow:global_step/sec: 4.46213 +INFO:tensorflow:step = 4801, loss = 0.711391, precision = 0.828125 (22.411 sec) +INFO:tensorflow:global_step/sec: 4.45767 +INFO:tensorflow:step = 4901, loss = 0.735842, precision = 0.890625 (22.433 sec) +INFO:tensorflow:global_step/sec: 4.46081 +INFO:tensorflow:step = 5001, loss = 0.658147, precision = 0.859375 (22.417 sec) +Saved checkpoint after 13 epoch(s) to data/resnet56/checkpoints/00013... +INFO:tensorflow:global_step/sec: 4.31289 +INFO:tensorflow:step = 5101, loss = 0.737742, precision = 0.8125 (23.186 sec) +INFO:tensorflow:global_step/sec: 4.46708 +INFO:tensorflow:step = 5201, loss = 0.690968, precision = 0.882812 (22.386 sec) +INFO:tensorflow:global_step/sec: 4.46364 +INFO:tensorflow:step = 5301, loss = 0.678122, precision = 0.859375 (22.403 sec) +INFO:tensorflow:global_step/sec: 4.46291 +INFO:tensorflow:step = 5401, loss = 0.740831, precision = 0.890625 (22.407 sec) +Saved checkpoint after 14 epoch(s) to data/resnet56/checkpoints/00014... +INFO:tensorflow:global_step/sec: 4.32233 +INFO:tensorflow:step = 5501, loss = 0.644347, precision = 0.867188 (23.136 sec) +INFO:tensorflow:global_step/sec: 4.46457 +INFO:tensorflow:step = 5601, loss = 0.81173, precision = 0.773438 (22.398 sec) +INFO:tensorflow:global_step/sec: 4.46155 +INFO:tensorflow:step = 5701, loss = 0.701578, precision = 0.851562 (22.414 sec) +INFO:tensorflow:global_step/sec: 4.46739 +INFO:tensorflow:step = 5801, loss = 0.954305, precision = 0.757812 (22.384 sec) +Saved checkpoint after 15 epoch(s) to data/resnet56/checkpoints/00015... +INFO:tensorflow:global_step/sec: 4.31862 +INFO:tensorflow:step = 5901, loss = 0.737447, precision = 0.828125 (23.156 sec) +INFO:tensorflow:global_step/sec: 4.4629 +INFO:tensorflow:step = 6001, loss = 0.627578, precision = 0.859375 (22.407 sec) +INFO:tensorflow:global_step/sec: 4.45843 +INFO:tensorflow:step = 6101, loss = 0.793367, precision = 0.820312 (22.429 sec) +INFO:tensorflow:global_step/sec: 4.46372 +INFO:tensorflow:step = 6201, loss = 0.682096, precision = 0.835938 (22.403 sec) +Saved checkpoint after 16 epoch(s) to data/resnet56/checkpoints/00016... +INFO:tensorflow:global_step/sec: 4.32309 +INFO:tensorflow:step = 6301, loss = 0.721186, precision = 0.859375 (23.132 sec) +INFO:tensorflow:global_step/sec: 4.45566 +INFO:tensorflow:step = 6401, loss = 0.628113, precision = 0.890625 (22.443 sec) +INFO:tensorflow:global_step/sec: 4.46174 +INFO:tensorflow:step = 6501, loss = 0.802043, precision = 0.828125 (22.413 sec) +INFO:tensorflow:global_step/sec: 4.46217 +INFO:tensorflow:step = 6601, loss = 0.652623, precision = 0.84375 (22.411 sec) +Saved checkpoint after 17 epoch(s) to data/resnet56/checkpoints/00017... +INFO:tensorflow:global_step/sec: 4.32147 +INFO:tensorflow:step = 6701, loss = 0.912854, precision = 0.78125 (23.140 sec) +INFO:tensorflow:global_step/sec: 4.46251 +INFO:tensorflow:step = 6801, loss = 0.769722, precision = 0.820312 (22.409 sec) +INFO:tensorflow:global_step/sec: 4.46156 +INFO:tensorflow:step = 6901, loss = 0.803574, precision = 0.820312 (22.414 sec) +INFO:tensorflow:global_step/sec: 4.45998 +INFO:tensorflow:step = 7001, loss = 0.609961, precision = 0.882812 (22.422 sec) +Saved checkpoint after 18 epoch(s) to data/resnet56/checkpoints/00018... +INFO:tensorflow:global_step/sec: 4.32236 +INFO:tensorflow:step = 7101, loss = 0.656177, precision = 0.875 (23.136 sec) +INFO:tensorflow:global_step/sec: 4.46779 +INFO:tensorflow:step = 7201, loss = 0.635703, precision = 0.859375 (22.382 sec) +INFO:tensorflow:global_step/sec: 4.46543 +INFO:tensorflow:step = 7301, loss = 0.800235, precision = 0.851562 (22.394 sec) +INFO:tensorflow:global_step/sec: 4.46226 +INFO:tensorflow:step = 7401, loss = 0.62449, precision = 0.875 (22.410 sec) +Saved checkpoint after 19 epoch(s) to data/resnet56/checkpoints/00019... +INFO:tensorflow:global_step/sec: 4.32109 +INFO:tensorflow:step = 7501, loss = 0.759465, precision = 0.851562 (23.142 sec) +INFO:tensorflow:global_step/sec: 4.46537 +INFO:tensorflow:step = 7601, loss = 0.719462, precision = 0.867188 (22.394 sec) +INFO:tensorflow:global_step/sec: 4.46668 +INFO:tensorflow:step = 7701, loss = 0.647946, precision = 0.851562 (22.388 sec) +INFO:tensorflow:global_step/sec: 4.46441 +INFO:tensorflow:step = 7801, loss = 0.694389, precision = 0.875 (22.399 sec) +Saved checkpoint after 20 epoch(s) to data/resnet56/checkpoints/00020... +INFO:tensorflow:global_step/sec: 4.30831 +INFO:tensorflow:step = 7901, loss = 0.656622, precision = 0.84375 (23.211 sec) +INFO:tensorflow:global_step/sec: 4.46249 +INFO:tensorflow:step = 8001, loss = 0.780149, precision = 0.820312 (22.409 sec) +INFO:tensorflow:global_step/sec: 4.45584 +INFO:tensorflow:step = 8101, loss = 0.699352, precision = 0.867188 (22.442 sec) +INFO:tensorflow:global_step/sec: 4.46396 +INFO:tensorflow:step = 8201, loss = 0.760523, precision = 0.84375 (22.402 sec) +Saved checkpoint after 21 epoch(s) to data/resnet56/checkpoints/00021... +INFO:tensorflow:global_step/sec: 4.31869 +INFO:tensorflow:step = 8301, loss = 0.728335, precision = 0.835938 (23.155 sec) +INFO:tensorflow:global_step/sec: 4.46518 +INFO:tensorflow:step = 8401, loss = 0.703712, precision = 0.875 (22.395 sec) +INFO:tensorflow:global_step/sec: 4.45911 +INFO:tensorflow:step = 8501, loss = 0.765481, precision = 0.835938 (22.426 sec) +INFO:tensorflow:global_step/sec: 4.4629 +INFO:tensorflow:step = 8601, loss = 0.657823, precision = 0.867188 (22.407 sec) +Saved checkpoint after 22 epoch(s) to data/resnet56/checkpoints/00022... +INFO:tensorflow:global_step/sec: 4.32522 +INFO:tensorflow:step = 8701, loss = 0.70517, precision = 0.859375 (23.120 sec) +INFO:tensorflow:global_step/sec: 4.46688 +INFO:tensorflow:step = 8801, loss = 0.682392, precision = 0.851562 (22.387 sec) +INFO:tensorflow:global_step/sec: 4.45758 +INFO:tensorflow:step = 8901, loss = 0.737128, precision = 0.851562 (22.434 sec) +Saved checkpoint after 23 epoch(s) to data/resnet56/checkpoints/00023... +INFO:tensorflow:global_step/sec: 4.32781 +INFO:tensorflow:step = 9001, loss = 0.64845, precision = 0.859375 (23.106 sec) +INFO:tensorflow:global_step/sec: 4.46069 +INFO:tensorflow:step = 9101, loss = 0.687264, precision = 0.84375 (22.418 sec) +INFO:tensorflow:global_step/sec: 4.45707 +INFO:tensorflow:step = 9201, loss = 0.708714, precision = 0.851562 (22.436 sec) +INFO:tensorflow:global_step/sec: 4.46006 +INFO:tensorflow:step = 9301, loss = 0.687788, precision = 0.851562 (22.421 sec) +Saved checkpoint after 24 epoch(s) to data/resnet56/checkpoints/00024... +INFO:tensorflow:global_step/sec: 4.32483 +INFO:tensorflow:step = 9401, loss = 0.651372, precision = 0.851562 (23.122 sec) +INFO:tensorflow:global_step/sec: 4.46154 +INFO:tensorflow:step = 9501, loss = 0.625121, precision = 0.875 (22.414 sec) +INFO:tensorflow:global_step/sec: 4.45798 +INFO:tensorflow:step = 9601, loss = 0.766614, precision = 0.8125 (22.432 sec) +INFO:tensorflow:global_step/sec: 4.46121 +INFO:tensorflow:step = 9701, loss = 0.653729, precision = 0.851562 (22.415 sec) +Saved checkpoint after 25 epoch(s) to data/resnet56/checkpoints/00025... +INFO:tensorflow:global_step/sec: 4.31127 +INFO:tensorflow:step = 9801, loss = 0.700852, precision = 0.84375 (23.195 sec) +INFO:tensorflow:global_step/sec: 4.45799 +INFO:tensorflow:step = 9901, loss = 0.680787, precision = 0.867188 (22.432 sec) +INFO:tensorflow:global_step/sec: 4.45562 +INFO:tensorflow:step = 10001, loss = 0.703334, precision = 0.859375 (22.444 sec) +INFO:tensorflow:global_step/sec: 4.45359 +INFO:tensorflow:step = 10101, loss = 0.623221, precision = 0.898438 (22.454 sec) +Saved checkpoint after 26 epoch(s) to data/resnet56/checkpoints/00026... +INFO:tensorflow:global_step/sec: 4.31643 +INFO:tensorflow:step = 10201, loss = 0.716224, precision = 0.851562 (23.167 sec) +INFO:tensorflow:global_step/sec: 4.45598 +INFO:tensorflow:step = 10301, loss = 0.618133, precision = 0.898438 (22.442 sec) +INFO:tensorflow:global_step/sec: 4.45422 +INFO:tensorflow:step = 10401, loss = 0.745341, precision = 0.828125 (22.451 sec) +INFO:tensorflow:global_step/sec: 4.45835 +INFO:tensorflow:step = 10501, loss = 0.592499, precision = 0.929688 (22.430 sec) +Saved checkpoint after 27 epoch(s) to data/resnet56/checkpoints/00027... +INFO:tensorflow:global_step/sec: 4.31891 +INFO:tensorflow:step = 10601, loss = 0.599902, precision = 0.90625 (23.154 sec) +INFO:tensorflow:global_step/sec: 4.46253 +INFO:tensorflow:step = 10701, loss = 0.626845, precision = 0.890625 (22.409 sec) +INFO:tensorflow:global_step/sec: 4.45905 +INFO:tensorflow:step = 10801, loss = 0.653764, precision = 0.835938 (22.426 sec) +INFO:tensorflow:global_step/sec: 4.46206 +INFO:tensorflow:step = 10901, loss = 0.6187, precision = 0.890625 (22.411 sec) +Saved checkpoint after 28 epoch(s) to data/resnet56/checkpoints/00028... +INFO:tensorflow:global_step/sec: 4.31632 +INFO:tensorflow:step = 11001, loss = 0.72997, precision = 0.84375 (23.168 sec) +INFO:tensorflow:global_step/sec: 4.45834 +INFO:tensorflow:step = 11101, loss = 0.682012, precision = 0.882812 (22.430 sec) +INFO:tensorflow:global_step/sec: 4.45618 +INFO:tensorflow:step = 11201, loss = 0.726139, precision = 0.867188 (22.441 sec) +INFO:tensorflow:global_step/sec: 4.46199 +INFO:tensorflow:step = 11301, loss = 0.519582, precision = 0.929688 (22.411 sec) +Saved checkpoint after 29 epoch(s) to data/resnet56/checkpoints/00029... +INFO:tensorflow:global_step/sec: 4.32124 +INFO:tensorflow:step = 11401, loss = 0.596202, precision = 0.90625 (23.142 sec) +INFO:tensorflow:global_step/sec: 4.46464 +INFO:tensorflow:step = 11501, loss = 0.695123, precision = 0.859375 (22.398 sec) +INFO:tensorflow:global_step/sec: 4.46151 +INFO:tensorflow:step = 11601, loss = 0.873707, precision = 0.78125 (22.414 sec) +INFO:tensorflow:global_step/sec: 4.46369 +INFO:tensorflow:step = 11701, loss = 0.730435, precision = 0.851562 (22.403 sec) +Saved checkpoint after 30 epoch(s) to data/resnet56/checkpoints/00030... +INFO:tensorflow:global_step/sec: 4.32285 +INFO:tensorflow:step = 11801, loss = 0.652161, precision = 0.859375 (23.133 sec) +INFO:tensorflow:global_step/sec: 4.45769 +INFO:tensorflow:step = 11901, loss = 0.689683, precision = 0.851562 (22.433 sec) +INFO:tensorflow:global_step/sec: 4.45789 +INFO:tensorflow:step = 12001, loss = 0.644094, precision = 0.875 (22.432 sec) +INFO:tensorflow:global_step/sec: 4.45966 +INFO:tensorflow:step = 12101, loss = 0.627765, precision = 0.882812 (22.423 sec) +Saved checkpoint after 31 epoch(s) to data/resnet56/checkpoints/00031... +INFO:tensorflow:global_step/sec: 4.31395 +INFO:tensorflow:step = 12201, loss = 0.811678, precision = 0.804688 (23.181 sec) +INFO:tensorflow:global_step/sec: 4.45634 +INFO:tensorflow:step = 12301, loss = 0.693974, precision = 0.859375 (22.440 sec) +INFO:tensorflow:global_step/sec: 4.46389 +INFO:tensorflow:step = 12401, loss = 0.646817, precision = 0.898438 (22.402 sec) +INFO:tensorflow:global_step/sec: 4.4598 +INFO:tensorflow:step = 12501, loss = 0.767573, precision = 0.820312 (22.423 sec) +Saved checkpoint after 32 epoch(s) to data/resnet56/checkpoints/00032... +INFO:tensorflow:global_step/sec: 4.31616 +INFO:tensorflow:step = 12601, loss = 0.717321, precision = 0.851562 (23.169 sec) +INFO:tensorflow:global_step/sec: 4.45869 +INFO:tensorflow:step = 12701, loss = 0.777066, precision = 0.8125 (22.428 sec) +INFO:tensorflow:global_step/sec: 4.45907 +INFO:tensorflow:step = 12801, loss = 0.664364, precision = 0.875 (22.426 sec) +INFO:tensorflow:global_step/sec: 4.45688 +INFO:tensorflow:step = 12901, loss = 0.716344, precision = 0.84375 (22.437 sec) +Saved checkpoint after 33 epoch(s) to data/resnet56/checkpoints/00033... +INFO:tensorflow:global_step/sec: 4.30311 +INFO:tensorflow:step = 13001, loss = 0.623185, precision = 0.890625 (23.239 sec) +INFO:tensorflow:global_step/sec: 4.46043 +INFO:tensorflow:step = 13101, loss = 0.829205, precision = 0.84375 (22.419 sec) +INFO:tensorflow:global_step/sec: 4.45598 +INFO:tensorflow:step = 13201, loss = 0.618836, precision = 0.890625 (22.442 sec) +Saved checkpoint after 34 epoch(s) to data/resnet56/checkpoints/00034... +INFO:tensorflow:global_step/sec: 4.31452 +INFO:tensorflow:step = 13301, loss = 0.52459, precision = 0.921875 (23.178 sec) +INFO:tensorflow:global_step/sec: 4.45903 +INFO:tensorflow:step = 13401, loss = 0.829758, precision = 0.796875 (22.426 sec) +INFO:tensorflow:global_step/sec: 4.45895 +INFO:tensorflow:step = 13501, loss = 0.678913, precision = 0.875 (22.427 sec) +INFO:tensorflow:global_step/sec: 4.4542 +INFO:tensorflow:step = 13601, loss = 0.664495, precision = 0.890625 (22.451 sec) +Saved checkpoint after 35 epoch(s) to data/resnet56/checkpoints/00035... +INFO:tensorflow:global_step/sec: 4.31381 +INFO:tensorflow:step = 13701, loss = 0.621205, precision = 0.898438 (23.181 sec) +INFO:tensorflow:global_step/sec: 4.46076 +INFO:tensorflow:step = 13801, loss = 0.712205, precision = 0.851562 (22.418 sec) +INFO:tensorflow:global_step/sec: 4.46005 +INFO:tensorflow:step = 13901, loss = 0.526218, precision = 0.9375 (22.421 sec) +INFO:tensorflow:global_step/sec: 4.45747 +INFO:tensorflow:step = 14001, loss = 0.811407, precision = 0.84375 (22.434 sec) +Saved checkpoint after 36 epoch(s) to data/resnet56/checkpoints/00036... +INFO:tensorflow:global_step/sec: 4.31114 +INFO:tensorflow:step = 14101, loss = 0.519218, precision = 0.929688 (23.196 sec) +INFO:tensorflow:global_step/sec: 4.45569 +INFO:tensorflow:step = 14201, loss = 0.698396, precision = 0.882812 (22.443 sec) +INFO:tensorflow:global_step/sec: 4.45375 +INFO:tensorflow:step = 14301, loss = 0.619853, precision = 0.875 (22.453 sec) +INFO:tensorflow:global_step/sec: 4.45073 +INFO:tensorflow:step = 14401, loss = 1.01784, precision = 0.789062 (22.468 sec) +Saved checkpoint after 37 epoch(s) to data/resnet56/checkpoints/00037... +INFO:tensorflow:global_step/sec: 4.31035 +INFO:tensorflow:step = 14501, loss = 0.705546, precision = 0.875 (23.200 sec) +INFO:tensorflow:global_step/sec: 4.45412 +INFO:tensorflow:step = 14601, loss = 0.60928, precision = 0.898438 (22.451 sec) +INFO:tensorflow:global_step/sec: 4.45664 +INFO:tensorflow:step = 14701, loss = 0.762566, precision = 0.820312 (22.438 sec) +INFO:tensorflow:global_step/sec: 4.45475 +INFO:tensorflow:step = 14801, loss = 0.623383, precision = 0.890625 (22.448 sec) +Saved checkpoint after 38 epoch(s) to data/resnet56/checkpoints/00038... +INFO:tensorflow:global_step/sec: 4.31861 +INFO:tensorflow:step = 14901, loss = 0.543638, precision = 0.914062 (23.156 sec) +INFO:tensorflow:global_step/sec: 4.46276 +INFO:tensorflow:step = 15001, loss = 0.815708, precision = 0.835938 (22.408 sec) +INFO:tensorflow:global_step/sec: 4.45625 +INFO:tensorflow:step = 15101, loss = 0.602661, precision = 0.90625 (22.440 sec) +INFO:tensorflow:global_step/sec: 4.46087 +INFO:tensorflow:step = 15201, loss = 0.678878, precision = 0.851562 (22.417 sec) +Saved checkpoint after 39 epoch(s) to data/resnet56/checkpoints/00039... +INFO:tensorflow:global_step/sec: 4.31628 +INFO:tensorflow:step = 15301, loss = 0.59399, precision = 0.914062 (23.168 sec) +INFO:tensorflow:global_step/sec: 4.45798 +INFO:tensorflow:step = 15401, loss = 0.734938, precision = 0.84375 (22.432 sec) +INFO:tensorflow:global_step/sec: 4.45917 +INFO:tensorflow:step = 15501, loss = 0.564672, precision = 0.898438 (22.426 sec) +INFO:tensorflow:global_step/sec: 4.465 +INFO:tensorflow:step = 15601, loss = 0.799991, precision = 0.835938 (22.396 sec) +Saved checkpoint after 40 epoch(s) to data/resnet56/checkpoints/00040... +INFO:tensorflow:global_step/sec: 4.31517 +INFO:tensorflow:step = 15701, loss = 0.673989, precision = 0.875 (23.174 sec) +INFO:tensorflow:global_step/sec: 4.45874 +INFO:tensorflow:step = 15801, loss = 0.717926, precision = 0.820312 (22.428 sec) +INFO:tensorflow:global_step/sec: 4.46574 +INFO:tensorflow:step = 15901, loss = 0.7067, precision = 0.851562 (22.393 sec) +INFO:tensorflow:global_step/sec: 4.46908 +INFO:tensorflow:step = 16001, loss = 0.797281, precision = 0.84375 (22.376 sec) +Saved checkpoint after 41 epoch(s) to data/resnet56/checkpoints/00041... +INFO:tensorflow:global_step/sec: 4.32259 +INFO:tensorflow:step = 16101, loss = 0.48504, precision = 0.9375 (23.134 sec) +INFO:tensorflow:global_step/sec: 4.45888 +INFO:tensorflow:step = 16201, loss = 0.67983, precision = 0.867188 (22.427 sec) +INFO:tensorflow:global_step/sec: 4.4646 +INFO:tensorflow:step = 16301, loss = 0.758278, precision = 0.84375 (22.398 sec) +INFO:tensorflow:global_step/sec: 4.46445 +INFO:tensorflow:step = 16401, loss = 0.603408, precision = 0.882812 (22.399 sec) +Saved checkpoint after 42 epoch(s) to data/resnet56/checkpoints/00042... +INFO:tensorflow:global_step/sec: 4.31584 +INFO:tensorflow:step = 16501, loss = 0.679688, precision = 0.875 (23.171 sec) +INFO:tensorflow:global_step/sec: 4.46342 +INFO:tensorflow:step = 16601, loss = 0.665068, precision = 0.875 (22.404 sec) +INFO:tensorflow:global_step/sec: 4.46521 +INFO:tensorflow:step = 16701, loss = 0.571283, precision = 0.898438 (22.395 sec) +INFO:tensorflow:global_step/sec: 4.46467 +INFO:tensorflow:step = 16801, loss = 0.686729, precision = 0.898438 (22.398 sec) +Saved checkpoint after 43 epoch(s) to data/resnet56/checkpoints/00043... +INFO:tensorflow:global_step/sec: 4.32524 +INFO:tensorflow:step = 16901, loss = 0.628125, precision = 0.867188 (23.120 sec) +INFO:tensorflow:global_step/sec: 4.46822 +INFO:tensorflow:step = 17001, loss = 0.672178, precision = 0.867188 (22.380 sec) +INFO:tensorflow:global_step/sec: 4.46754 +INFO:tensorflow:step = 17101, loss = 0.555898, precision = 0.921875 (22.384 sec) +INFO:tensorflow:global_step/sec: 4.4712 +INFO:tensorflow:step = 17201, loss = 0.58739, precision = 0.898438 (22.365 sec) +Saved checkpoint after 44 epoch(s) to data/resnet56/checkpoints/00044... +INFO:tensorflow:global_step/sec: 4.32427 +INFO:tensorflow:step = 17301, loss = 0.587773, precision = 0.890625 (23.125 sec) +INFO:tensorflow:global_step/sec: 4.46832 +INFO:tensorflow:step = 17401, loss = 0.684193, precision = 0.875 (22.380 sec) +INFO:tensorflow:global_step/sec: 4.47095 +INFO:tensorflow:step = 17501, loss = 0.791584, precision = 0.84375 (22.367 sec) +Saved checkpoint after 45 epoch(s) to data/resnet56/checkpoints/00045... +INFO:tensorflow:global_step/sec: 4.30772 +INFO:tensorflow:step = 17601, loss = 0.554155, precision = 0.9375 (23.214 sec) +INFO:tensorflow:global_step/sec: 4.46798 +INFO:tensorflow:step = 17701, loss = 0.575025, precision = 0.90625 (22.381 sec) +INFO:tensorflow:global_step/sec: 4.47222 +INFO:tensorflow:step = 17801, loss = 0.613981, precision = 0.875 (22.360 sec) +INFO:tensorflow:global_step/sec: 4.47328 +INFO:tensorflow:step = 17901, loss = 0.799553, precision = 0.859375 (22.355 sec) +Saved checkpoint after 46 epoch(s) to data/resnet56/checkpoints/00046... +INFO:tensorflow:global_step/sec: 4.32748 +INFO:tensorflow:step = 18001, loss = 0.662487, precision = 0.851562 (23.108 sec) +INFO:tensorflow:global_step/sec: 4.47201 +INFO:tensorflow:step = 18101, loss = 0.534046, precision = 0.929688 (22.361 sec) +INFO:tensorflow:global_step/sec: 4.47037 +INFO:tensorflow:step = 18201, loss = 0.659802, precision = 0.851562 (22.370 sec) +INFO:tensorflow:global_step/sec: 4.47305 +INFO:tensorflow:step = 18301, loss = 0.653311, precision = 0.859375 (22.356 sec) +Saved checkpoint after 47 epoch(s) to data/resnet56/checkpoints/00047... +INFO:tensorflow:global_step/sec: 4.32432 +INFO:tensorflow:step = 18401, loss = 0.6255, precision = 0.890625 (23.125 sec) +INFO:tensorflow:global_step/sec: 4.47791 +INFO:tensorflow:step = 18501, loss = 0.694052, precision = 0.859375 (22.332 sec) +INFO:tensorflow:global_step/sec: 4.47571 +INFO:tensorflow:step = 18601, loss = 0.599611, precision = 0.890625 (22.343 sec) +INFO:tensorflow:global_step/sec: 4.47577 +INFO:tensorflow:step = 18701, loss = 0.531962, precision = 0.929688 (22.343 sec) +Saved checkpoint after 48 epoch(s) to data/resnet56/checkpoints/00048... +INFO:tensorflow:global_step/sec: 4.33427 +INFO:tensorflow:step = 18801, loss = 0.622415, precision = 0.890625 (23.072 sec) +INFO:tensorflow:global_step/sec: 4.478 +INFO:tensorflow:step = 18901, loss = 0.676464, precision = 0.882812 (22.331 sec) +INFO:tensorflow:global_step/sec: 4.47874 +INFO:tensorflow:step = 19001, loss = 0.557365, precision = 0.921875 (22.328 sec) +INFO:tensorflow:global_step/sec: 4.4817 +INFO:tensorflow:step = 19101, loss = 0.623122, precision = 0.875 (22.313 sec) +Saved checkpoint after 49 epoch(s) to data/resnet56/checkpoints/00049... +INFO:tensorflow:global_step/sec: 4.34775 +INFO:tensorflow:step = 19201, loss = 0.607808, precision = 0.882812 (23.000 sec) +INFO:tensorflow:global_step/sec: 4.48069 +INFO:tensorflow:step = 19301, loss = 0.627917, precision = 0.882812 (22.318 sec) +INFO:tensorflow:global_step/sec: 4.47963 +INFO:tensorflow:step = 19401, loss = 0.794117, precision = 0.804688 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.47432 +INFO:tensorflow:step = 19501, loss = 0.751475, precision = 0.84375 (22.350 sec) +Saved checkpoint after 50 epoch(s) to data/resnet56/checkpoints/00050... +INFO:tensorflow:global_step/sec: 4.3357 +INFO:tensorflow:step = 19601, loss = 0.639322, precision = 0.90625 (23.064 sec) +INFO:tensorflow:global_step/sec: 4.47892 +INFO:tensorflow:step = 19701, loss = 0.770086, precision = 0.867188 (22.327 sec) +INFO:tensorflow:global_step/sec: 4.48625 +INFO:tensorflow:step = 19801, loss = 0.654616, precision = 0.882812 (22.290 sec) +INFO:tensorflow:global_step/sec: 4.48281 +INFO:tensorflow:step = 19901, loss = 0.62572, precision = 0.898438 (22.307 sec) +Saved checkpoint after 51 epoch(s) to data/resnet56/checkpoints/00051... +INFO:tensorflow:global_step/sec: 4.34296 +INFO:tensorflow:step = 20001, loss = 0.528804, precision = 0.929688 (23.026 sec) +INFO:tensorflow:global_step/sec: 4.48234 +INFO:tensorflow:step = 20101, loss = 0.672676, precision = 0.882812 (22.310 sec) +INFO:tensorflow:global_step/sec: 4.47466 +INFO:tensorflow:step = 20201, loss = 0.633192, precision = 0.867188 (22.348 sec) +INFO:tensorflow:global_step/sec: 4.48073 +INFO:tensorflow:step = 20301, loss = 0.608602, precision = 0.890625 (22.318 sec) +Saved checkpoint after 52 epoch(s) to data/resnet56/checkpoints/00052... +INFO:tensorflow:global_step/sec: 4.33368 +INFO:tensorflow:step = 20401, loss = 0.561774, precision = 0.90625 (23.075 sec) +INFO:tensorflow:global_step/sec: 4.4827 +INFO:tensorflow:step = 20501, loss = 0.665953, precision = 0.898438 (22.308 sec) +INFO:tensorflow:global_step/sec: 4.47807 +INFO:tensorflow:step = 20601, loss = 0.673195, precision = 0.882812 (22.331 sec) +INFO:tensorflow:global_step/sec: 4.48257 +INFO:tensorflow:step = 20701, loss = 0.612822, precision = 0.898438 (22.309 sec) +Saved checkpoint after 53 epoch(s) to data/resnet56/checkpoints/00053... +INFO:tensorflow:global_step/sec: 4.33969 +INFO:tensorflow:step = 20801, loss = 0.718612, precision = 0.859375 (23.043 sec) +INFO:tensorflow:global_step/sec: 4.47322 +INFO:tensorflow:step = 20901, loss = 0.59768, precision = 0.890625 (22.355 sec) +INFO:tensorflow:global_step/sec: 4.47811 +INFO:tensorflow:step = 21001, loss = 0.722637, precision = 0.835938 (22.331 sec) +INFO:tensorflow:global_step/sec: 4.47801 +INFO:tensorflow:step = 21101, loss = 0.764791, precision = 0.820312 (22.331 sec) +Saved checkpoint after 54 epoch(s) to data/resnet56/checkpoints/00054... +INFO:tensorflow:global_step/sec: 4.33691 +INFO:tensorflow:step = 21201, loss = 0.629764, precision = 0.898438 (23.058 sec) +INFO:tensorflow:global_step/sec: 4.47383 +INFO:tensorflow:step = 21301, loss = 0.621165, precision = 0.867188 (22.352 sec) +INFO:tensorflow:global_step/sec: 4.47918 +INFO:tensorflow:step = 21401, loss = 0.763009, precision = 0.828125 (22.325 sec) +INFO:tensorflow:global_step/sec: 4.47769 +INFO:tensorflow:step = 21501, loss = 0.657822, precision = 0.882812 (22.333 sec) +Saved checkpoint after 55 epoch(s) to data/resnet56/checkpoints/00055... +INFO:tensorflow:global_step/sec: 4.33836 +INFO:tensorflow:step = 21601, loss = 0.672901, precision = 0.882812 (23.050 sec) +INFO:tensorflow:global_step/sec: 4.47617 +INFO:tensorflow:step = 21701, loss = 0.669233, precision = 0.90625 (22.340 sec) +INFO:tensorflow:global_step/sec: 4.47543 +INFO:tensorflow:step = 21801, loss = 0.629404, precision = 0.90625 (22.344 sec) +Saved checkpoint after 56 epoch(s) to data/resnet56/checkpoints/00056... +INFO:tensorflow:global_step/sec: 4.33196 +INFO:tensorflow:step = 21901, loss = 0.660433, precision = 0.867188 (23.084 sec) +INFO:tensorflow:global_step/sec: 4.47726 +INFO:tensorflow:step = 22001, loss = 0.699404, precision = 0.859375 (22.335 sec) +INFO:tensorflow:global_step/sec: 4.47488 +INFO:tensorflow:step = 22101, loss = 0.67517, precision = 0.859375 (22.347 sec) +INFO:tensorflow:global_step/sec: 4.47443 +INFO:tensorflow:step = 22201, loss = 0.765808, precision = 0.859375 (22.349 sec) +Saved checkpoint after 57 epoch(s) to data/resnet56/checkpoints/00057... +INFO:tensorflow:global_step/sec: 4.32415 +INFO:tensorflow:step = 22301, loss = 0.662129, precision = 0.867188 (23.126 sec) +INFO:tensorflow:global_step/sec: 4.47592 +INFO:tensorflow:step = 22401, loss = 0.664909, precision = 0.875 (22.342 sec) +INFO:tensorflow:global_step/sec: 4.47265 +INFO:tensorflow:step = 22501, loss = 0.648577, precision = 0.898438 (22.358 sec) +INFO:tensorflow:global_step/sec: 4.48161 +INFO:tensorflow:step = 22601, loss = 0.625795, precision = 0.90625 (22.313 sec) +Saved checkpoint after 58 epoch(s) to data/resnet56/checkpoints/00058... +INFO:tensorflow:global_step/sec: 4.33708 +INFO:tensorflow:step = 22701, loss = 0.60033, precision = 0.882812 (23.057 sec) +INFO:tensorflow:global_step/sec: 4.48357 +INFO:tensorflow:step = 22801, loss = 0.628079, precision = 0.875 (22.304 sec) +INFO:tensorflow:global_step/sec: 4.47184 +INFO:tensorflow:step = 22901, loss = 0.599005, precision = 0.890625 (22.362 sec) +INFO:tensorflow:global_step/sec: 4.47864 +INFO:tensorflow:step = 23001, loss = 0.627557, precision = 0.882812 (22.328 sec) +Saved checkpoint after 59 epoch(s) to data/resnet56/checkpoints/00059... +INFO:tensorflow:global_step/sec: 4.33705 +INFO:tensorflow:step = 23101, loss = 0.687221, precision = 0.890625 (23.057 sec) +INFO:tensorflow:global_step/sec: 4.48071 +INFO:tensorflow:step = 23201, loss = 0.617225, precision = 0.875 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.47639 +INFO:tensorflow:step = 23301, loss = 0.609474, precision = 0.890625 (22.335 sec) +INFO:tensorflow:global_step/sec: 4.47951 +INFO:tensorflow:step = 23401, loss = 0.678997, precision = 0.851562 (22.324 sec) +Saved checkpoint after 60 epoch(s) to data/resnet56/checkpoints/00060... +INFO:tensorflow:global_step/sec: 4.33811 +INFO:tensorflow:step = 23501, loss = 0.555011, precision = 0.90625 (23.052 sec) +INFO:tensorflow:global_step/sec: 4.47989 +INFO:tensorflow:step = 23601, loss = 0.567935, precision = 0.921875 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.47758 +INFO:tensorflow:step = 23701, loss = 0.659649, precision = 0.875 (22.334 sec) +INFO:tensorflow:global_step/sec: 4.47985 +INFO:tensorflow:step = 23801, loss = 0.540094, precision = 0.929688 (22.322 sec) +Saved checkpoint after 61 epoch(s) to data/resnet56/checkpoints/00061... +INFO:tensorflow:global_step/sec: 4.34054 +INFO:tensorflow:step = 23901, loss = 0.645397, precision = 0.898438 (23.039 sec) +INFO:tensorflow:global_step/sec: 4.47622 +INFO:tensorflow:step = 24001, loss = 0.634777, precision = 0.859375 (22.340 sec) +INFO:tensorflow:global_step/sec: 4.47513 +INFO:tensorflow:step = 24101, loss = 0.62364, precision = 0.882812 (22.346 sec) +INFO:tensorflow:global_step/sec: 4.47479 +INFO:tensorflow:step = 24201, loss = 0.616874, precision = 0.890625 (22.347 sec) +Saved checkpoint after 62 epoch(s) to data/resnet56/checkpoints/00062... +INFO:tensorflow:global_step/sec: 4.34177 +INFO:tensorflow:step = 24301, loss = 0.757324, precision = 0.851562 (23.032 sec) +INFO:tensorflow:global_step/sec: 4.48225 +INFO:tensorflow:step = 24401, loss = 0.53275, precision = 0.890625 (22.310 sec) +INFO:tensorflow:global_step/sec: 4.47543 +INFO:tensorflow:step = 24501, loss = 0.697075, precision = 0.859375 (22.344 sec) +INFO:tensorflow:global_step/sec: 4.47614 +INFO:tensorflow:step = 24601, loss = 0.578508, precision = 0.90625 (22.341 sec) +Saved checkpoint after 63 epoch(s) to data/resnet56/checkpoints/00063... +INFO:tensorflow:global_step/sec: 4.33184 +INFO:tensorflow:step = 24701, loss = 0.559016, precision = 0.921875 (23.085 sec) +INFO:tensorflow:global_step/sec: 4.47353 +INFO:tensorflow:step = 24801, loss = 0.621101, precision = 0.90625 (22.354 sec) +INFO:tensorflow:global_step/sec: 4.47975 +INFO:tensorflow:step = 24901, loss = 0.635789, precision = 0.90625 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.47496 +INFO:tensorflow:step = 25001, loss = 0.666004, precision = 0.898438 (22.347 sec) +Saved checkpoint after 64 epoch(s) to data/resnet56/checkpoints/00064... +INFO:tensorflow:global_step/sec: 4.34354 +INFO:tensorflow:step = 25101, loss = 0.526246, precision = 0.90625 (23.023 sec) +INFO:tensorflow:global_step/sec: 4.4783 +INFO:tensorflow:step = 25201, loss = 0.542472, precision = 0.929688 (22.330 sec) +INFO:tensorflow:global_step/sec: 4.47995 +INFO:tensorflow:step = 25301, loss = 0.732964, precision = 0.835938 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.48549 +INFO:tensorflow:step = 25401, loss = 0.547594, precision = 0.929688 (22.294 sec) +Saved checkpoint after 65 epoch(s) to data/resnet56/checkpoints/00065... +INFO:tensorflow:global_step/sec: 4.33989 +INFO:tensorflow:step = 25501, loss = 0.714811, precision = 0.835938 (23.042 sec) +INFO:tensorflow:global_step/sec: 4.48 +INFO:tensorflow:step = 25601, loss = 0.675548, precision = 0.875 (22.321 sec) +INFO:tensorflow:global_step/sec: 4.47825 +INFO:tensorflow:step = 25701, loss = 0.759873, precision = 0.84375 (22.330 sec) +INFO:tensorflow:global_step/sec: 4.48126 +INFO:tensorflow:step = 25801, loss = 0.665174, precision = 0.914062 (22.315 sec) +Saved checkpoint after 66 epoch(s) to data/resnet56/checkpoints/00066... +INFO:tensorflow:global_step/sec: 4.3475 +INFO:tensorflow:step = 25901, loss = 0.646115, precision = 0.890625 (23.002 sec) +INFO:tensorflow:global_step/sec: 4.47579 +INFO:tensorflow:step = 26001, loss = 0.622689, precision = 0.882812 (22.342 sec) +INFO:tensorflow:global_step/sec: 4.47762 +INFO:tensorflow:step = 26101, loss = 0.652957, precision = 0.890625 (22.333 sec) +Saved checkpoint after 67 epoch(s) to data/resnet56/checkpoints/00067... +INFO:tensorflow:global_step/sec: 4.34179 +INFO:tensorflow:step = 26201, loss = 0.538071, precision = 0.898438 (23.032 sec) +INFO:tensorflow:global_step/sec: 4.47965 +INFO:tensorflow:step = 26301, loss = 0.584642, precision = 0.90625 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.47668 +INFO:tensorflow:step = 26401, loss = 0.662661, precision = 0.898438 (22.338 sec) +INFO:tensorflow:global_step/sec: 4.48339 +INFO:tensorflow:step = 26501, loss = 0.644137, precision = 0.882812 (22.305 sec) +Saved checkpoint after 68 epoch(s) to data/resnet56/checkpoints/00068... +INFO:tensorflow:global_step/sec: 4.3416 +INFO:tensorflow:step = 26601, loss = 0.530003, precision = 0.929688 (23.033 sec) +INFO:tensorflow:global_step/sec: 4.47724 +INFO:tensorflow:step = 26701, loss = 0.633854, precision = 0.882812 (22.335 sec) +INFO:tensorflow:global_step/sec: 4.47527 +INFO:tensorflow:step = 26801, loss = 0.560745, precision = 0.898438 (22.345 sec) +INFO:tensorflow:global_step/sec: 4.47963 +INFO:tensorflow:step = 26901, loss = 0.759418, precision = 0.851562 (22.323 sec) +Saved checkpoint after 69 epoch(s) to data/resnet56/checkpoints/00069... +INFO:tensorflow:global_step/sec: 4.34373 +INFO:tensorflow:step = 27001, loss = 0.572407, precision = 0.921875 (23.022 sec) +INFO:tensorflow:global_step/sec: 4.47935 +INFO:tensorflow:step = 27101, loss = 0.549663, precision = 0.882812 (22.325 sec) +INFO:tensorflow:global_step/sec: 4.47604 +INFO:tensorflow:step = 27201, loss = 0.620045, precision = 0.882812 (22.341 sec) +INFO:tensorflow:global_step/sec: 4.47855 +INFO:tensorflow:step = 27301, loss = 0.647443, precision = 0.84375 (22.329 sec) +Saved checkpoint after 70 epoch(s) to data/resnet56/checkpoints/00070... +INFO:tensorflow:global_step/sec: 4.32367 +INFO:tensorflow:step = 27401, loss = 0.551731, precision = 0.898438 (23.129 sec) +INFO:tensorflow:global_step/sec: 4.47313 +INFO:tensorflow:step = 27501, loss = 0.689272, precision = 0.867188 (22.356 sec) +INFO:tensorflow:global_step/sec: 4.47547 +INFO:tensorflow:step = 27601, loss = 0.612163, precision = 0.898438 (22.344 sec) +INFO:tensorflow:global_step/sec: 4.48048 +INFO:tensorflow:step = 27701, loss = 0.558836, precision = 0.898438 (22.319 sec) +Saved checkpoint after 71 epoch(s) to data/resnet56/checkpoints/00071... +INFO:tensorflow:global_step/sec: 4.34061 +INFO:tensorflow:step = 27801, loss = 0.689491, precision = 0.867188 (23.038 sec) +INFO:tensorflow:global_step/sec: 4.48146 +INFO:tensorflow:step = 27901, loss = 0.66182, precision = 0.882812 (22.314 sec) +INFO:tensorflow:global_step/sec: 4.4795 +INFO:tensorflow:step = 28001, loss = 0.562205, precision = 0.90625 (22.324 sec) +INFO:tensorflow:global_step/sec: 4.47432 +INFO:tensorflow:step = 28101, loss = 0.590328, precision = 0.921875 (22.350 sec) +Saved checkpoint after 72 epoch(s) to data/resnet56/checkpoints/00072... +INFO:tensorflow:global_step/sec: 4.34281 +INFO:tensorflow:step = 28201, loss = 0.600064, precision = 0.898438 (23.027 sec) +INFO:tensorflow:global_step/sec: 4.48329 +INFO:tensorflow:step = 28301, loss = 0.578669, precision = 0.90625 (22.305 sec) +INFO:tensorflow:global_step/sec: 4.47475 +INFO:tensorflow:step = 28401, loss = 0.893421, precision = 0.796875 (22.348 sec) +INFO:tensorflow:global_step/sec: 4.47717 +INFO:tensorflow:step = 28501, loss = 0.601729, precision = 0.898438 (22.336 sec) +Saved checkpoint after 73 epoch(s) to data/resnet56/checkpoints/00073... +INFO:tensorflow:global_step/sec: 4.33757 +INFO:tensorflow:step = 28601, loss = 0.672622, precision = 0.875 (23.055 sec) +INFO:tensorflow:global_step/sec: 4.47439 +INFO:tensorflow:step = 28701, loss = 0.519379, precision = 0.929688 (22.349 sec) +INFO:tensorflow:global_step/sec: 4.47163 +INFO:tensorflow:step = 28801, loss = 0.641164, precision = 0.882812 (22.363 sec) +INFO:tensorflow:global_step/sec: 4.47881 +INFO:tensorflow:step = 28901, loss = 0.608074, precision = 0.882812 (22.327 sec) +Saved checkpoint after 74 epoch(s) to data/resnet56/checkpoints/00074... +INFO:tensorflow:global_step/sec: 4.3395 +INFO:tensorflow:step = 29001, loss = 0.747144, precision = 0.867188 (23.044 sec) +INFO:tensorflow:global_step/sec: 4.4759 +INFO:tensorflow:step = 29101, loss = 0.745581, precision = 0.851562 (22.342 sec) +INFO:tensorflow:global_step/sec: 4.47412 +INFO:tensorflow:step = 29201, loss = 0.616241, precision = 0.851562 (22.351 sec) +INFO:tensorflow:global_step/sec: 4.47901 +INFO:tensorflow:step = 29301, loss = 0.654456, precision = 0.90625 (22.326 sec) +Saved checkpoint after 75 epoch(s) to data/resnet56/checkpoints/00075... +INFO:tensorflow:global_step/sec: 4.34002 +INFO:tensorflow:step = 29401, loss = 0.629244, precision = 0.90625 (23.041 sec) +INFO:tensorflow:global_step/sec: 4.48011 +INFO:tensorflow:step = 29501, loss = 0.600061, precision = 0.90625 (22.321 sec) +INFO:tensorflow:global_step/sec: 4.47822 +INFO:tensorflow:step = 29601, loss = 0.643823, precision = 0.921875 (22.330 sec) +INFO:tensorflow:global_step/sec: 4.48098 +INFO:tensorflow:step = 29701, loss = 0.565762, precision = 0.90625 (22.317 sec) +Saved checkpoint after 76 epoch(s) to data/resnet56/checkpoints/00076... +INFO:tensorflow:global_step/sec: 4.3453 +INFO:tensorflow:step = 29801, loss = 0.624435, precision = 0.875 (23.013 sec) +INFO:tensorflow:global_step/sec: 4.47154 +INFO:tensorflow:step = 29901, loss = 0.714645, precision = 0.875 (22.364 sec) +INFO:tensorflow:global_step/sec: 4.4753 +INFO:tensorflow:step = 30001, loss = 0.64738, precision = 0.882812 (22.345 sec) +INFO:tensorflow:global_step/sec: 4.47762 +INFO:tensorflow:step = 30101, loss = 0.535702, precision = 0.929688 (22.333 sec) +Saved checkpoint after 77 epoch(s) to data/resnet56/checkpoints/00077... +INFO:tensorflow:global_step/sec: 4.34367 +INFO:tensorflow:step = 30201, loss = 0.583189, precision = 0.914062 (23.022 sec) +INFO:tensorflow:global_step/sec: 4.47254 +INFO:tensorflow:step = 30301, loss = 0.684224, precision = 0.875 (22.359 sec) +INFO:tensorflow:global_step/sec: 4.4748 +INFO:tensorflow:step = 30401, loss = 0.624616, precision = 0.90625 (22.347 sec) +Saved checkpoint after 78 epoch(s) to data/resnet56/checkpoints/00078... +INFO:tensorflow:global_step/sec: 4.34285 +INFO:tensorflow:step = 30501, loss = 0.717701, precision = 0.84375 (23.026 sec) +INFO:tensorflow:global_step/sec: 4.48109 +INFO:tensorflow:step = 30601, loss = 0.640411, precision = 0.875 (22.316 sec) +INFO:tensorflow:global_step/sec: 4.4703 +INFO:tensorflow:step = 30701, loss = 0.634734, precision = 0.914062 (22.370 sec) +INFO:tensorflow:global_step/sec: 4.47577 +INFO:tensorflow:step = 30801, loss = 0.682454, precision = 0.867188 (22.343 sec) +Saved checkpoint after 79 epoch(s) to data/resnet56/checkpoints/00079... +INFO:tensorflow:global_step/sec: 4.34613 +INFO:tensorflow:step = 30901, loss = 0.564666, precision = 0.898438 (23.009 sec) +INFO:tensorflow:global_step/sec: 4.47579 +INFO:tensorflow:step = 31001, loss = 0.577395, precision = 0.90625 (22.342 sec) +INFO:tensorflow:global_step/sec: 4.47583 +INFO:tensorflow:step = 31101, loss = 0.573427, precision = 0.914062 (22.342 sec) +INFO:tensorflow:global_step/sec: 4.48426 +INFO:tensorflow:step = 31201, loss = 0.798483, precision = 0.828125 (22.300 sec) +Saved checkpoint after 80 epoch(s) to data/resnet56/checkpoints/00080... +INFO:tensorflow:global_step/sec: 4.33997 +INFO:tensorflow:step = 31301, loss = 0.460844, precision = 0.960938 (23.042 sec) +INFO:tensorflow:global_step/sec: 4.48185 +INFO:tensorflow:step = 31401, loss = 0.599156, precision = 0.890625 (22.312 sec) +INFO:tensorflow:global_step/sec: 4.47427 +INFO:tensorflow:step = 31501, loss = 0.552579, precision = 0.9375 (22.350 sec) +INFO:tensorflow:global_step/sec: 4.48057 +INFO:tensorflow:step = 31601, loss = 0.559604, precision = 0.914062 (22.319 sec) +Saved checkpoint after 81 epoch(s) to data/resnet56/checkpoints/00081... +INFO:tensorflow:global_step/sec: 4.34039 +INFO:tensorflow:step = 31701, loss = 0.672168, precision = 0.875 (23.040 sec) +INFO:tensorflow:global_step/sec: 4.47813 +INFO:tensorflow:step = 31801, loss = 0.664845, precision = 0.84375 (22.331 sec) +INFO:tensorflow:global_step/sec: 4.47698 +INFO:tensorflow:step = 31901, loss = 0.517187, precision = 0.9375 (22.336 sec) +INFO:tensorflow:global_step/sec: 4.4755 +INFO:tensorflow:step = 32001, loss = 0.643888, precision = 0.898438 (22.344 sec) +Saved checkpoint after 82 epoch(s) to data/resnet56/checkpoints/00082... +INFO:tensorflow:global_step/sec: 4.33596 +INFO:tensorflow:step = 32101, loss = 0.537358, precision = 0.90625 (23.063 sec) +INFO:tensorflow:global_step/sec: 4.47342 +INFO:tensorflow:step = 32201, loss = 0.559992, precision = 0.90625 (22.354 sec) +INFO:tensorflow:global_step/sec: 4.47984 +INFO:tensorflow:step = 32301, loss = 0.655004, precision = 0.898438 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.47915 +INFO:tensorflow:step = 32401, loss = 0.700973, precision = 0.882812 (22.326 sec) +Saved checkpoint after 83 epoch(s) to data/resnet56/checkpoints/00083... +INFO:tensorflow:global_step/sec: 4.34492 +INFO:tensorflow:step = 32501, loss = 0.658736, precision = 0.914062 (23.015 sec) +INFO:tensorflow:global_step/sec: 4.47362 +INFO:tensorflow:step = 32601, loss = 0.739758, precision = 0.84375 (22.353 sec) +INFO:tensorflow:global_step/sec: 4.47558 +INFO:tensorflow:step = 32701, loss = 0.616397, precision = 0.898438 (22.344 sec) +INFO:tensorflow:global_step/sec: 4.47873 +INFO:tensorflow:step = 32801, loss = 0.605727, precision = 0.90625 (22.328 sec) +Saved checkpoint after 84 epoch(s) to data/resnet56/checkpoints/00084... +INFO:tensorflow:global_step/sec: 4.34232 +INFO:tensorflow:step = 32901, loss = 0.593016, precision = 0.90625 (23.029 sec) +INFO:tensorflow:global_step/sec: 4.47638 +INFO:tensorflow:step = 33001, loss = 0.546129, precision = 0.921875 (22.339 sec) +INFO:tensorflow:global_step/sec: 4.4772 +INFO:tensorflow:step = 33101, loss = 0.655493, precision = 0.882812 (22.335 sec) +INFO:tensorflow:global_step/sec: 4.47777 +INFO:tensorflow:step = 33201, loss = 0.701782, precision = 0.851562 (22.333 sec) +Saved checkpoint after 85 epoch(s) to data/resnet56/checkpoints/00085... +INFO:tensorflow:global_step/sec: 4.34783 +INFO:tensorflow:step = 33301, loss = 0.679748, precision = 0.867188 (23.000 sec) +INFO:tensorflow:global_step/sec: 4.48276 +INFO:tensorflow:step = 33401, loss = 0.647157, precision = 0.875 (22.308 sec) +INFO:tensorflow:global_step/sec: 4.4813 +INFO:tensorflow:step = 33501, loss = 0.653915, precision = 0.882812 (22.315 sec) +INFO:tensorflow:global_step/sec: 4.47889 +INFO:tensorflow:step = 33601, loss = 0.691286, precision = 0.890625 (22.327 sec) +Saved checkpoint after 86 epoch(s) to data/resnet56/checkpoints/00086... +INFO:tensorflow:global_step/sec: 4.34331 +INFO:tensorflow:step = 33701, loss = 0.56487, precision = 0.882812 (23.024 sec) +INFO:tensorflow:global_step/sec: 4.48051 +INFO:tensorflow:step = 33801, loss = 0.521744, precision = 0.90625 (22.319 sec) +INFO:tensorflow:global_step/sec: 4.48362 +INFO:tensorflow:step = 33901, loss = 0.683376, precision = 0.875 (22.303 sec) +INFO:tensorflow:global_step/sec: 4.48076 +INFO:tensorflow:step = 34001, loss = 0.642107, precision = 0.898438 (22.318 sec) +Saved checkpoint after 87 epoch(s) to data/resnet56/checkpoints/00087... +INFO:tensorflow:global_step/sec: 4.3393 +INFO:tensorflow:step = 34101, loss = 0.62013, precision = 0.867188 (23.045 sec) +INFO:tensorflow:global_step/sec: 4.47403 +INFO:tensorflow:step = 34201, loss = 0.611267, precision = 0.90625 (22.351 sec) +INFO:tensorflow:global_step/sec: 4.48027 +INFO:tensorflow:step = 34301, loss = 0.657234, precision = 0.882812 (22.320 sec) +INFO:tensorflow:global_step/sec: 4.48324 +INFO:tensorflow:step = 34401, loss = 0.632644, precision = 0.890625 (22.305 sec) +Saved checkpoint after 88 epoch(s) to data/resnet56/checkpoints/00088... +INFO:tensorflow:global_step/sec: 4.3333 +INFO:tensorflow:step = 34501, loss = 0.661909, precision = 0.835938 (23.077 sec) +INFO:tensorflow:global_step/sec: 4.48456 +INFO:tensorflow:step = 34601, loss = 0.662691, precision = 0.882812 (22.299 sec) +INFO:tensorflow:global_step/sec: 4.48414 +INFO:tensorflow:step = 34701, loss = 0.595145, precision = 0.867188 (22.301 sec) +Saved checkpoint after 89 epoch(s) to data/resnet56/checkpoints/00089... +INFO:tensorflow:global_step/sec: 4.34203 +INFO:tensorflow:step = 34801, loss = 0.613056, precision = 0.882812 (23.031 sec) +INFO:tensorflow:global_step/sec: 4.48459 +INFO:tensorflow:step = 34901, loss = 0.541507, precision = 0.90625 (22.298 sec) +INFO:tensorflow:global_step/sec: 4.47997 +INFO:tensorflow:step = 35001, loss = 0.619292, precision = 0.882812 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.47716 +INFO:tensorflow:step = 35101, loss = 0.75025, precision = 0.851562 (22.336 sec) +Saved checkpoint after 90 epoch(s) to data/resnet56/checkpoints/00090... +INFO:tensorflow:global_step/sec: 4.34034 +INFO:tensorflow:step = 35201, loss = 0.835665, precision = 0.835938 (23.040 sec) +INFO:tensorflow:global_step/sec: 4.47985 +INFO:tensorflow:step = 35301, loss = 0.562269, precision = 0.9375 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.47772 +INFO:tensorflow:step = 35401, loss = 0.680498, precision = 0.851562 (22.333 sec) +INFO:tensorflow:global_step/sec: 4.47444 +INFO:tensorflow:step = 35501, loss = 0.556598, precision = 0.929688 (22.349 sec) +Saved checkpoint after 91 epoch(s) to data/resnet56/checkpoints/00091... +INFO:tensorflow:global_step/sec: 4.33981 +INFO:tensorflow:step = 35601, loss = 0.6121, precision = 0.921875 (23.043 sec) +INFO:tensorflow:global_step/sec: 4.48139 +INFO:tensorflow:step = 35701, loss = 0.520096, precision = 0.945312 (22.314 sec) +INFO:tensorflow:global_step/sec: 4.48246 +INFO:tensorflow:step = 35801, loss = 0.465138, precision = 0.9375 (22.309 sec) +INFO:tensorflow:global_step/sec: 4.4839 +INFO:tensorflow:step = 35901, loss = 0.505517, precision = 0.929688 (22.302 sec) +Saved checkpoint after 92 epoch(s) to data/resnet56/checkpoints/00092... +INFO:tensorflow:global_step/sec: 4.3445 +INFO:tensorflow:step = 36001, loss = 0.56204, precision = 0.890625 (23.018 sec) +INFO:tensorflow:global_step/sec: 4.47519 +INFO:tensorflow:step = 36101, loss = 0.396846, precision = 0.976562 (22.345 sec) +INFO:tensorflow:global_step/sec: 4.47793 +INFO:tensorflow:step = 36201, loss = 0.442978, precision = 0.921875 (22.332 sec) +INFO:tensorflow:global_step/sec: 4.48186 +INFO:tensorflow:step = 36301, loss = 0.440646, precision = 0.960938 (22.312 sec) +Saved checkpoint after 93 epoch(s) to data/resnet56/checkpoints/00093... +INFO:tensorflow:global_step/sec: 4.34313 +INFO:tensorflow:step = 36401, loss = 0.486813, precision = 0.945312 (23.025 sec) +INFO:tensorflow:global_step/sec: 4.4799 +INFO:tensorflow:step = 36501, loss = 0.381302, precision = 0.976562 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.47822 +INFO:tensorflow:step = 36601, loss = 0.429124, precision = 0.945312 (22.330 sec) +INFO:tensorflow:global_step/sec: 4.47654 +INFO:tensorflow:step = 36701, loss = 0.47484, precision = 0.929688 (22.339 sec) +Saved checkpoint after 94 epoch(s) to data/resnet56/checkpoints/00094... +INFO:tensorflow:global_step/sec: 4.32386 +INFO:tensorflow:step = 36801, loss = 0.412389, precision = 0.945312 (23.128 sec) +INFO:tensorflow:global_step/sec: 4.47529 +INFO:tensorflow:step = 36901, loss = 0.396952, precision = 0.96875 (22.345 sec) +INFO:tensorflow:global_step/sec: 4.47541 +INFO:tensorflow:step = 37001, loss = 0.390477, precision = 0.953125 (22.344 sec) +INFO:tensorflow:global_step/sec: 4.47774 +INFO:tensorflow:step = 37101, loss = 0.330942, precision = 0.992188 (22.333 sec) +Saved checkpoint after 95 epoch(s) to data/resnet56/checkpoints/00095... +INFO:tensorflow:global_step/sec: 4.34205 +INFO:tensorflow:step = 37201, loss = 0.386468, precision = 0.953125 (23.031 sec) +INFO:tensorflow:global_step/sec: 4.47964 +INFO:tensorflow:step = 37301, loss = 0.360481, precision = 0.96875 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.47682 +INFO:tensorflow:step = 37401, loss = 0.534517, precision = 0.9375 (22.337 sec) +INFO:tensorflow:global_step/sec: 4.48362 +INFO:tensorflow:step = 37501, loss = 0.338728, precision = 0.984375 (22.303 sec) +Saved checkpoint after 96 epoch(s) to data/resnet56/checkpoints/00096... +INFO:tensorflow:global_step/sec: 4.34685 +INFO:tensorflow:step = 37601, loss = 0.33943, precision = 0.96875 (23.005 sec) +INFO:tensorflow:global_step/sec: 4.47697 +INFO:tensorflow:step = 37701, loss = 0.374605, precision = 0.945312 (22.336 sec) +INFO:tensorflow:global_step/sec: 4.48373 +INFO:tensorflow:step = 37801, loss = 0.325652, precision = 0.976562 (22.303 sec) +INFO:tensorflow:global_step/sec: 4.47936 +INFO:tensorflow:step = 37901, loss = 0.356839, precision = 0.960938 (22.325 sec) +Saved checkpoint after 97 epoch(s) to data/resnet56/checkpoints/00097... +INFO:tensorflow:global_step/sec: 4.34856 +INFO:tensorflow:step = 38001, loss = 0.327103, precision = 0.96875 (22.996 sec) +INFO:tensorflow:global_step/sec: 4.47705 +INFO:tensorflow:step = 38101, loss = 0.382744, precision = 0.945312 (22.336 sec) +INFO:tensorflow:global_step/sec: 4.47596 +INFO:tensorflow:step = 38201, loss = 0.429803, precision = 0.929688 (22.342 sec) +INFO:tensorflow:global_step/sec: 4.47858 +INFO:tensorflow:step = 38301, loss = 0.325897, precision = 0.96875 (22.328 sec) +Saved checkpoint after 98 epoch(s) to data/resnet56/checkpoints/00098... +INFO:tensorflow:global_step/sec: 4.34182 +INFO:tensorflow:step = 38401, loss = 0.376224, precision = 0.953125 (23.032 sec) +INFO:tensorflow:global_step/sec: 4.48366 +INFO:tensorflow:step = 38501, loss = 0.411681, precision = 0.921875 (22.303 sec) +INFO:tensorflow:global_step/sec: 4.48795 +INFO:tensorflow:step = 38601, loss = 0.377877, precision = 0.945312 (22.282 sec) +INFO:tensorflow:global_step/sec: 4.48599 +INFO:tensorflow:step = 38701, loss = 0.372027, precision = 0.960938 (22.292 sec) +Saved checkpoint after 99 epoch(s) to data/resnet56/checkpoints/00099... +INFO:tensorflow:global_step/sec: 4.34256 +INFO:tensorflow:step = 38801, loss = 0.338772, precision = 0.953125 (23.028 sec) +INFO:tensorflow:global_step/sec: 4.4757 +INFO:tensorflow:step = 38901, loss = 0.326271, precision = 0.976562 (22.343 sec) +INFO:tensorflow:global_step/sec: 4.48563 +INFO:tensorflow:step = 39001, loss = 0.334021, precision = 0.960938 (22.293 sec) +Saved checkpoint after 100 epoch(s) to data/resnet56/checkpoints/00100... +INFO:tensorflow:global_step/sec: 4.34941 +INFO:tensorflow:step = 39101, loss = 0.288725, precision = 0.984375 (22.992 sec) +INFO:tensorflow:global_step/sec: 4.48832 +INFO:tensorflow:step = 39201, loss = 0.257883, precision = 1.0 (22.280 sec) +INFO:tensorflow:global_step/sec: 4.48933 +INFO:tensorflow:step = 39301, loss = 0.297077, precision = 0.96875 (22.275 sec) +INFO:tensorflow:global_step/sec: 4.48674 +INFO:tensorflow:step = 39401, loss = 0.275303, precision = 0.984375 (22.288 sec) +Saved checkpoint after 101 epoch(s) to data/resnet56/checkpoints/00101... +INFO:tensorflow:global_step/sec: 4.349 +INFO:tensorflow:step = 39501, loss = 0.288223, precision = 0.96875 (22.994 sec) +INFO:tensorflow:global_step/sec: 4.48821 +INFO:tensorflow:step = 39601, loss = 0.384414, precision = 0.960938 (22.280 sec) +INFO:tensorflow:global_step/sec: 4.48147 +INFO:tensorflow:step = 39701, loss = 0.26662, precision = 0.992188 (22.314 sec) +INFO:tensorflow:global_step/sec: 4.48824 +INFO:tensorflow:step = 39801, loss = 0.294838, precision = 0.96875 (22.280 sec) +Saved checkpoint after 102 epoch(s) to data/resnet56/checkpoints/00102... +INFO:tensorflow:global_step/sec: 4.34559 +INFO:tensorflow:step = 39901, loss = 0.355804, precision = 0.945312 (23.012 sec) +INFO:tensorflow:global_step/sec: 4.48779 +INFO:tensorflow:step = 40001, loss = 0.286997, precision = 0.96875 (22.283 sec) +INFO:tensorflow:global_step/sec: 4.4836 +INFO:tensorflow:step = 40101, loss = 0.36991, precision = 0.953125 (22.303 sec) +INFO:tensorflow:global_step/sec: 4.48958 +INFO:tensorflow:step = 40201, loss = 0.254701, precision = 0.984375 (22.274 sec) +Saved checkpoint after 103 epoch(s) to data/resnet56/checkpoints/00103... +INFO:tensorflow:global_step/sec: 4.34825 +INFO:tensorflow:step = 40301, loss = 0.305098, precision = 0.976562 (22.998 sec) +INFO:tensorflow:global_step/sec: 4.49055 +INFO:tensorflow:step = 40401, loss = 0.309185, precision = 0.96875 (22.269 sec) +INFO:tensorflow:global_step/sec: 4.48873 +INFO:tensorflow:step = 40501, loss = 0.246391, precision = 0.984375 (22.278 sec) +INFO:tensorflow:global_step/sec: 4.48569 +INFO:tensorflow:step = 40601, loss = 0.25039, precision = 0.984375 (22.293 sec) +Saved checkpoint after 104 epoch(s) to data/resnet56/checkpoints/00104... +INFO:tensorflow:global_step/sec: 4.35119 +INFO:tensorflow:step = 40701, loss = 0.266932, precision = 0.976562 (22.982 sec) +INFO:tensorflow:global_step/sec: 4.48528 +INFO:tensorflow:step = 40801, loss = 0.271679, precision = 0.976562 (22.295 sec) +INFO:tensorflow:global_step/sec: 4.48386 +INFO:tensorflow:step = 40901, loss = 0.240861, precision = 0.96875 (22.302 sec) +INFO:tensorflow:global_step/sec: 4.48583 +INFO:tensorflow:step = 41001, loss = 0.309205, precision = 0.976562 (22.292 sec) +Saved checkpoint after 105 epoch(s) to data/resnet56/checkpoints/00105... +INFO:tensorflow:global_step/sec: 4.3463 +INFO:tensorflow:step = 41101, loss = 0.24821, precision = 0.992188 (23.009 sec) +INFO:tensorflow:global_step/sec: 4.48993 +INFO:tensorflow:step = 41201, loss = 0.299468, precision = 0.960938 (22.271 sec) +INFO:tensorflow:global_step/sec: 4.48647 +INFO:tensorflow:step = 41301, loss = 0.345025, precision = 0.96875 (22.289 sec) +INFO:tensorflow:global_step/sec: 4.48144 +INFO:tensorflow:step = 41401, loss = 0.235546, precision = 1.0 (22.314 sec) +Saved checkpoint after 106 epoch(s) to data/resnet56/checkpoints/00106... +INFO:tensorflow:global_step/sec: 4.34517 +INFO:tensorflow:step = 41501, loss = 0.300669, precision = 0.96875 (23.014 sec) +INFO:tensorflow:global_step/sec: 4.4814 +INFO:tensorflow:step = 41601, loss = 0.210292, precision = 0.992188 (22.314 sec) +INFO:tensorflow:global_step/sec: 4.48339 +INFO:tensorflow:step = 41701, loss = 0.273247, precision = 0.96875 (22.305 sec) +INFO:tensorflow:global_step/sec: 4.48042 +INFO:tensorflow:step = 41801, loss = 0.245743, precision = 0.984375 (22.319 sec) +Saved checkpoint after 107 epoch(s) to data/resnet56/checkpoints/00107... +INFO:tensorflow:global_step/sec: 4.33776 +INFO:tensorflow:step = 41901, loss = 0.224736, precision = 0.992188 (23.053 sec) +INFO:tensorflow:global_step/sec: 4.47813 +INFO:tensorflow:step = 42001, loss = 0.251512, precision = 0.976562 (22.331 sec) +INFO:tensorflow:global_step/sec: 4.47972 +INFO:tensorflow:step = 42101, loss = 0.249587, precision = 0.976562 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.48522 +INFO:tensorflow:step = 42201, loss = 0.240219, precision = 0.976562 (22.295 sec) +Saved checkpoint after 108 epoch(s) to data/resnet56/checkpoints/00108... +INFO:tensorflow:global_step/sec: 4.34102 +INFO:tensorflow:step = 42301, loss = 0.25533, precision = 0.976562 (23.036 sec) +INFO:tensorflow:global_step/sec: 4.47833 +INFO:tensorflow:step = 42401, loss = 0.243456, precision = 0.992188 (22.330 sec) +INFO:tensorflow:global_step/sec: 4.48129 +INFO:tensorflow:step = 42501, loss = 0.256046, precision = 0.976562 (22.315 sec) +INFO:tensorflow:global_step/sec: 4.47707 +INFO:tensorflow:step = 42601, loss = 0.279839, precision = 0.96875 (22.336 sec) +Saved checkpoint after 109 epoch(s) to data/resnet56/checkpoints/00109... +INFO:tensorflow:global_step/sec: 4.34756 +INFO:tensorflow:step = 42701, loss = 0.280275, precision = 0.96875 (23.001 sec) +INFO:tensorflow:global_step/sec: 4.47851 +INFO:tensorflow:step = 42801, loss = 0.234846, precision = 0.976562 (22.329 sec) +INFO:tensorflow:global_step/sec: 4.47383 +INFO:tensorflow:step = 42901, loss = 0.233972, precision = 0.976562 (22.352 sec) +INFO:tensorflow:global_step/sec: 4.47416 +INFO:tensorflow:step = 43001, loss = 0.23397, precision = 0.96875 (22.351 sec) +Saved checkpoint after 110 epoch(s) to data/resnet56/checkpoints/00110... +INFO:tensorflow:global_step/sec: 4.34654 +INFO:tensorflow:step = 43101, loss = 0.246549, precision = 0.96875 (23.007 sec) +INFO:tensorflow:global_step/sec: 4.48409 +INFO:tensorflow:step = 43201, loss = 0.279334, precision = 0.945312 (22.301 sec) +INFO:tensorflow:global_step/sec: 4.48103 +INFO:tensorflow:step = 43301, loss = 0.218195, precision = 0.992188 (22.316 sec) +Saved checkpoint after 111 epoch(s) to data/resnet56/checkpoints/00111... +INFO:tensorflow:global_step/sec: 4.34718 +INFO:tensorflow:step = 43401, loss = 0.192685, precision = 1.0 (23.003 sec) +INFO:tensorflow:global_step/sec: 4.48474 +INFO:tensorflow:step = 43501, loss = 0.248958, precision = 0.96875 (22.298 sec) +INFO:tensorflow:global_step/sec: 4.47636 +INFO:tensorflow:step = 43601, loss = 0.233495, precision = 0.976562 (22.340 sec) +INFO:tensorflow:global_step/sec: 4.47568 +INFO:tensorflow:step = 43701, loss = 0.193737, precision = 1.0 (22.343 sec) +Saved checkpoint after 112 epoch(s) to data/resnet56/checkpoints/00112... +INFO:tensorflow:global_step/sec: 4.33993 +INFO:tensorflow:step = 43801, loss = 0.237111, precision = 0.96875 (23.042 sec) +INFO:tensorflow:global_step/sec: 4.48388 +INFO:tensorflow:step = 43901, loss = 0.238463, precision = 0.984375 (22.302 sec) +INFO:tensorflow:global_step/sec: 4.48472 +INFO:tensorflow:step = 44001, loss = 0.265159, precision = 0.945312 (22.298 sec) +INFO:tensorflow:global_step/sec: 4.47543 +INFO:tensorflow:step = 44101, loss = 0.24876, precision = 0.960938 (22.344 sec) +Saved checkpoint after 113 epoch(s) to data/resnet56/checkpoints/00113... +INFO:tensorflow:global_step/sec: 4.34298 +INFO:tensorflow:step = 44201, loss = 0.19425, precision = 0.992188 (23.026 sec) +INFO:tensorflow:global_step/sec: 4.47977 +INFO:tensorflow:step = 44301, loss = 0.243087, precision = 0.960938 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.48392 +INFO:tensorflow:step = 44401, loss = 0.249867, precision = 0.96875 (22.302 sec) +INFO:tensorflow:global_step/sec: 4.47989 +INFO:tensorflow:step = 44501, loss = 0.22165, precision = 0.984375 (22.322 sec) +Saved checkpoint after 114 epoch(s) to data/resnet56/checkpoints/00114... +INFO:tensorflow:global_step/sec: 4.3426 +INFO:tensorflow:step = 44601, loss = 0.179656, precision = 1.0 (23.028 sec) +INFO:tensorflow:global_step/sec: 4.48289 +INFO:tensorflow:step = 44701, loss = 0.210221, precision = 0.984375 (22.307 sec) +INFO:tensorflow:global_step/sec: 4.4824 +INFO:tensorflow:step = 44801, loss = 0.187466, precision = 1.0 (22.309 sec) +INFO:tensorflow:global_step/sec: 4.48283 +INFO:tensorflow:step = 44901, loss = 0.260017, precision = 0.953125 (22.307 sec) +Saved checkpoint after 115 epoch(s) to data/resnet56/checkpoints/00115... +INFO:tensorflow:global_step/sec: 4.34147 +INFO:tensorflow:step = 45001, loss = 0.222655, precision = 0.976562 (23.034 sec) +INFO:tensorflow:global_step/sec: 4.48651 +INFO:tensorflow:step = 45101, loss = 0.178365, precision = 1.0 (22.289 sec) +INFO:tensorflow:global_step/sec: 4.47934 +INFO:tensorflow:step = 45201, loss = 0.17727, precision = 1.0 (22.325 sec) +INFO:tensorflow:global_step/sec: 4.48161 +INFO:tensorflow:step = 45301, loss = 0.195759, precision = 0.984375 (22.313 sec) +Saved checkpoint after 116 epoch(s) to data/resnet56/checkpoints/00116... +INFO:tensorflow:global_step/sec: 4.339 +INFO:tensorflow:step = 45401, loss = 0.184987, precision = 1.0 (23.047 sec) +INFO:tensorflow:global_step/sec: 4.48939 +INFO:tensorflow:step = 45501, loss = 0.210008, precision = 0.984375 (22.275 sec) +INFO:tensorflow:global_step/sec: 4.47894 +INFO:tensorflow:step = 45601, loss = 0.273858, precision = 0.960938 (22.327 sec) +INFO:tensorflow:global_step/sec: 4.48194 +INFO:tensorflow:step = 45701, loss = 0.193758, precision = 0.984375 (22.312 sec) +Saved checkpoint after 117 epoch(s) to data/resnet56/checkpoints/00117... +INFO:tensorflow:global_step/sec: 4.34338 +INFO:tensorflow:step = 45801, loss = 0.193953, precision = 0.984375 (23.024 sec) +INFO:tensorflow:global_step/sec: 4.47802 +INFO:tensorflow:step = 45901, loss = 0.209495, precision = 0.984375 (22.331 sec) +INFO:tensorflow:global_step/sec: 4.4868 +INFO:tensorflow:step = 46001, loss = 0.262984, precision = 0.9375 (22.288 sec) +INFO:tensorflow:global_step/sec: 4.4823 +INFO:tensorflow:step = 46101, loss = 0.25817, precision = 0.96875 (22.310 sec) +Saved checkpoint after 118 epoch(s) to data/resnet56/checkpoints/00118... +INFO:tensorflow:global_step/sec: 4.34303 +INFO:tensorflow:step = 46201, loss = 0.2217, precision = 0.984375 (23.025 sec) +INFO:tensorflow:global_step/sec: 4.48511 +INFO:tensorflow:step = 46301, loss = 0.207195, precision = 0.984375 (22.296 sec) +INFO:tensorflow:global_step/sec: 4.48103 +INFO:tensorflow:step = 46401, loss = 0.237368, precision = 0.96875 (22.316 sec) +INFO:tensorflow:global_step/sec: 4.48291 +INFO:tensorflow:step = 46501, loss = 0.301492, precision = 0.9375 (22.307 sec) +Saved checkpoint after 119 epoch(s) to data/resnet56/checkpoints/00119... +INFO:tensorflow:global_step/sec: 4.33829 +INFO:tensorflow:step = 46601, loss = 0.177623, precision = 0.992188 (23.051 sec) +INFO:tensorflow:global_step/sec: 4.482 +INFO:tensorflow:step = 46701, loss = 0.203984, precision = 0.984375 (22.311 sec) +INFO:tensorflow:global_step/sec: 4.48457 +INFO:tensorflow:step = 46801, loss = 0.200425, precision = 0.976562 (22.299 sec) +INFO:tensorflow:global_step/sec: 4.47776 +INFO:tensorflow:step = 46901, loss = 0.172613, precision = 1.0 (22.333 sec) +Saved checkpoint after 120 epoch(s) to data/resnet56/checkpoints/00120... +INFO:tensorflow:global_step/sec: 4.33559 +INFO:tensorflow:step = 47001, loss = 0.24131, precision = 0.960938 (23.065 sec) +INFO:tensorflow:global_step/sec: 4.48291 +INFO:tensorflow:step = 47101, loss = 0.248983, precision = 0.953125 (22.307 sec) +INFO:tensorflow:global_step/sec: 4.48423 +INFO:tensorflow:step = 47201, loss = 0.234525, precision = 0.953125 (22.300 sec) +INFO:tensorflow:global_step/sec: 4.47729 +INFO:tensorflow:step = 47301, loss = 0.193822, precision = 0.976562 (22.335 sec) +Saved checkpoint after 121 epoch(s) to data/resnet56/checkpoints/00121... +INFO:tensorflow:global_step/sec: 4.33783 +INFO:tensorflow:step = 47401, loss = 0.212934, precision = 0.976562 (23.053 sec) +INFO:tensorflow:global_step/sec: 4.48521 +INFO:tensorflow:step = 47501, loss = 0.195632, precision = 0.992188 (22.295 sec) +INFO:tensorflow:global_step/sec: 4.4798 +INFO:tensorflow:step = 47601, loss = 0.219186, precision = 0.976562 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.479 +INFO:tensorflow:step = 47701, loss = 0.20215, precision = 0.96875 (22.326 sec) +Saved checkpoint after 122 epoch(s) to data/resnet56/checkpoints/00122... +INFO:tensorflow:global_step/sec: 4.34762 +INFO:tensorflow:step = 47801, loss = 0.187719, precision = 0.984375 (23.001 sec) +INFO:tensorflow:global_step/sec: 4.48012 +INFO:tensorflow:step = 47901, loss = 0.1874, precision = 1.0 (22.321 sec) +INFO:tensorflow:global_step/sec: 4.482 +INFO:tensorflow:step = 48001, loss = 0.230754, precision = 0.976562 (22.312 sec) +Saved checkpoint after 123 epoch(s) to data/resnet56/checkpoints/00123... +INFO:tensorflow:global_step/sec: 4.34144 +INFO:tensorflow:step = 48101, loss = 0.201421, precision = 0.976562 (23.034 sec) +INFO:tensorflow:global_step/sec: 4.47557 +INFO:tensorflow:step = 48201, loss = 0.165825, precision = 1.0 (22.343 sec) +INFO:tensorflow:global_step/sec: 4.48165 +INFO:tensorflow:step = 48301, loss = 0.264286, precision = 0.960938 (22.313 sec) +INFO:tensorflow:global_step/sec: 4.48304 +INFO:tensorflow:step = 48401, loss = 0.225219, precision = 0.976562 (22.306 sec) +Saved checkpoint after 124 epoch(s) to data/resnet56/checkpoints/00124... +INFO:tensorflow:global_step/sec: 4.3409 +INFO:tensorflow:step = 48501, loss = 0.193306, precision = 0.984375 (23.037 sec) +INFO:tensorflow:global_step/sec: 4.48242 +INFO:tensorflow:step = 48601, loss = 0.190381, precision = 0.976562 (22.309 sec) +INFO:tensorflow:global_step/sec: 4.48642 +INFO:tensorflow:step = 48701, loss = 0.20198, precision = 0.984375 (22.290 sec) +INFO:tensorflow:global_step/sec: 4.47669 +INFO:tensorflow:step = 48801, loss = 0.195457, precision = 0.976562 (22.338 sec) +Saved checkpoint after 125 epoch(s) to data/resnet56/checkpoints/00125... +INFO:tensorflow:global_step/sec: 4.34613 +INFO:tensorflow:step = 48901, loss = 0.210935, precision = 0.976562 (23.009 sec) +INFO:tensorflow:global_step/sec: 4.47447 +INFO:tensorflow:step = 49001, loss = 0.192821, precision = 0.992188 (22.349 sec) +INFO:tensorflow:global_step/sec: 4.48047 +INFO:tensorflow:step = 49101, loss = 0.231033, precision = 0.976562 (22.319 sec) +INFO:tensorflow:global_step/sec: 4.47873 +INFO:tensorflow:step = 49201, loss = 0.183505, precision = 0.984375 (22.328 sec) +Saved checkpoint after 126 epoch(s) to data/resnet56/checkpoints/00126... +INFO:tensorflow:global_step/sec: 4.34338 +INFO:tensorflow:step = 49301, loss = 0.200501, precision = 0.976562 (23.024 sec) +INFO:tensorflow:global_step/sec: 4.48086 +INFO:tensorflow:step = 49401, loss = 0.247788, precision = 0.960938 (22.317 sec) +INFO:tensorflow:global_step/sec: 4.47511 +INFO:tensorflow:step = 49501, loss = 0.262762, precision = 0.96875 (22.346 sec) +INFO:tensorflow:global_step/sec: 4.48023 +INFO:tensorflow:step = 49601, loss = 0.194367, precision = 0.96875 (22.320 sec) +Saved checkpoint after 127 epoch(s) to data/resnet56/checkpoints/00127... +INFO:tensorflow:global_step/sec: 4.34194 +INFO:tensorflow:step = 49701, loss = 0.182765, precision = 0.992188 (23.031 sec) +INFO:tensorflow:global_step/sec: 4.48237 +INFO:tensorflow:step = 49801, loss = 0.21972, precision = 0.976562 (22.309 sec) +INFO:tensorflow:global_step/sec: 4.47712 +INFO:tensorflow:step = 49901, loss = 0.171858, precision = 0.992188 (22.336 sec) +INFO:tensorflow:global_step/sec: 4.48064 +INFO:tensorflow:step = 50001, loss = 0.17638, precision = 0.992188 (22.318 sec) +Saved checkpoint after 128 epoch(s) to data/resnet56/checkpoints/00128... +INFO:tensorflow:global_step/sec: 4.3388 +INFO:tensorflow:step = 50101, loss = 0.200966, precision = 0.984375 (23.048 sec) +INFO:tensorflow:global_step/sec: 4.48539 +INFO:tensorflow:step = 50201, loss = 0.233755, precision = 0.96875 (22.295 sec) +INFO:tensorflow:global_step/sec: 4.47623 +INFO:tensorflow:step = 50301, loss = 0.171946, precision = 0.984375 (22.340 sec) +INFO:tensorflow:global_step/sec: 4.47669 +INFO:tensorflow:step = 50401, loss = 0.208355, precision = 0.976562 (22.338 sec) +Saved checkpoint after 129 epoch(s) to data/resnet56/checkpoints/00129... +INFO:tensorflow:global_step/sec: 4.3338 +INFO:tensorflow:step = 50501, loss = 0.246959, precision = 0.976562 (23.075 sec) +INFO:tensorflow:global_step/sec: 4.47872 +INFO:tensorflow:step = 50601, loss = 0.178178, precision = 0.984375 (22.328 sec) +INFO:tensorflow:global_step/sec: 4.47227 +INFO:tensorflow:step = 50701, loss = 0.207603, precision = 0.976562 (22.360 sec) +INFO:tensorflow:global_step/sec: 4.48074 +INFO:tensorflow:step = 50801, loss = 0.231859, precision = 0.976562 (22.318 sec) +Saved checkpoint after 130 epoch(s) to data/resnet56/checkpoints/00130... +INFO:tensorflow:global_step/sec: 4.34069 +INFO:tensorflow:step = 50901, loss = 0.172081, precision = 0.992188 (23.038 sec) +INFO:tensorflow:global_step/sec: 4.48334 +INFO:tensorflow:step = 51001, loss = 0.153538, precision = 1.0 (22.305 sec) +INFO:tensorflow:global_step/sec: 4.47645 +INFO:tensorflow:step = 51101, loss = 0.17441, precision = 0.984375 (22.339 sec) +INFO:tensorflow:global_step/sec: 4.48145 +INFO:tensorflow:step = 51201, loss = 0.254012, precision = 0.953125 (22.314 sec) +Saved checkpoint after 131 epoch(s) to data/resnet56/checkpoints/00131... +INFO:tensorflow:global_step/sec: 4.33846 +INFO:tensorflow:step = 51301, loss = 0.219521, precision = 0.976562 (23.050 sec) +INFO:tensorflow:global_step/sec: 4.47799 +INFO:tensorflow:step = 51401, loss = 0.27856, precision = 0.953125 (22.331 sec) +INFO:tensorflow:global_step/sec: 4.47534 +INFO:tensorflow:step = 51501, loss = 0.160786, precision = 0.992188 (22.345 sec) +INFO:tensorflow:global_step/sec: 4.48282 +INFO:tensorflow:step = 51601, loss = 0.20776, precision = 0.960938 (22.307 sec) +Saved checkpoint after 132 epoch(s) to data/resnet56/checkpoints/00132... +INFO:tensorflow:global_step/sec: 4.31808 +INFO:tensorflow:step = 51701, loss = 0.252294, precision = 0.945312 (23.158 sec) +INFO:tensorflow:global_step/sec: 4.48097 +INFO:tensorflow:step = 51801, loss = 0.184217, precision = 0.976562 (22.317 sec) +INFO:tensorflow:global_step/sec: 4.48308 +INFO:tensorflow:step = 51901, loss = 0.193138, precision = 0.976562 (22.306 sec) +INFO:tensorflow:global_step/sec: 4.48254 +INFO:tensorflow:step = 52001, loss = 0.190958, precision = 0.984375 (22.309 sec) +Saved checkpoint after 133 epoch(s) to data/resnet56/checkpoints/00133... +INFO:tensorflow:global_step/sec: 4.34351 +INFO:tensorflow:step = 52101, loss = 0.169937, precision = 0.992188 (23.023 sec) +INFO:tensorflow:global_step/sec: 4.48319 +INFO:tensorflow:step = 52201, loss = 0.231622, precision = 0.96875 (22.305 sec) +INFO:tensorflow:global_step/sec: 4.47884 +INFO:tensorflow:step = 52301, loss = 0.220849, precision = 0.976562 (22.327 sec) +Saved checkpoint after 134 epoch(s) to data/resnet56/checkpoints/00134... +INFO:tensorflow:global_step/sec: 4.34468 +INFO:tensorflow:step = 52401, loss = 0.212435, precision = 0.976562 (23.017 sec) +INFO:tensorflow:global_step/sec: 4.48068 +INFO:tensorflow:step = 52501, loss = 0.200846, precision = 0.984375 (22.318 sec) +INFO:tensorflow:global_step/sec: 4.48375 +INFO:tensorflow:step = 52601, loss = 0.164221, precision = 0.992188 (22.303 sec) +INFO:tensorflow:global_step/sec: 4.48177 +INFO:tensorflow:step = 52701, loss = 0.169099, precision = 0.992188 (22.313 sec) +Saved checkpoint after 135 epoch(s) to data/resnet56/checkpoints/00135... +INFO:tensorflow:global_step/sec: 4.33768 +INFO:tensorflow:step = 52801, loss = 0.271096, precision = 0.945312 (23.054 sec) +INFO:tensorflow:global_step/sec: 4.47956 +INFO:tensorflow:step = 52901, loss = 0.176045, precision = 0.992188 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.47534 +INFO:tensorflow:step = 53001, loss = 0.2366, precision = 0.976562 (22.345 sec) +INFO:tensorflow:global_step/sec: 4.47484 +INFO:tensorflow:step = 53101, loss = 0.149071, precision = 1.0 (22.347 sec) +Saved checkpoint after 136 epoch(s) to data/resnet56/checkpoints/00136... +INFO:tensorflow:global_step/sec: 4.33933 +INFO:tensorflow:step = 53201, loss = 0.161579, precision = 1.0 (23.045 sec) +INFO:tensorflow:global_step/sec: 4.48373 +INFO:tensorflow:step = 53301, loss = 0.181923, precision = 0.984375 (22.303 sec) +INFO:tensorflow:global_step/sec: 4.4789 +INFO:tensorflow:step = 53401, loss = 0.155796, precision = 1.0 (22.327 sec) +INFO:tensorflow:global_step/sec: 4.48422 +INFO:tensorflow:step = 53501, loss = 0.144522, precision = 1.0 (22.300 sec) +Saved checkpoint after 137 epoch(s) to data/resnet56/checkpoints/00137... +INFO:tensorflow:global_step/sec: 4.34049 +INFO:tensorflow:step = 53601, loss = 0.145669, precision = 1.0 (23.039 sec) +INFO:tensorflow:global_step/sec: 4.48142 +INFO:tensorflow:step = 53701, loss = 0.147998, precision = 1.0 (22.314 sec) +INFO:tensorflow:global_step/sec: 4.48171 +INFO:tensorflow:step = 53801, loss = 0.14366, precision = 1.0 (22.313 sec) +INFO:tensorflow:global_step/sec: 4.48139 +INFO:tensorflow:step = 53901, loss = 0.143514, precision = 1.0 (22.314 sec) +Saved checkpoint after 138 epoch(s) to data/resnet56/checkpoints/00138... +INFO:tensorflow:global_step/sec: 4.33924 +INFO:tensorflow:step = 54001, loss = 0.137567, precision = 1.0 (23.046 sec) +INFO:tensorflow:global_step/sec: 4.47542 +INFO:tensorflow:step = 54101, loss = 0.17957, precision = 0.984375 (22.344 sec) +INFO:tensorflow:global_step/sec: 4.48675 +INFO:tensorflow:step = 54201, loss = 0.156771, precision = 0.992188 (22.288 sec) +INFO:tensorflow:global_step/sec: 4.4836 +INFO:tensorflow:step = 54301, loss = 0.160249, precision = 0.992188 (22.304 sec) +Saved checkpoint after 139 epoch(s) to data/resnet56/checkpoints/00139... +INFO:tensorflow:global_step/sec: 4.33506 +INFO:tensorflow:step = 54401, loss = 0.160879, precision = 0.992188 (23.068 sec) +INFO:tensorflow:global_step/sec: 4.48021 +INFO:tensorflow:step = 54501, loss = 0.143232, precision = 1.0 (22.320 sec) +INFO:tensorflow:global_step/sec: 4.47303 +INFO:tensorflow:step = 54601, loss = 0.154995, precision = 1.0 (22.356 sec) +INFO:tensorflow:global_step/sec: 4.47648 +INFO:tensorflow:step = 54701, loss = 0.139482, precision = 1.0 (22.339 sec) +Saved checkpoint after 140 epoch(s) to data/resnet56/checkpoints/00140... +INFO:tensorflow:global_step/sec: 4.33303 +INFO:tensorflow:step = 54801, loss = 0.142696, precision = 1.0 (23.079 sec) +INFO:tensorflow:global_step/sec: 4.47933 +INFO:tensorflow:step = 54901, loss = 0.138538, precision = 1.0 (22.325 sec) +INFO:tensorflow:global_step/sec: 4.47866 +INFO:tensorflow:step = 55001, loss = 0.164991, precision = 0.992188 (22.328 sec) +INFO:tensorflow:global_step/sec: 4.48238 +INFO:tensorflow:step = 55101, loss = 0.146175, precision = 1.0 (22.310 sec) +Saved checkpoint after 141 epoch(s) to data/resnet56/checkpoints/00141... +INFO:tensorflow:global_step/sec: 4.33851 +INFO:tensorflow:step = 55201, loss = 0.136705, precision = 1.0 (23.049 sec) +INFO:tensorflow:global_step/sec: 4.48255 +INFO:tensorflow:step = 55301, loss = 0.139411, precision = 1.0 (22.309 sec) +INFO:tensorflow:global_step/sec: 4.48042 +INFO:tensorflow:step = 55401, loss = 0.143066, precision = 1.0 (22.319 sec) +INFO:tensorflow:global_step/sec: 4.48267 +INFO:tensorflow:step = 55501, loss = 0.158531, precision = 0.992188 (22.308 sec) +Saved checkpoint after 142 epoch(s) to data/resnet56/checkpoints/00142... +INFO:tensorflow:global_step/sec: 4.34279 +INFO:tensorflow:step = 55601, loss = 0.14717, precision = 0.992188 (23.027 sec) +INFO:tensorflow:global_step/sec: 4.47954 +INFO:tensorflow:step = 55701, loss = 0.146559, precision = 0.992188 (22.324 sec) +INFO:tensorflow:global_step/sec: 4.47985 +INFO:tensorflow:step = 55801, loss = 0.137211, precision = 1.0 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.47472 +INFO:tensorflow:step = 55901, loss = 0.158828, precision = 0.992188 (22.348 sec) +Saved checkpoint after 143 epoch(s) to data/resnet56/checkpoints/00143... +INFO:tensorflow:global_step/sec: 4.33274 +INFO:tensorflow:step = 56001, loss = 0.182425, precision = 0.984375 (23.080 sec) +INFO:tensorflow:global_step/sec: 4.47698 +INFO:tensorflow:step = 56101, loss = 0.138114, precision = 1.0 (22.336 sec) +INFO:tensorflow:global_step/sec: 4.47812 +INFO:tensorflow:step = 56201, loss = 0.159647, precision = 0.992188 (22.331 sec) +INFO:tensorflow:global_step/sec: 4.47673 +INFO:tensorflow:step = 56301, loss = 0.139376, precision = 1.0 (22.338 sec) +Saved checkpoint after 144 epoch(s) to data/resnet56/checkpoints/00144... +INFO:tensorflow:global_step/sec: 4.33179 +INFO:tensorflow:step = 56401, loss = 0.14621, precision = 1.0 (23.085 sec) +INFO:tensorflow:global_step/sec: 4.4806 +INFO:tensorflow:step = 56501, loss = 0.13547, precision = 1.0 (22.318 sec) +INFO:tensorflow:global_step/sec: 4.48532 +INFO:tensorflow:step = 56601, loss = 0.138424, precision = 1.0 (22.295 sec) +Saved checkpoint after 145 epoch(s) to data/resnet56/checkpoints/00145... +INFO:tensorflow:global_step/sec: 4.34248 +INFO:tensorflow:step = 56701, loss = 0.140022, precision = 1.0 (23.028 sec) +INFO:tensorflow:global_step/sec: 4.47968 +INFO:tensorflow:step = 56801, loss = 0.170526, precision = 0.992188 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.48045 +INFO:tensorflow:step = 56901, loss = 0.134399, precision = 1.0 (22.319 sec) +INFO:tensorflow:global_step/sec: 4.47981 +INFO:tensorflow:step = 57001, loss = 0.134355, precision = 1.0 (22.322 sec) +Saved checkpoint after 146 epoch(s) to data/resnet56/checkpoints/00146... +INFO:tensorflow:global_step/sec: 4.34094 +INFO:tensorflow:step = 57101, loss = 0.132593, precision = 1.0 (23.037 sec) +INFO:tensorflow:global_step/sec: 4.48563 +INFO:tensorflow:step = 57201, loss = 0.137958, precision = 1.0 (22.293 sec) +INFO:tensorflow:global_step/sec: 4.47805 +INFO:tensorflow:step = 57301, loss = 0.138448, precision = 1.0 (22.331 sec) +INFO:tensorflow:global_step/sec: 4.48145 +INFO:tensorflow:step = 57401, loss = 0.167958, precision = 0.984375 (22.314 sec) +Saved checkpoint after 147 epoch(s) to data/resnet56/checkpoints/00147... +INFO:tensorflow:global_step/sec: 4.33084 +INFO:tensorflow:step = 57501, loss = 0.154431, precision = 0.984375 (23.090 sec) +INFO:tensorflow:global_step/sec: 4.48561 +INFO:tensorflow:step = 57601, loss = 0.1426, precision = 1.0 (22.293 sec) +INFO:tensorflow:global_step/sec: 4.4814 +INFO:tensorflow:step = 57701, loss = 0.134405, precision = 1.0 (22.314 sec) +INFO:tensorflow:global_step/sec: 4.47833 +INFO:tensorflow:step = 57801, loss = 0.141328, precision = 0.992188 (22.330 sec) +Saved checkpoint after 148 epoch(s) to data/resnet56/checkpoints/00148... +INFO:tensorflow:global_step/sec: 4.33785 +INFO:tensorflow:step = 57901, loss = 0.134456, precision = 1.0 (23.053 sec) +INFO:tensorflow:global_step/sec: 4.48032 +INFO:tensorflow:step = 58001, loss = 0.132753, precision = 1.0 (22.320 sec) +INFO:tensorflow:global_step/sec: 4.47772 +INFO:tensorflow:step = 58101, loss = 0.157741, precision = 0.984375 (22.333 sec) +INFO:tensorflow:global_step/sec: 4.48172 +INFO:tensorflow:step = 58201, loss = 0.148143, precision = 0.992188 (22.313 sec) +Saved checkpoint after 149 epoch(s) to data/resnet56/checkpoints/00149... +INFO:tensorflow:global_step/sec: 4.34384 +INFO:tensorflow:step = 58301, loss = 0.133841, precision = 1.0 (23.021 sec) +INFO:tensorflow:global_step/sec: 4.4812 +INFO:tensorflow:step = 58401, loss = 0.132977, precision = 1.0 (22.315 sec) +INFO:tensorflow:global_step/sec: 4.48208 +INFO:tensorflow:step = 58501, loss = 0.134882, precision = 1.0 (22.311 sec) +INFO:tensorflow:global_step/sec: 4.4815 +INFO:tensorflow:step = 58601, loss = 0.132694, precision = 1.0 (22.314 sec) +Saved checkpoint after 150 epoch(s) to data/resnet56/checkpoints/00150... +INFO:tensorflow:global_step/sec: 4.34224 +INFO:tensorflow:step = 58701, loss = 0.133808, precision = 1.0 (23.030 sec) +INFO:tensorflow:global_step/sec: 4.47921 +INFO:tensorflow:step = 58801, loss = 0.129201, precision = 1.0 (22.325 sec) +INFO:tensorflow:global_step/sec: 4.46929 +INFO:tensorflow:step = 58901, loss = 0.140288, precision = 1.0 (22.375 sec) +INFO:tensorflow:global_step/sec: 4.4825 +INFO:tensorflow:step = 59001, loss = 0.133575, precision = 1.0 (22.309 sec) +Saved checkpoint after 151 epoch(s) to data/resnet56/checkpoints/00151... +INFO:tensorflow:global_step/sec: 4.32784 +INFO:tensorflow:step = 59101, loss = 0.144807, precision = 0.992188 (23.106 sec) +INFO:tensorflow:global_step/sec: 4.47973 +INFO:tensorflow:step = 59201, loss = 0.130317, precision = 1.0 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.47977 +INFO:tensorflow:step = 59301, loss = 0.130915, precision = 1.0 (22.323 sec) +INFO:tensorflow:global_step/sec: 4.48327 +INFO:tensorflow:step = 59401, loss = 0.136418, precision = 1.0 (22.305 sec) +Saved checkpoint after 152 epoch(s) to data/resnet56/checkpoints/00152... +INFO:tensorflow:global_step/sec: 4.33346 +INFO:tensorflow:step = 59501, loss = 0.128513, precision = 1.0 (23.076 sec) +INFO:tensorflow:global_step/sec: 4.48858 +INFO:tensorflow:step = 59601, loss = 0.131417, precision = 1.0 (22.279 sec) +INFO:tensorflow:global_step/sec: 4.48755 +INFO:tensorflow:step = 59701, loss = 0.131489, precision = 1.0 (22.284 sec) +INFO:tensorflow:global_step/sec: 4.48466 +INFO:tensorflow:step = 59801, loss = 0.131894, precision = 1.0 (22.298 sec) +Saved checkpoint after 153 epoch(s) to data/resnet56/checkpoints/00153... +INFO:tensorflow:global_step/sec: 4.34303 +INFO:tensorflow:step = 59901, loss = 0.131072, precision = 1.0 (23.025 sec) +INFO:tensorflow:global_step/sec: 4.4846 +INFO:tensorflow:step = 60001, loss = 0.135706, precision = 0.992188 (22.298 sec) +INFO:tensorflow:global_step/sec: 4.48388 +INFO:tensorflow:step = 60101, loss = 0.133754, precision = 1.0 (22.302 sec) +INFO:tensorflow:global_step/sec: 4.48525 +INFO:tensorflow:step = 60201, loss = 0.12826, precision = 1.0 (22.295 sec) +Saved checkpoint after 154 epoch(s) to data/resnet56/checkpoints/00154... +INFO:tensorflow:global_step/sec: 4.34269 +INFO:tensorflow:step = 60301, loss = 0.128271, precision = 1.0 (23.027 sec) +INFO:tensorflow:global_step/sec: 4.48577 +INFO:tensorflow:step = 60401, loss = 0.134825, precision = 1.0 (22.293 sec) +INFO:tensorflow:global_step/sec: 4.4882 +INFO:tensorflow:step = 60501, loss = 0.129867, precision = 1.0 (22.281 sec) +INFO:tensorflow:global_step/sec: 4.49154 +INFO:tensorflow:step = 60601, loss = 0.132661, precision = 1.0 (22.264 sec) +Saved checkpoint after 155 epoch(s) to data/resnet56/checkpoints/00155... +INFO:tensorflow:global_step/sec: 4.34361 +INFO:tensorflow:step = 60701, loss = 0.130404, precision = 1.0 (23.022 sec) +INFO:tensorflow:global_step/sec: 4.48301 +INFO:tensorflow:step = 60801, loss = 0.141821, precision = 1.0 (22.306 sec) +INFO:tensorflow:global_step/sec: 4.48388 +INFO:tensorflow:step = 60901, loss = 0.130439, precision = 1.0 (22.302 sec) +Saved checkpoint after 156 epoch(s) to data/resnet56/checkpoints/00156... +INFO:tensorflow:global_step/sec: 4.33937 +INFO:tensorflow:step = 61001, loss = 0.129646, precision = 1.0 (23.045 sec) +INFO:tensorflow:global_step/sec: 4.48644 +INFO:tensorflow:step = 61101, loss = 0.134808, precision = 1.0 (22.289 sec) +INFO:tensorflow:global_step/sec: 4.48773 +INFO:tensorflow:step = 61201, loss = 0.13193, precision = 1.0 (22.283 sec) +INFO:tensorflow:global_step/sec: 4.48406 +INFO:tensorflow:step = 61301, loss = 0.129897, precision = 1.0 (22.301 sec) +Saved checkpoint after 157 epoch(s) to data/resnet56/checkpoints/00157... +INFO:tensorflow:global_step/sec: 4.32243 +INFO:tensorflow:step = 61401, loss = 0.130939, precision = 1.0 (23.135 sec) +INFO:tensorflow:global_step/sec: 4.48269 +INFO:tensorflow:step = 61501, loss = 0.127118, precision = 1.0 (22.308 sec) +INFO:tensorflow:global_step/sec: 4.49009 +INFO:tensorflow:step = 61601, loss = 0.129474, precision = 1.0 (22.271 sec) +INFO:tensorflow:global_step/sec: 4.48176 +INFO:tensorflow:step = 61701, loss = 0.127278, precision = 1.0 (22.313 sec) +Saved checkpoint after 158 epoch(s) to data/resnet56/checkpoints/00158... +INFO:tensorflow:global_step/sec: 4.33774 +INFO:tensorflow:step = 61801, loss = 0.133745, precision = 1.0 (23.054 sec) +INFO:tensorflow:global_step/sec: 4.47851 +INFO:tensorflow:step = 61901, loss = 0.161725, precision = 0.984375 (22.329 sec) +INFO:tensorflow:global_step/sec: 4.48109 +INFO:tensorflow:step = 62001, loss = 0.127429, precision = 1.0 (22.316 sec) +INFO:tensorflow:global_step/sec: 4.4733 +INFO:tensorflow:step = 62101, loss = 0.12917, precision = 1.0 (22.355 sec) +Saved checkpoint after 159 epoch(s) to data/resnet56/checkpoints/00159... +INFO:tensorflow:global_step/sec: 4.33521 +INFO:tensorflow:step = 62201, loss = 0.128144, precision = 1.0 (23.067 sec) +INFO:tensorflow:global_step/sec: 4.47615 +INFO:tensorflow:step = 62301, loss = 0.126545, precision = 1.0 (22.340 sec) +INFO:tensorflow:global_step/sec: 4.47715 +INFO:tensorflow:step = 62401, loss = 0.12681, precision = 1.0 (22.336 sec) +INFO:tensorflow:global_step/sec: 4.47607 +INFO:tensorflow:step = 62501, loss = 0.131018, precision = 1.0 (22.341 sec) +Saved checkpoint after 160 epoch(s) to data/resnet56/checkpoints/00160... +INFO:tensorflow:global_step/sec: 4.33473 +INFO:tensorflow:step = 62601, loss = 0.134579, precision = 1.0 (23.070 sec) +INFO:tensorflow:global_step/sec: 4.48048 +INFO:tensorflow:step = 62701, loss = 0.143882, precision = 0.992188 (22.319 sec) +INFO:tensorflow:global_step/sec: 4.47906 +INFO:tensorflow:step = 62801, loss = 0.125715, precision = 1.0 (22.326 sec) +INFO:tensorflow:global_step/sec: 4.4815 +INFO:tensorflow:step = 62901, loss = 0.127995, precision = 1.0 (22.314 sec) +Saved checkpoint after 161 epoch(s) to data/resnet56/checkpoints/00161... +INFO:tensorflow:global_step/sec: 4.3358 +INFO:tensorflow:step = 63001, loss = 0.125639, precision = 1.0 (23.064 sec) +INFO:tensorflow:global_step/sec: 4.47921 +INFO:tensorflow:step = 63101, loss = 0.124965, precision = 1.0 (22.325 sec) +INFO:tensorflow:global_step/sec: 4.47788 +INFO:tensorflow:step = 63201, loss = 0.126725, precision = 1.0 (22.332 sec) +INFO:tensorflow:global_step/sec: 4.48471 +INFO:tensorflow:step = 63301, loss = 0.125203, precision = 1.0 (22.298 sec) +Saved checkpoint after 162 epoch(s) to data/resnet56/checkpoints/00162... +INFO:tensorflow:global_step/sec: 4.3371 +INFO:tensorflow:step = 63401, loss = 0.126243, precision = 1.0 (23.057 sec) +INFO:tensorflow:global_step/sec: 4.48173 +INFO:tensorflow:step = 63501, loss = 0.124489, precision = 1.0 (22.313 sec) +INFO:tensorflow:global_step/sec: 4.47798 +INFO:tensorflow:step = 63601, loss = 0.125195, precision = 1.0 (22.332 sec) +INFO:tensorflow:global_step/sec: 4.48272 +INFO:tensorflow:step = 63701, loss = 0.128869, precision = 1.0 (22.308 sec) +Saved checkpoint after 163 epoch(s) to data/resnet56/checkpoints/00163... +INFO:tensorflow:global_step/sec: 4.34311 +INFO:tensorflow:step = 63801, loss = 0.125886, precision = 1.0 (23.025 sec) +INFO:tensorflow:global_step/sec: 4.47832 +INFO:tensorflow:step = 63901, loss = 0.129613, precision = 1.0 (22.330 sec) +INFO:tensorflow:global_step/sec: 4.48262 +INFO:tensorflow:step = 64001, loss = 0.124661, precision = 1.0 (22.308 sec) +INFO:tensorflow:global_step/sec: 4.48411 +INFO:tensorflow:step = 64101, loss = 0.125143, precision = 1.0 (22.301 sec) +Saved checkpoint after 164 epoch(s) to data/resnet56/checkpoints/00164... +INFO:tensorflow:global_step/sec: 4.3437 +INFO:tensorflow:step = 64201, loss = 0.135453, precision = 0.992188 (23.022 sec) +INFO:tensorflow:global_step/sec: 4.48249 +INFO:tensorflow:step = 64301, loss = 0.126442, precision = 1.0 (22.309 sec) +INFO:tensorflow:global_step/sec: 4.48253 +INFO:tensorflow:step = 64401, loss = 0.152345, precision = 0.984375 (22.309 sec) +INFO:tensorflow:global_step/sec: 4.48186 +INFO:tensorflow:step = 64501, loss = 0.12349, precision = 1.0 (22.312 sec) +Saved checkpoint after 165 epoch(s) to data/resnet56/checkpoints/00165... +INFO:tensorflow:global_step/sec: 4.33857 +INFO:tensorflow:step = 64601, loss = 0.133, precision = 0.992188 (23.049 sec) +INFO:tensorflow:global_step/sec: 4.47874 +INFO:tensorflow:step = 64701, loss = 0.123719, precision = 1.0 (22.328 sec) +INFO:tensorflow:global_step/sec: 4.47748 +INFO:tensorflow:step = 64801, loss = 0.128324, precision = 0.992188 (22.334 sec) +INFO:tensorflow:global_step/sec: 4.47989 +INFO:tensorflow:step = 64901, loss = 0.125893, precision = 1.0 (22.322 sec) +Saved checkpoint after 166 epoch(s) to data/resnet56/checkpoints/00166... +INFO:tensorflow:global_step/sec: 4.33694 +INFO:tensorflow:step = 65001, loss = 0.125396, precision = 1.0 (23.058 sec) +INFO:tensorflow:global_step/sec: 4.48446 +INFO:tensorflow:step = 65101, loss = 0.123658, precision = 1.0 (22.299 sec) +INFO:tensorflow:global_step/sec: 4.47953 +INFO:tensorflow:step = 65201, loss = 0.138463, precision = 1.0 (22.324 sec) +Saved checkpoint after 167 epoch(s) to data/resnet56/checkpoints/00167... +INFO:tensorflow:global_step/sec: 4.33663 +INFO:tensorflow:step = 65301, loss = 0.12838, precision = 1.0 (23.059 sec) +INFO:tensorflow:global_step/sec: 4.48216 +INFO:tensorflow:step = 65401, loss = 0.122669, precision = 1.0 (22.311 sec) +INFO:tensorflow:global_step/sec: 4.47724 +INFO:tensorflow:step = 65501, loss = 0.122214, precision = 1.0 (22.335 sec) +INFO:tensorflow:global_step/sec: 4.4791 +INFO:tensorflow:step = 65601, loss = 0.124259, precision = 1.0 (22.326 sec) +Saved checkpoint after 168 epoch(s) to data/resnet56/checkpoints/00168... +INFO:tensorflow:global_step/sec: 4.33045 +INFO:tensorflow:step = 65701, loss = 0.123013, precision = 1.0 (23.092 sec) +INFO:tensorflow:global_step/sec: 4.48721 +INFO:tensorflow:step = 65801, loss = 0.123734, precision = 1.0 (22.285 sec) +INFO:tensorflow:global_step/sec: 4.48269 +INFO:tensorflow:step = 65901, loss = 0.124381, precision = 1.0 (22.308 sec) +INFO:tensorflow:global_step/sec: 4.47774 +INFO:tensorflow:step = 66001, loss = 0.121759, precision = 1.0 (22.333 sec) +Saved checkpoint after 169 epoch(s) to data/resnet56/checkpoints/00169... +INFO:tensorflow:global_step/sec: 4.33476 +INFO:tensorflow:step = 66101, loss = 0.124492, precision = 1.0 (23.069 sec) +INFO:tensorflow:global_step/sec: 4.48523 +INFO:tensorflow:step = 66201, loss = 0.137384, precision = 0.992188 (22.295 sec) +INFO:tensorflow:global_step/sec: 4.47598 +INFO:tensorflow:step = 66301, loss = 0.134893, precision = 1.0 (22.341 sec) +INFO:tensorflow:global_step/sec: 4.47661 +INFO:tensorflow:step = 66401, loss = 0.126795, precision = 1.0 (22.338 sec) +Saved checkpoint after 170 epoch(s) to data/resnet56/checkpoints/00170... +INFO:tensorflow:global_step/sec: 4.31497 +INFO:tensorflow:step = 66501, loss = 0.122441, precision = 1.0 (23.175 sec) +INFO:tensorflow:global_step/sec: 4.4825 +INFO:tensorflow:step = 66601, loss = 0.125501, precision = 1.0 (22.309 sec) +INFO:tensorflow:global_step/sec: 4.47707 +INFO:tensorflow:step = 66701, loss = 0.120738, precision = 1.0 (22.336 sec) +INFO:tensorflow:global_step/sec: 4.47815 +INFO:tensorflow:step = 66801, loss = 0.120692, precision = 1.0 (22.331 sec) +Saved checkpoint after 171 epoch(s) to data/resnet56/checkpoints/00171... +INFO:tensorflow:global_step/sec: 4.33143 +INFO:tensorflow:step = 66901, loss = 0.138936, precision = 0.992188 (23.087 sec) +INFO:tensorflow:global_step/sec: 4.47911 +INFO:tensorflow:step = 67001, loss = 0.122162, precision = 1.0 (22.326 sec) +INFO:tensorflow:global_step/sec: 4.47735 +INFO:tensorflow:step = 67101, loss = 0.128169, precision = 1.0 (22.335 sec) +INFO:tensorflow:global_step/sec: 4.47897 +INFO:tensorflow:step = 67201, loss = 0.121105, precision = 1.0 (22.327 sec) +Saved checkpoint after 172 epoch(s) to data/resnet56/checkpoints/00172... +INFO:tensorflow:global_step/sec: 4.33922 +INFO:tensorflow:step = 67301, loss = 0.12241, precision = 1.0 (23.046 sec) +INFO:tensorflow:global_step/sec: 4.47899 +INFO:tensorflow:step = 67401, loss = 0.122816, precision = 1.0 (22.326 sec) +INFO:tensorflow:global_step/sec: 4.4784 +INFO:tensorflow:step = 67501, loss = 0.127149, precision = 1.0 (22.329 sec) +INFO:tensorflow:global_step/sec: 4.48018 +INFO:tensorflow:step = 67601, loss = 0.119974, precision = 1.0 (22.321 sec) +Saved checkpoint after 173 epoch(s) to data/resnet56/checkpoints/00173... +INFO:tensorflow:global_step/sec: 4.34183 +INFO:tensorflow:step = 67701, loss = 0.132654, precision = 0.992188 (23.032 sec) +INFO:tensorflow:global_step/sec: 4.48177 +INFO:tensorflow:step = 67801, loss = 0.119521, precision = 1.0 (22.313 sec) +INFO:tensorflow:global_step/sec: 4.48036 +INFO:tensorflow:step = 67901, loss = 0.127807, precision = 0.992188 (22.320 sec) +INFO:tensorflow:global_step/sec: 4.47754 +INFO:tensorflow:step = 68001, loss = 0.123621, precision = 1.0 (22.334 sec) +Saved checkpoint after 174 epoch(s) to data/resnet56/checkpoints/00174... +INFO:tensorflow:global_step/sec: 4.33457 +INFO:tensorflow:step = 68101, loss = 0.121717, precision = 1.0 (23.070 sec) +INFO:tensorflow:global_step/sec: 4.47978 +INFO:tensorflow:step = 68201, loss = 0.121072, precision = 1.0 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.47941 +INFO:tensorflow:step = 68301, loss = 0.132922, precision = 0.992188 (22.324 sec) +INFO:tensorflow:global_step/sec: 4.48033 +INFO:tensorflow:step = 68401, loss = 0.128112, precision = 1.0 (22.320 sec) +Saved checkpoint after 175 epoch(s) to data/resnet56/checkpoints/00175... +INFO:tensorflow:global_step/sec: 4.33442 +INFO:tensorflow:step = 68501, loss = 0.124797, precision = 1.0 (23.071 sec) +INFO:tensorflow:global_step/sec: 4.48112 +INFO:tensorflow:step = 68601, loss = 0.121132, precision = 1.0 (22.316 sec) +INFO:tensorflow:global_step/sec: 4.48206 +INFO:tensorflow:step = 68701, loss = 0.118618, precision = 1.0 (22.311 sec) +INFO:tensorflow:global_step/sec: 4.47812 +INFO:tensorflow:step = 68801, loss = 0.117399, precision = 1.0 (22.331 sec) +Saved checkpoint after 176 epoch(s) to data/resnet56/checkpoints/00176... +INFO:tensorflow:global_step/sec: 4.3387 +INFO:tensorflow:step = 68901, loss = 0.117801, precision = 1.0 (23.048 sec) +INFO:tensorflow:global_step/sec: 4.47823 +INFO:tensorflow:step = 69001, loss = 0.119858, precision = 1.0 (22.330 sec) +INFO:tensorflow:global_step/sec: 4.475 +INFO:tensorflow:step = 69101, loss = 0.128117, precision = 1.0 (22.346 sec) +INFO:tensorflow:global_step/sec: 4.4804 +INFO:tensorflow:step = 69201, loss = 0.122545, precision = 1.0 (22.319 sec) +Saved checkpoint after 177 epoch(s) to data/resnet56/checkpoints/00177... +INFO:tensorflow:global_step/sec: 4.34532 +INFO:tensorflow:step = 69301, loss = 0.117356, precision = 1.0 (23.013 sec) +INFO:tensorflow:global_step/sec: 4.48588 +INFO:tensorflow:step = 69401, loss = 0.119963, precision = 1.0 (22.292 sec) +INFO:tensorflow:global_step/sec: 4.47882 +INFO:tensorflow:step = 69501, loss = 0.120702, precision = 1.0 (22.327 sec) +Saved checkpoint after 178 epoch(s) to data/resnet56/checkpoints/00178... +INFO:tensorflow:global_step/sec: 4.3371 +INFO:tensorflow:step = 69601, loss = 0.118072, precision = 1.0 (23.057 sec) +INFO:tensorflow:global_step/sec: 4.48256 +INFO:tensorflow:step = 69701, loss = 0.123653, precision = 1.0 (22.309 sec) +INFO:tensorflow:global_step/sec: 4.48435 +INFO:tensorflow:step = 69801, loss = 0.124796, precision = 1.0 (22.300 sec) +INFO:tensorflow:global_step/sec: 4.48377 +INFO:tensorflow:step = 69901, loss = 0.1224, precision = 1.0 (22.303 sec) +Saved checkpoint after 179 epoch(s) to data/resnet56/checkpoints/00179... +INFO:tensorflow:global_step/sec: 4.3431 +INFO:tensorflow:step = 70001, loss = 0.11705, precision = 1.0 (23.025 sec) +INFO:tensorflow:global_step/sec: 4.47824 +INFO:tensorflow:step = 70101, loss = 0.119858, precision = 1.0 (22.330 sec) +INFO:tensorflow:global_step/sec: 4.48117 +INFO:tensorflow:step = 70201, loss = 0.116001, precision = 1.0 (22.316 sec) +INFO:tensorflow:global_step/sec: 4.47945 +INFO:tensorflow:step = 70301, loss = 0.11881, precision = 1.0 (22.324 sec) +Saved checkpoint after 180 epoch(s) to data/resnet56/checkpoints/00180... +INFO:tensorflow:global_step/sec: 4.34097 +INFO:tensorflow:step = 70401, loss = 0.119357, precision = 1.0 (23.036 sec) +INFO:tensorflow:global_step/sec: 4.48274 +INFO:tensorflow:step = 70501, loss = 0.127663, precision = 0.992188 (22.308 sec) +INFO:tensorflow:global_step/sec: 4.47991 +INFO:tensorflow:step = 70601, loss = 0.117966, precision = 1.0 (22.322 sec) +INFO:tensorflow:global_step/sec: 4.48177 +INFO:tensorflow:step = 70701, loss = 0.117792, precision = 1.0 (22.313 sec) +Saved checkpoint after 181 epoch(s) to data/resnet56/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1k80_gc/resnet164_b_train.log b/tensorflow/CIFAR10/logs/1k80_gc/resnet164_b_train.log new file mode 100644 index 0000000..2f5db18 --- /dev/null +++ b/tensorflow/CIFAR10/logs/1k80_gc/resnet164_b_train.log @@ -0,0 +1,2222 @@ +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/1.69m params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (256x10, 2.56k/2.56k params) + logit/biases (10, 10/10 params) + unit_1_0/common_bn_relu/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (1x1x16x16, 256/256 params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub3/bn3/beta (16, 16/16 params) + unit_1_0/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_0/sub_add/project/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_1/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_1/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub3/bn3/beta (16, 16/16 params) + unit_1_1/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_10/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_10/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_10/sub2/bn2/beta (16, 16/16 params) + unit_1_10/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/sub3/bn3/beta (16, 16/16 params) + unit_1_10/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_11/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_11/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_11/sub2/bn2/beta (16, 16/16 params) + unit_1_11/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/sub3/bn3/beta (16, 16/16 params) + unit_1_11/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_12/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_12/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_12/sub2/bn2/beta (16, 16/16 params) + unit_1_12/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/sub3/bn3/beta (16, 16/16 params) + unit_1_12/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_13/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_13/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_13/sub2/bn2/beta (16, 16/16 params) + unit_1_13/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/sub3/bn3/beta (16, 16/16 params) + unit_1_13/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_14/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_14/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_14/sub2/bn2/beta (16, 16/16 params) + unit_1_14/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/sub3/bn3/beta (16, 16/16 params) + unit_1_14/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_15/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_15/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_15/sub2/bn2/beta (16, 16/16 params) + unit_1_15/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/sub3/bn3/beta (16, 16/16 params) + unit_1_15/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_16/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_16/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_16/sub2/bn2/beta (16, 16/16 params) + unit_1_16/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/sub3/bn3/beta (16, 16/16 params) + unit_1_16/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_17/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_17/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_17/sub2/bn2/beta (16, 16/16 params) + unit_1_17/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/sub3/bn3/beta (16, 16/16 params) + unit_1_17/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_2/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_2/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub3/bn3/beta (16, 16/16 params) + unit_1_2/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_3/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_3/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub3/bn3/beta (16, 16/16 params) + unit_1_3/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_4/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_4/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub3/bn3/beta (16, 16/16 params) + unit_1_4/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_5/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_5/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub3/bn3/beta (16, 16/16 params) + unit_1_5/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_6/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_6/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub3/bn3/beta (16, 16/16 params) + unit_1_6/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_7/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_7/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub3/bn3/beta (16, 16/16 params) + unit_1_7/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_8/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_8/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub3/bn3/beta (16, 16/16 params) + unit_1_8/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_9/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_9/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_9/sub2/bn2/beta (16, 16/16 params) + unit_1_9/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/sub3/bn3/beta (16, 16/16 params) + unit_1_9/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_2_0/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_2_0/sub1/conv1/DW (1x1x64x32, 2.05k/2.05k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_0/sub3/bn3/beta (32, 32/32 params) + unit_2_0/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_0/sub_add/project/DW (1x1x64x128, 8.19k/8.19k params) + unit_2_1/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_1/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub3/bn3/beta (32, 32/32 params) + unit_2_1/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_10/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_10/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_10/sub2/bn2/beta (32, 32/32 params) + unit_2_10/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/sub3/bn3/beta (32, 32/32 params) + unit_2_10/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_11/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_11/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_11/sub2/bn2/beta (32, 32/32 params) + unit_2_11/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/sub3/bn3/beta (32, 32/32 params) + unit_2_11/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_12/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_12/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_12/sub2/bn2/beta (32, 32/32 params) + unit_2_12/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/sub3/bn3/beta (32, 32/32 params) + unit_2_12/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_13/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_13/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_13/sub2/bn2/beta (32, 32/32 params) + unit_2_13/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/sub3/bn3/beta (32, 32/32 params) + unit_2_13/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_14/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_14/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_14/sub2/bn2/beta (32, 32/32 params) + unit_2_14/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/sub3/bn3/beta (32, 32/32 params) + unit_2_14/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_15/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_15/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_15/sub2/bn2/beta (32, 32/32 params) + unit_2_15/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/sub3/bn3/beta (32, 32/32 params) + unit_2_15/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_16/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_16/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_16/sub2/bn2/beta (32, 32/32 params) + unit_2_16/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/sub3/bn3/beta (32, 32/32 params) + unit_2_16/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_17/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_17/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_17/sub2/bn2/beta (32, 32/32 params) + unit_2_17/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/sub3/bn3/beta (32, 32/32 params) + unit_2_17/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_2/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_2/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub3/bn3/beta (32, 32/32 params) + unit_2_2/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_3/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_3/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub3/bn3/beta (32, 32/32 params) + unit_2_3/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_4/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_4/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub3/bn3/beta (32, 32/32 params) + unit_2_4/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_5/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_5/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub3/bn3/beta (32, 32/32 params) + unit_2_5/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_6/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_6/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub3/bn3/beta (32, 32/32 params) + unit_2_6/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_7/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_7/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub3/bn3/beta (32, 32/32 params) + unit_2_7/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_8/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_8/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub3/bn3/beta (32, 32/32 params) + unit_2_8/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_9/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_9/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_9/sub2/bn2/beta (32, 32/32 params) + unit_2_9/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/sub3/bn3/beta (32, 32/32 params) + unit_2_9/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_3_0/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_3_0/sub1/conv1/DW (1x1x128x64, 8.19k/8.19k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_0/sub3/bn3/beta (64, 64/64 params) + unit_3_0/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_0/sub_add/project/DW (1x1x128x256, 32.77k/32.77k params) + unit_3_1/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_1/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub3/bn3/beta (64, 64/64 params) + unit_3_1/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_10/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_10/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_10/sub2/bn2/beta (64, 64/64 params) + unit_3_10/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/sub3/bn3/beta (64, 64/64 params) + unit_3_10/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_11/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_11/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_11/sub2/bn2/beta (64, 64/64 params) + unit_3_11/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/sub3/bn3/beta (64, 64/64 params) + unit_3_11/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_12/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_12/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_12/sub2/bn2/beta (64, 64/64 params) + unit_3_12/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/sub3/bn3/beta (64, 64/64 params) + unit_3_12/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_13/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_13/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_13/sub2/bn2/beta (64, 64/64 params) + unit_3_13/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/sub3/bn3/beta (64, 64/64 params) + unit_3_13/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_14/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_14/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_14/sub2/bn2/beta (64, 64/64 params) + unit_3_14/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/sub3/bn3/beta (64, 64/64 params) + unit_3_14/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_15/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_15/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_15/sub2/bn2/beta (64, 64/64 params) + unit_3_15/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/sub3/bn3/beta (64, 64/64 params) + unit_3_15/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_16/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_16/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_16/sub2/bn2/beta (64, 64/64 params) + unit_3_16/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/sub3/bn3/beta (64, 64/64 params) + unit_3_16/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_17/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_17/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_17/sub2/bn2/beta (64, 64/64 params) + unit_3_17/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/sub3/bn3/beta (64, 64/64 params) + unit_3_17/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_2/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_2/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub3/bn3/beta (64, 64/64 params) + unit_3_2/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_3/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_3/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub3/bn3/beta (64, 64/64 params) + unit_3_3/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_4/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_4/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub3/bn3/beta (64, 64/64 params) + unit_3_4/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_5/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_5/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub3/bn3/beta (64, 64/64 params) + unit_3_5/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_6/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_6/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub3/bn3/beta (64, 64/64 params) + unit_3_6/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_7/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_7/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub3/bn3/beta (64, 64/64 params) + unit_3_7/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_8/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_8/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub3/bn3/beta (64, 64/64 params) + unit_3_8/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_9/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_9/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_9/sub2/bn2/beta (64, 64/64 params) + unit_3_9/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/sub3/bn3/beta (64, 64/64 params) + unit_3_9/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_last/final_bn/beta (256, 256/256 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/62.59b flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub_add/project/Conv2D (536.87m/536.87m flops) + unit_2_0/sub_add/project/Conv2D (536.87m/536.87m flops) + unit_3_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_0/sub_add/project/Conv2D (268.44m/268.44m flops) + unit_1_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_0/sub1/conv1/Conv2D (134.22m/134.22m flops) + unit_3_0/sub1/conv1/Conv2D (134.22m/134.22m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + unit_1_0/sub1/conv1/Conv2D (67.11m/67.11m flops) + logit/xw_plus_b (1.28k/656.64k flops) + logit/xw_plus_b/MatMul (655.36k/655.36k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (655.36k/655.36k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (655.36k/655.36k flops) + +======================End of Report========================== +2017-07-30 00:29:08.491549: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero +2017-07-30 00:29:08.492356: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla K80 +major: 3 minor: 7 memoryClockRate (GHz) 0.8235 +pciBusID 0000:00:04.0 +Total memory: 11.17GiB +Free memory: 11.09GiB +2017-07-30 00:29:08.492389: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-07-30 00:29:08.492402: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-07-30 00:29:08.492419: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:00:04.0) +2017-07-30 00:29:10.242314: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 00:29:10.242378: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 8 visible devices +2017-07-30 00:29:10.245133: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0xa1ac7b0 executing computations on platform Host. Devices: +2017-07-30 00:29:10.245162: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-07-30 00:29:10.246111: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 00:29:10.246156: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 8 visible devices +2017-07-30 00:29:10.247206: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0xb86e820 executing computations on platform CUDA. Devices: +2017-07-30 00:29:10.247228: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla K80, Compute Capability 3.7 +2017-07-30 00:29:11.558236: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:247] PoolAllocator: After 1152 get requests, put_count=1100 evicted_count=1000 eviction_rate=0.909091 and unsatisfied allocation rate=1 +2017-07-30 00:29:11.558328: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:259] Raising pool_size_limit_ from 100 to 110 +INFO:tensorflow:step = 1, loss = 8.40695, precision = 0.0625 +INFO:tensorflow:global_step/sec: 1.3357 +INFO:tensorflow:step = 101, loss = 7.91483, precision = 0.289062 (74.868 sec) +INFO:tensorflow:global_step/sec: 1.37166 +INFO:tensorflow:step = 201, loss = 7.6816, precision = 0.3125 (72.905 sec) +INFO:tensorflow:global_step/sec: 1.37164 +INFO:tensorflow:step = 301, loss = 7.36041, precision = 0.398438 (72.905 sec) +total_params: 1691146 +Saved checkpoint after 1 epoch(s) to ../data/resnet164/checkpoints/00001... +INFO:tensorflow:global_step/sec: 1.3181 +INFO:tensorflow:step = 401, loss = 7.70667, precision = 0.34375 (75.867 sec) +INFO:tensorflow:global_step/sec: 1.37262 +INFO:tensorflow:step = 501, loss = 6.69559, precision = 0.460938 (72.853 sec) +INFO:tensorflow:global_step/sec: 1.37057 +INFO:tensorflow:step = 601, loss = 6.08644, precision = 0.578125 (72.963 sec) +INFO:tensorflow:global_step/sec: 1.37022 +INFO:tensorflow:step = 701, loss = 5.52997, precision = 0.546875 (72.981 sec) +Saved checkpoint after 2 epoch(s) to ../data/resnet164/checkpoints/00002... +INFO:tensorflow:global_step/sec: 1.32412 +INFO:tensorflow:step = 801, loss = 5.1018, precision = 0.609375 (75.522 sec) +INFO:tensorflow:global_step/sec: 1.37053 +INFO:tensorflow:step = 901, loss = 4.54882, precision = 0.640625 (72.964 sec) +INFO:tensorflow:global_step/sec: 1.37016 +INFO:tensorflow:step = 1001, loss = 4.15433, precision = 0.703125 (72.984 sec) +INFO:tensorflow:global_step/sec: 1.37092 +INFO:tensorflow:step = 1101, loss = 3.78356, precision = 0.71875 (72.943 sec) +Saved checkpoint after 3 epoch(s) to ../data/resnet164/checkpoints/00003... +INFO:tensorflow:global_step/sec: 1.32766 +INFO:tensorflow:step = 1201, loss = 3.70718, precision = 0.640625 (75.321 sec) +INFO:tensorflow:global_step/sec: 1.37145 +INFO:tensorflow:step = 1301, loss = 3.35475, precision = 0.695312 (72.915 sec) +INFO:tensorflow:global_step/sec: 1.37169 +INFO:tensorflow:step = 1401, loss = 3.18919, precision = 0.648438 (72.903 sec) +INFO:tensorflow:global_step/sec: 1.37124 +INFO:tensorflow:step = 1501, loss = 2.81857, precision = 0.71875 (72.927 sec) +Saved checkpoint after 4 epoch(s) to ../data/resnet164/checkpoints/00004... +INFO:tensorflow:global_step/sec: 1.32568 +INFO:tensorflow:step = 1601, loss = 2.65146, precision = 0.765625 (75.433 sec) +INFO:tensorflow:global_step/sec: 1.37113 +INFO:tensorflow:step = 1701, loss = 2.48633, precision = 0.75 (72.932 sec) +INFO:tensorflow:global_step/sec: 1.37207 +INFO:tensorflow:step = 1801, loss = 2.17718, precision = 0.789062 (72.883 sec) +INFO:tensorflow:global_step/sec: 1.37099 +INFO:tensorflow:step = 1901, loss = 2.1845, precision = 0.742188 (72.940 sec) +Saved checkpoint after 5 epoch(s) to ../data/resnet164/checkpoints/00005... +INFO:tensorflow:global_step/sec: 1.32379 +INFO:tensorflow:step = 2001, loss = 1.86284, precision = 0.828125 (75.541 sec) +INFO:tensorflow:global_step/sec: 1.36936 +INFO:tensorflow:step = 2101, loss = 1.93679, precision = 0.75 (73.026 sec) +INFO:tensorflow:global_step/sec: 1.37027 +INFO:tensorflow:step = 2201, loss = 1.89831, precision = 0.765625 (72.978 sec) +INFO:tensorflow:global_step/sec: 1.37023 +INFO:tensorflow:step = 2301, loss = 1.85456, precision = 0.703125 (72.980 sec) +Saved checkpoint after 6 epoch(s) to ../data/resnet164/checkpoints/00006... +INFO:tensorflow:global_step/sec: 1.32682 +INFO:tensorflow:step = 2401, loss = 1.71237, precision = 0.742188 (75.368 sec) +INFO:tensorflow:global_step/sec: 1.36915 +INFO:tensorflow:step = 2501, loss = 1.58765, precision = 0.71875 (73.038 sec) +INFO:tensorflow:global_step/sec: 1.36986 +INFO:tensorflow:step = 2601, loss = 1.43923, precision = 0.78125 (73.001 sec) +INFO:tensorflow:global_step/sec: 1.36993 +INFO:tensorflow:step = 2701, loss = 1.45246, precision = 0.78125 (72.996 sec) +Saved checkpoint after 7 epoch(s) to ../data/resnet164/checkpoints/00007... +INFO:tensorflow:global_step/sec: 1.32641 +INFO:tensorflow:step = 2801, loss = 1.22975, precision = 0.835938 (75.391 sec) +INFO:tensorflow:global_step/sec: 1.36996 +INFO:tensorflow:step = 2901, loss = 1.33211, precision = 0.789062 (72.995 sec) +INFO:tensorflow:global_step/sec: 1.37045 +INFO:tensorflow:step = 3001, loss = 1.32043, precision = 0.804688 (72.969 sec) +INFO:tensorflow:global_step/sec: 1.37004 +INFO:tensorflow:step = 3101, loss = 1.23295, precision = 0.8125 (72.991 sec) +Saved checkpoint after 8 epoch(s) to ../data/resnet164/checkpoints/00008... +INFO:tensorflow:global_step/sec: 1.32897 +INFO:tensorflow:step = 3201, loss = 1.16917, precision = 0.8125 (75.246 sec) +INFO:tensorflow:global_step/sec: 1.36982 +INFO:tensorflow:step = 3301, loss = 1.32483, precision = 0.726562 (73.002 sec) +INFO:tensorflow:global_step/sec: 1.3705 +INFO:tensorflow:step = 3401, loss = 1.00311, precision = 0.84375 (72.966 sec) +INFO:tensorflow:global_step/sec: 1.37053 +INFO:tensorflow:step = 3501, loss = 0.993852, precision = 0.828125 (72.964 sec) +Saved checkpoint after 9 epoch(s) to ../data/resnet164/checkpoints/00009... +INFO:tensorflow:global_step/sec: 1.32078 +INFO:tensorflow:step = 3601, loss = 0.997532, precision = 0.828125 (75.713 sec) +INFO:tensorflow:global_step/sec: 1.37012 +INFO:tensorflow:step = 3701, loss = 1.17555, precision = 0.765625 (72.986 sec) +INFO:tensorflow:global_step/sec: 1.37081 +INFO:tensorflow:step = 3801, loss = 0.933248, precision = 0.875 (72.950 sec) +INFO:tensorflow:global_step/sec: 1.3705 +INFO:tensorflow:step = 3901, loss = 1.05485, precision = 0.8125 (72.966 sec) +Saved checkpoint after 10 epoch(s) to ../data/resnet164/checkpoints/00010... +INFO:tensorflow:global_step/sec: 1.32553 +INFO:tensorflow:step = 4001, loss = 0.842509, precision = 0.835938 (75.442 sec) +INFO:tensorflow:global_step/sec: 1.36958 +INFO:tensorflow:step = 4101, loss = 0.995981, precision = 0.796875 (73.015 sec) +INFO:tensorflow:global_step/sec: 1.37005 +INFO:tensorflow:step = 4201, loss = 0.975255, precision = 0.820312 (72.990 sec) +Saved checkpoint after 11 epoch(s) to ../data/resnet164/checkpoints/00011... +INFO:tensorflow:global_step/sec: 1.3256 +INFO:tensorflow:step = 4301, loss = 0.971976, precision = 0.789062 (75.437 sec) +INFO:tensorflow:global_step/sec: 1.36915 +INFO:tensorflow:step = 4401, loss = 0.870302, precision = 0.796875 (73.038 sec) +INFO:tensorflow:global_step/sec: 1.37027 +INFO:tensorflow:step = 4501, loss = 0.922395, precision = 0.796875 (72.978 sec) +INFO:tensorflow:global_step/sec: 1.37011 +INFO:tensorflow:step = 4601, loss = 0.967541, precision = 0.757812 (72.987 sec) +Saved checkpoint after 12 epoch(s) to ../data/resnet164/checkpoints/00012... +INFO:tensorflow:global_step/sec: 1.32664 +INFO:tensorflow:step = 4701, loss = 0.90638, precision = 0.828125 (75.378 sec) +INFO:tensorflow:global_step/sec: 1.37129 +INFO:tensorflow:step = 4801, loss = 0.779839, precision = 0.875 (72.924 sec) +INFO:tensorflow:global_step/sec: 1.37044 +INFO:tensorflow:step = 4901, loss = 0.835832, precision = 0.820312 (72.969 sec) +INFO:tensorflow:global_step/sec: 1.37109 +INFO:tensorflow:step = 5001, loss = 0.826647, precision = 0.84375 (72.935 sec) +Saved checkpoint after 13 epoch(s) to ../data/resnet164/checkpoints/00013... +INFO:tensorflow:global_step/sec: 1.32464 +INFO:tensorflow:step = 5101, loss = 0.793596, precision = 0.835938 (75.492 sec) +INFO:tensorflow:global_step/sec: 1.37108 +INFO:tensorflow:step = 5201, loss = 0.892478, precision = 0.828125 (72.935 sec) +INFO:tensorflow:global_step/sec: 1.37063 +INFO:tensorflow:step = 5301, loss = 0.988922, precision = 0.804688 (72.959 sec) +INFO:tensorflow:global_step/sec: 1.37065 +INFO:tensorflow:step = 5401, loss = 0.900933, precision = 0.835938 (72.959 sec) +Saved checkpoint after 14 epoch(s) to ../data/resnet164/checkpoints/00014... +INFO:tensorflow:global_step/sec: 1.32562 +INFO:tensorflow:step = 5501, loss = 0.788703, precision = 0.835938 (75.436 sec) +INFO:tensorflow:global_step/sec: 1.37088 +INFO:tensorflow:step = 5601, loss = 0.96695, precision = 0.757812 (72.946 sec) +INFO:tensorflow:global_step/sec: 1.37059 +INFO:tensorflow:step = 5701, loss = 0.864157, precision = 0.8125 (72.961 sec) +INFO:tensorflow:global_step/sec: 1.37079 +INFO:tensorflow:step = 5801, loss = 0.856458, precision = 0.859375 (72.951 sec) +Saved checkpoint after 15 epoch(s) to ../data/resnet164/checkpoints/00015... +INFO:tensorflow:global_step/sec: 1.32836 +INFO:tensorflow:step = 5901, loss = 0.78816, precision = 0.828125 (75.281 sec) +INFO:tensorflow:global_step/sec: 1.37141 +INFO:tensorflow:step = 6001, loss = 0.686418, precision = 0.882812 (72.918 sec) +INFO:tensorflow:global_step/sec: 1.37058 +INFO:tensorflow:step = 6101, loss = 0.723384, precision = 0.890625 (72.962 sec) +INFO:tensorflow:global_step/sec: 1.37035 +INFO:tensorflow:step = 6201, loss = 0.847929, precision = 0.820312 (72.974 sec) +Saved checkpoint after 16 epoch(s) to ../data/resnet164/checkpoints/00016... +INFO:tensorflow:global_step/sec: 1.32641 +INFO:tensorflow:step = 6301, loss = 0.787762, precision = 0.835938 (75.391 sec) +INFO:tensorflow:global_step/sec: 1.37099 +INFO:tensorflow:step = 6401, loss = 0.829899, precision = 0.859375 (72.940 sec) +INFO:tensorflow:global_step/sec: 1.37127 +INFO:tensorflow:step = 6501, loss = 0.795241, precision = 0.8125 (72.925 sec) +INFO:tensorflow:global_step/sec: 1.37164 +INFO:tensorflow:step = 6601, loss = 0.845229, precision = 0.835938 (72.905 sec) +Saved checkpoint after 17 epoch(s) to ../data/resnet164/checkpoints/00017... +INFO:tensorflow:global_step/sec: 1.32651 +INFO:tensorflow:step = 6701, loss = 0.859774, precision = 0.828125 (75.386 sec) +INFO:tensorflow:global_step/sec: 1.37035 +INFO:tensorflow:step = 6801, loss = 0.735282, precision = 0.859375 (72.974 sec) +INFO:tensorflow:global_step/sec: 1.37016 +INFO:tensorflow:step = 6901, loss = 0.825414, precision = 0.78125 (72.984 sec) +INFO:tensorflow:global_step/sec: 1.37051 +INFO:tensorflow:step = 7001, loss = 0.716614, precision = 0.851562 (72.965 sec) +Saved checkpoint after 18 epoch(s) to ../data/resnet164/checkpoints/00018... +INFO:tensorflow:global_step/sec: 1.32709 +INFO:tensorflow:step = 7101, loss = 0.725851, precision = 0.875 (75.353 sec) +INFO:tensorflow:global_step/sec: 1.37312 +INFO:tensorflow:step = 7201, loss = 0.740376, precision = 0.851562 (72.827 sec) +INFO:tensorflow:global_step/sec: 1.37246 +INFO:tensorflow:step = 7301, loss = 0.794508, precision = 0.851562 (72.862 sec) +INFO:tensorflow:global_step/sec: 1.37193 +INFO:tensorflow:step = 7401, loss = 0.659346, precision = 0.859375 (72.890 sec) +Saved checkpoint after 19 epoch(s) to ../data/resnet164/checkpoints/00019... +INFO:tensorflow:global_step/sec: 1.3259 +INFO:tensorflow:step = 7501, loss = 0.76372, precision = 0.851562 (75.421 sec) +INFO:tensorflow:global_step/sec: 1.37149 +INFO:tensorflow:step = 7601, loss = 0.847067, precision = 0.828125 (72.914 sec) +INFO:tensorflow:global_step/sec: 1.37199 +INFO:tensorflow:step = 7701, loss = 0.853, precision = 0.828125 (72.887 sec) +INFO:tensorflow:global_step/sec: 1.3725 +INFO:tensorflow:step = 7801, loss = 0.824415, precision = 0.804688 (72.860 sec) +Saved checkpoint after 20 epoch(s) to ../data/resnet164/checkpoints/00020... +INFO:tensorflow:global_step/sec: 1.3284 +INFO:tensorflow:step = 7901, loss = 0.621461, precision = 0.898438 (75.279 sec) +INFO:tensorflow:global_step/sec: 1.3716 +INFO:tensorflow:step = 8001, loss = 0.853694, precision = 0.8125 (72.908 sec) +INFO:tensorflow:global_step/sec: 1.36993 +INFO:tensorflow:step = 8101, loss = 0.669246, precision = 0.875 (72.996 sec) +INFO:tensorflow:global_step/sec: 1.36883 +INFO:tensorflow:step = 8201, loss = 0.765451, precision = 0.851562 (73.055 sec) +Saved checkpoint after 21 epoch(s) to ../data/resnet164/checkpoints/00021... +INFO:tensorflow:global_step/sec: 1.32737 +INFO:tensorflow:step = 8301, loss = 0.69459, precision = 0.875 (75.337 sec) +INFO:tensorflow:global_step/sec: 1.36767 +INFO:tensorflow:step = 8401, loss = 0.775848, precision = 0.898438 (73.117 sec) +INFO:tensorflow:global_step/sec: 1.36796 +INFO:tensorflow:step = 8501, loss = 0.788227, precision = 0.8125 (73.101 sec) +INFO:tensorflow:global_step/sec: 1.36783 +INFO:tensorflow:step = 8601, loss = 0.621537, precision = 0.898438 (73.108 sec) +Saved checkpoint after 22 epoch(s) to ../data/resnet164/checkpoints/00022... +INFO:tensorflow:global_step/sec: 1.32168 +INFO:tensorflow:step = 8701, loss = 0.732961, precision = 0.835938 (75.661 sec) +INFO:tensorflow:global_step/sec: 1.3667 +INFO:tensorflow:step = 8801, loss = 0.819157, precision = 0.78125 (73.169 sec) +INFO:tensorflow:global_step/sec: 1.36705 +INFO:tensorflow:step = 8901, loss = 0.685816, precision = 0.851562 (73.150 sec) +Saved checkpoint after 23 epoch(s) to ../data/resnet164/checkpoints/00023... +INFO:tensorflow:global_step/sec: 1.32173 +INFO:tensorflow:step = 9001, loss = 0.759089, precision = 0.859375 (75.659 sec) +INFO:tensorflow:global_step/sec: 1.36806 +INFO:tensorflow:step = 9101, loss = 0.73286, precision = 0.835938 (73.096 sec) +INFO:tensorflow:global_step/sec: 1.36704 +INFO:tensorflow:step = 9201, loss = 0.67288, precision = 0.882812 (73.151 sec) +INFO:tensorflow:global_step/sec: 1.36773 +INFO:tensorflow:step = 9301, loss = 0.748707, precision = 0.882812 (73.114 sec) +Saved checkpoint after 24 epoch(s) to ../data/resnet164/checkpoints/00024... +INFO:tensorflow:global_step/sec: 1.32179 +INFO:tensorflow:step = 9401, loss = 0.743169, precision = 0.820312 (75.655 sec) +INFO:tensorflow:global_step/sec: 1.36774 +INFO:tensorflow:step = 9501, loss = 0.964531, precision = 0.789062 (73.113 sec) +INFO:tensorflow:global_step/sec: 1.36703 +INFO:tensorflow:step = 9601, loss = 0.721257, precision = 0.835938 (73.152 sec) +INFO:tensorflow:global_step/sec: 1.36721 +INFO:tensorflow:step = 9701, loss = 0.765272, precision = 0.859375 (73.142 sec) +Saved checkpoint after 25 epoch(s) to ../data/resnet164/checkpoints/00025... +INFO:tensorflow:global_step/sec: 1.31943 +INFO:tensorflow:step = 9801, loss = 0.571354, precision = 0.914062 (75.790 sec) +INFO:tensorflow:global_step/sec: 1.3689 +INFO:tensorflow:step = 9901, loss = 0.758683, precision = 0.859375 (73.051 sec) +INFO:tensorflow:global_step/sec: 1.3696 +INFO:tensorflow:step = 10001, loss = 0.750885, precision = 0.84375 (73.014 sec) +INFO:tensorflow:global_step/sec: 1.36833 +INFO:tensorflow:step = 10101, loss = 0.667607, precision = 0.890625 (73.082 sec) +Saved checkpoint after 26 epoch(s) to ../data/resnet164/checkpoints/00026... +INFO:tensorflow:global_step/sec: 1.3226 +INFO:tensorflow:step = 10201, loss = 0.768349, precision = 0.84375 (75.609 sec) +INFO:tensorflow:global_step/sec: 1.36772 +INFO:tensorflow:step = 10301, loss = 0.645212, precision = 0.882812 (73.115 sec) +INFO:tensorflow:global_step/sec: 1.36736 +INFO:tensorflow:step = 10401, loss = 0.714994, precision = 0.859375 (73.134 sec) +INFO:tensorflow:global_step/sec: 1.36801 +INFO:tensorflow:step = 10501, loss = 0.756118, precision = 0.835938 (73.099 sec) +Saved checkpoint after 27 epoch(s) to ../data/resnet164/checkpoints/00027... +INFO:tensorflow:global_step/sec: 1.32744 +INFO:tensorflow:step = 10601, loss = 0.756358, precision = 0.835938 (75.333 sec) +INFO:tensorflow:global_step/sec: 1.36849 +INFO:tensorflow:step = 10701, loss = 0.653143, precision = 0.898438 (73.073 sec) +INFO:tensorflow:global_step/sec: 1.36697 +INFO:tensorflow:step = 10801, loss = 0.747757, precision = 0.835938 (73.155 sec) +INFO:tensorflow:global_step/sec: 1.36741 +INFO:tensorflow:step = 10901, loss = 0.669361, precision = 0.859375 (73.131 sec) +Saved checkpoint after 28 epoch(s) to ../data/resnet164/checkpoints/00028... +INFO:tensorflow:global_step/sec: 1.3223 +INFO:tensorflow:step = 11001, loss = 0.878276, precision = 0.796875 (75.626 sec) +INFO:tensorflow:global_step/sec: 1.36837 +INFO:tensorflow:step = 11101, loss = 0.713048, precision = 0.867188 (73.079 sec) +INFO:tensorflow:global_step/sec: 1.36736 +INFO:tensorflow:step = 11201, loss = 0.746622, precision = 0.851562 (73.134 sec) +INFO:tensorflow:global_step/sec: 1.3678 +INFO:tensorflow:step = 11301, loss = 0.709027, precision = 0.882812 (73.110 sec) +Saved checkpoint after 29 epoch(s) to ../data/resnet164/checkpoints/00029... +INFO:tensorflow:global_step/sec: 1.31623 +INFO:tensorflow:step = 11401, loss = 0.682184, precision = 0.84375 (75.974 sec) +INFO:tensorflow:global_step/sec: 1.36921 +INFO:tensorflow:step = 11501, loss = 0.707811, precision = 0.867188 (73.035 sec) +INFO:tensorflow:global_step/sec: 1.36768 +INFO:tensorflow:step = 11601, loss = 0.759826, precision = 0.859375 (73.117 sec) +INFO:tensorflow:global_step/sec: 1.36708 +INFO:tensorflow:step = 11701, loss = 0.790435, precision = 0.8125 (73.149 sec) +Saved checkpoint after 30 epoch(s) to ../data/resnet164/checkpoints/00030... +INFO:tensorflow:global_step/sec: 1.32141 +INFO:tensorflow:step = 11801, loss = 0.808823, precision = 0.796875 (75.677 sec) +INFO:tensorflow:global_step/sec: 1.36769 +INFO:tensorflow:step = 11901, loss = 0.610608, precision = 0.882812 (73.116 sec) +INFO:tensorflow:global_step/sec: 1.36769 +INFO:tensorflow:step = 12001, loss = 0.678239, precision = 0.890625 (73.116 sec) +INFO:tensorflow:global_step/sec: 1.36753 +INFO:tensorflow:step = 12101, loss = 0.665997, precision = 0.882812 (73.124 sec) +Saved checkpoint after 31 epoch(s) to ../data/resnet164/checkpoints/00031... +INFO:tensorflow:global_step/sec: 1.32411 +INFO:tensorflow:step = 12201, loss = 0.726527, precision = 0.867188 (75.523 sec) +INFO:tensorflow:global_step/sec: 1.36807 +INFO:tensorflow:step = 12301, loss = 0.775641, precision = 0.859375 (73.095 sec) +INFO:tensorflow:global_step/sec: 1.36772 +INFO:tensorflow:step = 12401, loss = 0.700809, precision = 0.867188 (73.114 sec) +INFO:tensorflow:global_step/sec: 1.36734 +INFO:tensorflow:step = 12501, loss = 0.700601, precision = 0.867188 (73.135 sec) +Saved checkpoint after 32 epoch(s) to ../data/resnet164/checkpoints/00032... +INFO:tensorflow:global_step/sec: 1.32158 +INFO:tensorflow:step = 12601, loss = 0.667879, precision = 0.882812 (75.667 sec) +INFO:tensorflow:global_step/sec: 1.36825 +INFO:tensorflow:step = 12701, loss = 0.735662, precision = 0.851562 (73.086 sec) +INFO:tensorflow:global_step/sec: 1.36861 +INFO:tensorflow:step = 12801, loss = 0.699526, precision = 0.851562 (73.067 sec) +INFO:tensorflow:global_step/sec: 1.36782 +INFO:tensorflow:step = 12901, loss = 0.628842, precision = 0.90625 (73.109 sec) +Saved checkpoint after 33 epoch(s) to ../data/resnet164/checkpoints/00033... +INFO:tensorflow:global_step/sec: 1.32075 +INFO:tensorflow:step = 13001, loss = 0.707931, precision = 0.882812 (75.715 sec) +INFO:tensorflow:global_step/sec: 1.36787 +INFO:tensorflow:step = 13101, loss = 0.732072, precision = 0.875 (73.106 sec) +INFO:tensorflow:global_step/sec: 1.36883 +INFO:tensorflow:step = 13201, loss = 0.717937, precision = 0.84375 (73.055 sec) +Saved checkpoint after 34 epoch(s) to ../data/resnet164/checkpoints/00034... +INFO:tensorflow:global_step/sec: 1.32243 +INFO:tensorflow:step = 13301, loss = 0.638171, precision = 0.898438 (75.618 sec) +INFO:tensorflow:global_step/sec: 1.3681 +INFO:tensorflow:step = 13401, loss = 0.792747, precision = 0.859375 (73.094 sec) +INFO:tensorflow:global_step/sec: 1.36744 +INFO:tensorflow:step = 13501, loss = 0.708341, precision = 0.875 (73.129 sec) +INFO:tensorflow:global_step/sec: 1.36841 +INFO:tensorflow:step = 13601, loss = 0.718572, precision = 0.882812 (73.078 sec) +Saved checkpoint after 35 epoch(s) to ../data/resnet164/checkpoints/00035... +INFO:tensorflow:global_step/sec: 1.32161 +INFO:tensorflow:step = 13701, loss = 0.624056, precision = 0.914062 (75.665 sec) +INFO:tensorflow:global_step/sec: 1.36815 +INFO:tensorflow:step = 13801, loss = 0.707361, precision = 0.867188 (73.091 sec) +INFO:tensorflow:global_step/sec: 1.36738 +INFO:tensorflow:step = 13901, loss = 0.749224, precision = 0.828125 (73.132 sec) +INFO:tensorflow:global_step/sec: 1.369 +INFO:tensorflow:step = 14001, loss = 0.756292, precision = 0.875 (73.046 sec) +Saved checkpoint after 36 epoch(s) to ../data/resnet164/checkpoints/00036... +INFO:tensorflow:global_step/sec: 1.32208 +INFO:tensorflow:step = 14101, loss = 0.712808, precision = 0.875 (75.638 sec) +INFO:tensorflow:global_step/sec: 1.36801 +INFO:tensorflow:step = 14201, loss = 0.816605, precision = 0.828125 (73.099 sec) +INFO:tensorflow:global_step/sec: 1.36876 +INFO:tensorflow:step = 14301, loss = 0.752767, precision = 0.882812 (73.059 sec) +INFO:tensorflow:global_step/sec: 1.36878 +INFO:tensorflow:step = 14401, loss = 0.861724, precision = 0.804688 (73.058 sec) +Saved checkpoint after 37 epoch(s) to ../data/resnet164/checkpoints/00037... +INFO:tensorflow:global_step/sec: 1.32305 +INFO:tensorflow:step = 14501, loss = 0.735637, precision = 0.851562 (75.583 sec) +INFO:tensorflow:global_step/sec: 1.36785 +INFO:tensorflow:step = 14601, loss = 0.70498, precision = 0.84375 (73.107 sec) +INFO:tensorflow:global_step/sec: 1.36759 +INFO:tensorflow:step = 14701, loss = 0.73935, precision = 0.867188 (73.121 sec) +INFO:tensorflow:global_step/sec: 1.3685 +INFO:tensorflow:step = 14801, loss = 0.692426, precision = 0.867188 (73.073 sec) +Saved checkpoint after 38 epoch(s) to ../data/resnet164/checkpoints/00038... +INFO:tensorflow:global_step/sec: 1.32158 +INFO:tensorflow:step = 14901, loss = 0.597915, precision = 0.90625 (75.667 sec) +INFO:tensorflow:global_step/sec: 1.36789 +INFO:tensorflow:step = 15001, loss = 0.557536, precision = 0.921875 (73.105 sec) +INFO:tensorflow:global_step/sec: 1.36772 +INFO:tensorflow:step = 15101, loss = 0.625316, precision = 0.921875 (73.115 sec) +INFO:tensorflow:global_step/sec: 1.36874 +INFO:tensorflow:step = 15201, loss = 0.828637, precision = 0.835938 (73.060 sec) +Saved checkpoint after 39 epoch(s) to ../data/resnet164/checkpoints/00039... +INFO:tensorflow:global_step/sec: 1.31785 +INFO:tensorflow:step = 15301, loss = 0.682034, precision = 0.875 (75.881 sec) +INFO:tensorflow:global_step/sec: 1.3675 +INFO:tensorflow:step = 15401, loss = 0.6394, precision = 0.898438 (73.126 sec) +INFO:tensorflow:global_step/sec: 1.36753 +INFO:tensorflow:step = 15501, loss = 0.644359, precision = 0.90625 (73.124 sec) +INFO:tensorflow:global_step/sec: 1.36901 +INFO:tensorflow:step = 15601, loss = 0.601404, precision = 0.914062 (73.045 sec) +Saved checkpoint after 40 epoch(s) to ../data/resnet164/checkpoints/00040... +INFO:tensorflow:global_step/sec: 1.32891 +INFO:tensorflow:step = 15701, loss = 0.767928, precision = 0.835938 (75.250 sec) +INFO:tensorflow:global_step/sec: 1.36934 +INFO:tensorflow:step = 15801, loss = 0.625963, precision = 0.898438 (73.028 sec) +INFO:tensorflow:global_step/sec: 1.36773 +INFO:tensorflow:step = 15901, loss = 0.546079, precision = 0.921875 (73.114 sec) +INFO:tensorflow:global_step/sec: 1.36799 +INFO:tensorflow:step = 16001, loss = 0.651166, precision = 0.882812 (73.100 sec) +Saved checkpoint after 41 epoch(s) to ../data/resnet164/checkpoints/00041... +INFO:tensorflow:global_step/sec: 1.32212 +INFO:tensorflow:step = 16101, loss = 0.723789, precision = 0.828125 (75.636 sec) +INFO:tensorflow:global_step/sec: 1.36821 +INFO:tensorflow:step = 16201, loss = 0.705243, precision = 0.882812 (73.088 sec) +INFO:tensorflow:global_step/sec: 1.36803 +INFO:tensorflow:step = 16301, loss = 0.622612, precision = 0.898438 (73.098 sec) +INFO:tensorflow:global_step/sec: 1.36961 +INFO:tensorflow:step = 16401, loss = 0.646639, precision = 0.890625 (73.014 sec) +Saved checkpoint after 42 epoch(s) to ../data/resnet164/checkpoints/00042... +INFO:tensorflow:global_step/sec: 1.32729 +INFO:tensorflow:step = 16501, loss = 0.741348, precision = 0.851562 (75.341 sec) +INFO:tensorflow:global_step/sec: 1.36815 +INFO:tensorflow:step = 16601, loss = 0.639543, precision = 0.898438 (73.091 sec) +INFO:tensorflow:global_step/sec: 1.36779 +INFO:tensorflow:step = 16701, loss = 0.649675, precision = 0.890625 (73.110 sec) +INFO:tensorflow:global_step/sec: 1.36902 +INFO:tensorflow:step = 16801, loss = 0.716072, precision = 0.859375 (73.045 sec) +Saved checkpoint after 43 epoch(s) to ../data/resnet164/checkpoints/00043... +INFO:tensorflow:global_step/sec: 1.32075 +INFO:tensorflow:step = 16901, loss = 0.705193, precision = 0.890625 (75.714 sec) +INFO:tensorflow:global_step/sec: 1.36792 +INFO:tensorflow:step = 17001, loss = 0.671323, precision = 0.890625 (73.104 sec) +INFO:tensorflow:global_step/sec: 1.36759 +INFO:tensorflow:step = 17101, loss = 0.803898, precision = 0.851562 (73.121 sec) +INFO:tensorflow:global_step/sec: 1.36865 +INFO:tensorflow:step = 17201, loss = 0.533568, precision = 0.945312 (73.065 sec) +Saved checkpoint after 44 epoch(s) to ../data/resnet164/checkpoints/00044... +INFO:tensorflow:global_step/sec: 1.32015 +INFO:tensorflow:step = 17301, loss = 0.718161, precision = 0.875 (75.749 sec) +INFO:tensorflow:global_step/sec: 1.36806 +INFO:tensorflow:step = 17401, loss = 0.681598, precision = 0.867188 (73.096 sec) +INFO:tensorflow:global_step/sec: 1.36821 +INFO:tensorflow:step = 17501, loss = 0.573041, precision = 0.921875 (73.088 sec) +Saved checkpoint after 45 epoch(s) to ../data/resnet164/checkpoints/00045... +INFO:tensorflow:global_step/sec: 1.32374 +INFO:tensorflow:step = 17601, loss = 0.658535, precision = 0.898438 (75.544 sec) +INFO:tensorflow:global_step/sec: 1.36811 +INFO:tensorflow:step = 17701, loss = 0.741184, precision = 0.890625 (73.093 sec) +INFO:tensorflow:global_step/sec: 1.3686 +INFO:tensorflow:step = 17801, loss = 0.753713, precision = 0.890625 (73.067 sec) +INFO:tensorflow:global_step/sec: 1.36813 +INFO:tensorflow:step = 17901, loss = 0.747369, precision = 0.859375 (73.093 sec) +Saved checkpoint after 46 epoch(s) to ../data/resnet164/checkpoints/00046... +INFO:tensorflow:global_step/sec: 1.3275 +INFO:tensorflow:step = 18001, loss = 0.654825, precision = 0.882812 (75.329 sec) +INFO:tensorflow:global_step/sec: 1.36888 +INFO:tensorflow:step = 18101, loss = 0.658791, precision = 0.875 (73.052 sec) +INFO:tensorflow:global_step/sec: 1.36724 +INFO:tensorflow:step = 18201, loss = 0.665614, precision = 0.898438 (73.140 sec) +INFO:tensorflow:global_step/sec: 1.36713 +INFO:tensorflow:step = 18301, loss = 0.675275, precision = 0.875 (73.146 sec) +Saved checkpoint after 47 epoch(s) to ../data/resnet164/checkpoints/00047... +INFO:tensorflow:global_step/sec: 1.32763 +INFO:tensorflow:step = 18401, loss = 0.626043, precision = 0.875 (75.322 sec) +INFO:tensorflow:global_step/sec: 1.3695 +INFO:tensorflow:step = 18501, loss = 0.679343, precision = 0.867188 (73.019 sec) +INFO:tensorflow:global_step/sec: 1.3688 +INFO:tensorflow:step = 18601, loss = 0.712253, precision = 0.859375 (73.057 sec) +INFO:tensorflow:global_step/sec: 1.36889 +INFO:tensorflow:step = 18701, loss = 0.73903, precision = 0.859375 (73.052 sec) +Saved checkpoint after 48 epoch(s) to ../data/resnet164/checkpoints/00048... +INFO:tensorflow:global_step/sec: 1.32293 +INFO:tensorflow:step = 18801, loss = 0.657232, precision = 0.859375 (75.590 sec) +INFO:tensorflow:global_step/sec: 1.37075 +INFO:tensorflow:step = 18901, loss = 0.772407, precision = 0.859375 (72.953 sec) +INFO:tensorflow:global_step/sec: 1.36877 +INFO:tensorflow:step = 19001, loss = 0.624692, precision = 0.890625 (73.058 sec) +INFO:tensorflow:global_step/sec: 1.37116 +INFO:tensorflow:step = 19101, loss = 0.722398, precision = 0.84375 (72.931 sec) +Saved checkpoint after 49 epoch(s) to ../data/resnet164/checkpoints/00049... +INFO:tensorflow:global_step/sec: 1.31685 +INFO:tensorflow:step = 19201, loss = 0.753151, precision = 0.882812 (75.939 sec) +INFO:tensorflow:global_step/sec: 1.36964 +INFO:tensorflow:step = 19301, loss = 0.64368, precision = 0.921875 (73.012 sec) +INFO:tensorflow:global_step/sec: 1.36959 +INFO:tensorflow:step = 19401, loss = 0.66534, precision = 0.882812 (73.015 sec) +INFO:tensorflow:global_step/sec: 1.36875 +INFO:tensorflow:step = 19501, loss = 0.567771, precision = 0.929688 (73.059 sec) +Saved checkpoint after 50 epoch(s) to ../data/resnet164/checkpoints/00050... +INFO:tensorflow:global_step/sec: 1.32241 +INFO:tensorflow:step = 19601, loss = 0.696229, precision = 0.882812 (75.620 sec) +INFO:tensorflow:global_step/sec: 1.36913 +INFO:tensorflow:step = 19701, loss = 0.6483, precision = 0.890625 (73.039 sec) +INFO:tensorflow:global_step/sec: 1.36842 +INFO:tensorflow:step = 19801, loss = 0.777489, precision = 0.851562 (73.077 sec) +INFO:tensorflow:global_step/sec: 1.36865 +INFO:tensorflow:step = 19901, loss = 0.740897, precision = 0.867188 (73.064 sec) +Saved checkpoint after 51 epoch(s) to ../data/resnet164/checkpoints/00051... +INFO:tensorflow:global_step/sec: 1.32179 +INFO:tensorflow:step = 20001, loss = 0.611825, precision = 0.914062 (75.655 sec) +INFO:tensorflow:global_step/sec: 1.36991 +INFO:tensorflow:step = 20101, loss = 0.697011, precision = 0.867188 (72.997 sec) +INFO:tensorflow:global_step/sec: 1.37001 +INFO:tensorflow:step = 20201, loss = 0.656831, precision = 0.890625 (72.992 sec) +INFO:tensorflow:global_step/sec: 1.3685 +INFO:tensorflow:step = 20301, loss = 0.718254, precision = 0.84375 (73.073 sec) +Saved checkpoint after 52 epoch(s) to ../data/resnet164/checkpoints/00052... +INFO:tensorflow:global_step/sec: 1.32707 +INFO:tensorflow:step = 20401, loss = 0.800646, precision = 0.84375 (75.354 sec) +INFO:tensorflow:global_step/sec: 1.36955 +INFO:tensorflow:step = 20501, loss = 0.586259, precision = 0.921875 (73.016 sec) +INFO:tensorflow:global_step/sec: 1.36942 +INFO:tensorflow:step = 20601, loss = 0.616532, precision = 0.929688 (73.024 sec) +INFO:tensorflow:global_step/sec: 1.36817 +INFO:tensorflow:step = 20701, loss = 0.576592, precision = 0.921875 (73.090 sec) +Saved checkpoint after 53 epoch(s) to ../data/resnet164/checkpoints/00053... +INFO:tensorflow:global_step/sec: 1.32231 +INFO:tensorflow:step = 20801, loss = 0.625493, precision = 0.90625 (75.625 sec) +INFO:tensorflow:global_step/sec: 1.36889 +INFO:tensorflow:step = 20901, loss = 0.709452, precision = 0.882812 (73.052 sec) +INFO:tensorflow:global_step/sec: 1.36814 +INFO:tensorflow:step = 21001, loss = 0.58369, precision = 0.929688 (73.092 sec) +INFO:tensorflow:global_step/sec: 1.36772 +INFO:tensorflow:step = 21101, loss = 0.689312, precision = 0.882812 (73.114 sec) +Saved checkpoint after 54 epoch(s) to ../data/resnet164/checkpoints/00054... +INFO:tensorflow:global_step/sec: 1.32222 +INFO:tensorflow:step = 21201, loss = 0.707705, precision = 0.851562 (75.631 sec) +INFO:tensorflow:global_step/sec: 1.3689 +INFO:tensorflow:step = 21301, loss = 0.68219, precision = 0.859375 (73.051 sec) +INFO:tensorflow:global_step/sec: 1.37002 +INFO:tensorflow:step = 21401, loss = 0.676659, precision = 0.90625 (72.992 sec) +INFO:tensorflow:global_step/sec: 1.36861 +INFO:tensorflow:step = 21501, loss = 0.557692, precision = 0.929688 (73.066 sec) +Saved checkpoint after 55 epoch(s) to ../data/resnet164/checkpoints/00055... +INFO:tensorflow:global_step/sec: 1.32148 +INFO:tensorflow:step = 21601, loss = 0.678778, precision = 0.867188 (75.673 sec) +INFO:tensorflow:global_step/sec: 1.36875 +INFO:tensorflow:step = 21701, loss = 0.639832, precision = 0.898438 (73.059 sec) +INFO:tensorflow:global_step/sec: 1.36864 +INFO:tensorflow:step = 21801, loss = 0.754384, precision = 0.820312 (73.065 sec) +Saved checkpoint after 56 epoch(s) to ../data/resnet164/checkpoints/00056... +INFO:tensorflow:global_step/sec: 1.32332 +INFO:tensorflow:step = 21901, loss = 0.884339, precision = 0.804688 (75.568 sec) +INFO:tensorflow:global_step/sec: 1.3701 +INFO:tensorflow:step = 22001, loss = 0.726293, precision = 0.851562 (72.988 sec) +INFO:tensorflow:global_step/sec: 1.37323 +INFO:tensorflow:step = 22101, loss = 0.760113, precision = 0.867188 (72.821 sec) +INFO:tensorflow:global_step/sec: 1.37068 +INFO:tensorflow:step = 22201, loss = 0.688005, precision = 0.875 (72.957 sec) +Saved checkpoint after 57 epoch(s) to ../data/resnet164/checkpoints/00057... +INFO:tensorflow:global_step/sec: 1.32569 +INFO:tensorflow:step = 22301, loss = 0.590794, precision = 0.914062 (75.432 sec) +INFO:tensorflow:global_step/sec: 1.37074 +INFO:tensorflow:step = 22401, loss = 0.614374, precision = 0.914062 (72.953 sec) +INFO:tensorflow:global_step/sec: 1.37151 +INFO:tensorflow:step = 22501, loss = 0.7237, precision = 0.867188 (72.913 sec) +INFO:tensorflow:global_step/sec: 1.37105 +INFO:tensorflow:step = 22601, loss = 0.830606, precision = 0.84375 (72.937 sec) +Saved checkpoint after 58 epoch(s) to ../data/resnet164/checkpoints/00058... +INFO:tensorflow:global_step/sec: 1.33085 +INFO:tensorflow:step = 22701, loss = 0.663583, precision = 0.882812 (75.140 sec) +INFO:tensorflow:global_step/sec: 1.37022 +INFO:tensorflow:step = 22801, loss = 0.633595, precision = 0.890625 (72.981 sec) +INFO:tensorflow:global_step/sec: 1.37059 +INFO:tensorflow:step = 22901, loss = 0.77055, precision = 0.867188 (72.962 sec) +INFO:tensorflow:global_step/sec: 1.36954 +INFO:tensorflow:step = 23001, loss = 0.838063, precision = 0.875 (73.017 sec) +Saved checkpoint after 59 epoch(s) to ../data/resnet164/checkpoints/00059... +INFO:tensorflow:global_step/sec: 1.31863 +INFO:tensorflow:step = 23101, loss = 0.710584, precision = 0.867188 (75.837 sec) +INFO:tensorflow:global_step/sec: 1.37022 +INFO:tensorflow:step = 23201, loss = 0.634347, precision = 0.929688 (72.981 sec) +INFO:tensorflow:global_step/sec: 1.37009 +INFO:tensorflow:step = 23301, loss = 0.621303, precision = 0.875 (72.988 sec) +INFO:tensorflow:global_step/sec: 1.37081 +INFO:tensorflow:step = 23401, loss = 0.581712, precision = 0.914062 (72.950 sec) +Saved checkpoint after 60 epoch(s) to ../data/resnet164/checkpoints/00060... +INFO:tensorflow:global_step/sec: 1.32778 +INFO:tensorflow:step = 23501, loss = 0.690813, precision = 0.875 (75.313 sec) +INFO:tensorflow:global_step/sec: 1.37076 +INFO:tensorflow:step = 23601, loss = 0.601019, precision = 0.914062 (72.952 sec) +INFO:tensorflow:global_step/sec: 1.3712 +INFO:tensorflow:step = 23701, loss = 0.69342, precision = 0.875 (72.929 sec) +INFO:tensorflow:global_step/sec: 1.37173 +INFO:tensorflow:step = 23801, loss = 0.72075, precision = 0.859375 (72.901 sec) +Saved checkpoint after 61 epoch(s) to ../data/resnet164/checkpoints/00061... +INFO:tensorflow:global_step/sec: 1.32353 +INFO:tensorflow:step = 23901, loss = 0.598254, precision = 0.921875 (75.555 sec) +INFO:tensorflow:global_step/sec: 1.37161 +INFO:tensorflow:step = 24001, loss = 0.665181, precision = 0.882812 (72.907 sec) +INFO:tensorflow:global_step/sec: 1.37074 +INFO:tensorflow:step = 24101, loss = 0.784104, precision = 0.84375 (72.953 sec) +INFO:tensorflow:global_step/sec: 1.37182 +INFO:tensorflow:step = 24201, loss = 0.71747, precision = 0.851562 (72.896 sec) +Saved checkpoint after 62 epoch(s) to ../data/resnet164/checkpoints/00062... +INFO:tensorflow:global_step/sec: 1.32424 +INFO:tensorflow:step = 24301, loss = 0.664248, precision = 0.875 (75.515 sec) +INFO:tensorflow:global_step/sec: 1.37073 +INFO:tensorflow:step = 24401, loss = 0.893594, precision = 0.820312 (72.954 sec) +INFO:tensorflow:global_step/sec: 1.3719 +INFO:tensorflow:step = 24501, loss = 0.740997, precision = 0.875 (72.891 sec) +INFO:tensorflow:global_step/sec: 1.37141 +INFO:tensorflow:step = 24601, loss = 0.636155, precision = 0.882812 (72.918 sec) +Saved checkpoint after 63 epoch(s) to ../data/resnet164/checkpoints/00063... +INFO:tensorflow:global_step/sec: 1.32712 +INFO:tensorflow:step = 24701, loss = 0.702703, precision = 0.898438 (75.351 sec) +INFO:tensorflow:global_step/sec: 1.3709 +INFO:tensorflow:step = 24801, loss = 0.646487, precision = 0.898438 (72.945 sec) +INFO:tensorflow:global_step/sec: 1.37085 +INFO:tensorflow:step = 24901, loss = 0.598775, precision = 0.921875 (72.947 sec) +INFO:tensorflow:global_step/sec: 1.37098 +INFO:tensorflow:step = 25001, loss = 0.619533, precision = 0.90625 (72.940 sec) +Saved checkpoint after 64 epoch(s) to ../data/resnet164/checkpoints/00064... +INFO:tensorflow:global_step/sec: 1.32998 +INFO:tensorflow:step = 25101, loss = 0.652483, precision = 0.90625 (75.189 sec) +INFO:tensorflow:global_step/sec: 1.37198 +INFO:tensorflow:step = 25201, loss = 0.641927, precision = 0.898438 (72.887 sec) +INFO:tensorflow:global_step/sec: 1.37139 +INFO:tensorflow:step = 25301, loss = 0.611304, precision = 0.882812 (72.919 sec) +INFO:tensorflow:global_step/sec: 1.3723 +INFO:tensorflow:step = 25401, loss = 0.691827, precision = 0.914062 (72.870 sec) +Saved checkpoint after 65 epoch(s) to ../data/resnet164/checkpoints/00065... +INFO:tensorflow:global_step/sec: 1.32611 +INFO:tensorflow:step = 25501, loss = 0.602786, precision = 0.914062 (75.408 sec) +INFO:tensorflow:global_step/sec: 1.37053 +INFO:tensorflow:step = 25601, loss = 0.653085, precision = 0.890625 (72.964 sec) +INFO:tensorflow:global_step/sec: 1.37153 +INFO:tensorflow:step = 25701, loss = 0.646101, precision = 0.890625 (72.911 sec) +INFO:tensorflow:global_step/sec: 1.37056 +INFO:tensorflow:step = 25801, loss = 0.779574, precision = 0.828125 (72.963 sec) +Saved checkpoint after 66 epoch(s) to ../data/resnet164/checkpoints/00066... +INFO:tensorflow:global_step/sec: 1.32327 +INFO:tensorflow:step = 25901, loss = 0.664174, precision = 0.898438 (75.570 sec) +INFO:tensorflow:global_step/sec: 1.3708 +INFO:tensorflow:step = 26001, loss = 0.664288, precision = 0.882812 (72.950 sec) +INFO:tensorflow:global_step/sec: 1.3717 +INFO:tensorflow:step = 26101, loss = 0.52374, precision = 0.929688 (72.902 sec) +Saved checkpoint after 67 epoch(s) to ../data/resnet164/checkpoints/00067... +INFO:tensorflow:global_step/sec: 1.3298 +INFO:tensorflow:step = 26201, loss = 0.688218, precision = 0.859375 (75.209 sec) +INFO:tensorflow:global_step/sec: 1.37085 +INFO:tensorflow:step = 26301, loss = 0.685365, precision = 0.867188 (72.937 sec) +INFO:tensorflow:global_step/sec: 1.37093 +INFO:tensorflow:step = 26401, loss = 0.730734, precision = 0.882812 (72.943 sec) +INFO:tensorflow:global_step/sec: 1.3705 +INFO:tensorflow:step = 26501, loss = 0.91031, precision = 0.820312 (72.966 sec) +Saved checkpoint after 68 epoch(s) to ../data/resnet164/checkpoints/00068... +INFO:tensorflow:global_step/sec: 1.32555 +INFO:tensorflow:step = 26601, loss = 0.525819, precision = 0.929688 (75.440 sec) +INFO:tensorflow:global_step/sec: 1.37173 +INFO:tensorflow:step = 26701, loss = 0.687481, precision = 0.84375 (72.901 sec) +INFO:tensorflow:global_step/sec: 1.37131 +INFO:tensorflow:step = 26801, loss = 0.648203, precision = 0.867188 (72.923 sec) +INFO:tensorflow:global_step/sec: 1.37186 +INFO:tensorflow:step = 26901, loss = 0.727597, precision = 0.875 (72.894 sec) +Saved checkpoint after 69 epoch(s) to ../data/resnet164/checkpoints/00069... +INFO:tensorflow:global_step/sec: 1.32891 +INFO:tensorflow:step = 27001, loss = 0.6078, precision = 0.898438 (75.250 sec) +INFO:tensorflow:global_step/sec: 1.37236 +INFO:tensorflow:step = 27101, loss = 0.552421, precision = 0.9375 (72.867 sec) +INFO:tensorflow:global_step/sec: 1.37071 +INFO:tensorflow:step = 27201, loss = 0.72547, precision = 0.867188 (72.955 sec) +INFO:tensorflow:global_step/sec: 1.37023 +INFO:tensorflow:step = 27301, loss = 0.653861, precision = 0.898438 (72.980 sec) +Saved checkpoint after 70 epoch(s) to ../data/resnet164/checkpoints/00070... +INFO:tensorflow:global_step/sec: 1.31984 +INFO:tensorflow:step = 27401, loss = 0.733394, precision = 0.84375 (75.767 sec) +INFO:tensorflow:global_step/sec: 1.37194 +INFO:tensorflow:step = 27501, loss = 0.543096, precision = 0.945312 (72.890 sec) +INFO:tensorflow:global_step/sec: 1.37106 +INFO:tensorflow:step = 27601, loss = 0.70547, precision = 0.882812 (72.937 sec) +INFO:tensorflow:global_step/sec: 1.37136 +INFO:tensorflow:step = 27701, loss = 0.623387, precision = 0.914062 (72.920 sec) +Saved checkpoint after 71 epoch(s) to ../data/resnet164/checkpoints/00071... +INFO:tensorflow:global_step/sec: 1.32754 +INFO:tensorflow:step = 27801, loss = 0.556437, precision = 0.921875 (75.328 sec) +INFO:tensorflow:global_step/sec: 1.37216 +INFO:tensorflow:step = 27901, loss = 0.627702, precision = 0.898438 (72.877 sec) +INFO:tensorflow:global_step/sec: 1.37139 +INFO:tensorflow:step = 28001, loss = 0.643678, precision = 0.898438 (72.919 sec) +INFO:tensorflow:global_step/sec: 1.37139 +INFO:tensorflow:step = 28101, loss = 0.713449, precision = 0.851562 (72.919 sec) +Saved checkpoint after 72 epoch(s) to ../data/resnet164/checkpoints/00072... +INFO:tensorflow:global_step/sec: 1.33027 +INFO:tensorflow:step = 28201, loss = 0.597122, precision = 0.9375 (75.173 sec) +INFO:tensorflow:global_step/sec: 1.37114 +INFO:tensorflow:step = 28301, loss = 0.64567, precision = 0.882812 (72.932 sec) +INFO:tensorflow:global_step/sec: 1.37146 +INFO:tensorflow:step = 28401, loss = 0.796537, precision = 0.84375 (72.915 sec) +INFO:tensorflow:global_step/sec: 1.37254 +INFO:tensorflow:step = 28501, loss = 0.555004, precision = 0.9375 (72.857 sec) +Saved checkpoint after 73 epoch(s) to ../data/resnet164/checkpoints/00073... +INFO:tensorflow:global_step/sec: 1.33219 +INFO:tensorflow:step = 28601, loss = 0.594347, precision = 0.921875 (75.064 sec) +INFO:tensorflow:global_step/sec: 1.37118 +INFO:tensorflow:step = 28701, loss = 0.717624, precision = 0.851562 (72.930 sec) +INFO:tensorflow:global_step/sec: 1.37151 +INFO:tensorflow:step = 28801, loss = 0.687951, precision = 0.867188 (72.912 sec) +INFO:tensorflow:global_step/sec: 1.37127 +INFO:tensorflow:step = 28901, loss = 0.639704, precision = 0.914062 (72.925 sec) +Saved checkpoint after 74 epoch(s) to ../data/resnet164/checkpoints/00074... +INFO:tensorflow:global_step/sec: 1.33007 +INFO:tensorflow:step = 29001, loss = 0.761097, precision = 0.875 (75.184 sec) +INFO:tensorflow:global_step/sec: 1.37175 +INFO:tensorflow:step = 29101, loss = 0.689865, precision = 0.867188 (72.900 sec) +INFO:tensorflow:global_step/sec: 1.37093 +INFO:tensorflow:step = 29201, loss = 0.683645, precision = 0.875 (72.943 sec) +INFO:tensorflow:global_step/sec: 1.37209 +INFO:tensorflow:step = 29301, loss = 0.670484, precision = 0.882812 (72.882 sec) +Saved checkpoint after 75 epoch(s) to ../data/resnet164/checkpoints/00075... +INFO:tensorflow:global_step/sec: 1.32694 +INFO:tensorflow:step = 29401, loss = 0.688609, precision = 0.890625 (75.361 sec) +INFO:tensorflow:global_step/sec: 1.37125 +INFO:tensorflow:step = 29501, loss = 0.54424, precision = 0.921875 (72.926 sec) +INFO:tensorflow:global_step/sec: 1.37123 +INFO:tensorflow:step = 29601, loss = 0.701649, precision = 0.867188 (72.928 sec) +INFO:tensorflow:global_step/sec: 1.37066 +INFO:tensorflow:step = 29701, loss = 0.678402, precision = 0.882812 (72.957 sec) +Saved checkpoint after 76 epoch(s) to ../data/resnet164/checkpoints/00076... +INFO:tensorflow:global_step/sec: 1.32509 +INFO:tensorflow:step = 29801, loss = 0.670686, precision = 0.890625 (75.467 sec) +INFO:tensorflow:global_step/sec: 1.37118 +INFO:tensorflow:step = 29901, loss = 0.57613, precision = 0.921875 (72.930 sec) +INFO:tensorflow:global_step/sec: 1.37169 +INFO:tensorflow:step = 30001, loss = 0.600019, precision = 0.921875 (72.903 sec) +INFO:tensorflow:global_step/sec: 1.37143 +INFO:tensorflow:step = 30101, loss = 0.639701, precision = 0.914062 (72.917 sec) +Saved checkpoint after 77 epoch(s) to ../data/resnet164/checkpoints/00077... +INFO:tensorflow:global_step/sec: 1.32432 +INFO:tensorflow:step = 30201, loss = 0.69099, precision = 0.859375 (75.511 sec) +INFO:tensorflow:global_step/sec: 1.37181 +INFO:tensorflow:step = 30301, loss = 0.597253, precision = 0.898438 (72.896 sec) +INFO:tensorflow:global_step/sec: 1.37092 +INFO:tensorflow:step = 30401, loss = 0.639272, precision = 0.898438 (72.944 sec) +Saved checkpoint after 78 epoch(s) to ../data/resnet164/checkpoints/00078... +INFO:tensorflow:global_step/sec: 1.32852 +INFO:tensorflow:step = 30501, loss = 0.581394, precision = 0.914062 (75.272 sec) +INFO:tensorflow:global_step/sec: 1.37367 +INFO:tensorflow:step = 30601, loss = 0.67999, precision = 0.867188 (72.798 sec) +INFO:tensorflow:global_step/sec: 1.37374 +INFO:tensorflow:step = 30701, loss = 0.53876, precision = 0.9375 (72.794 sec) +INFO:tensorflow:global_step/sec: 1.37265 +INFO:tensorflow:step = 30801, loss = 0.673661, precision = 0.875 (72.852 sec) +Saved checkpoint after 79 epoch(s) to ../data/resnet164/checkpoints/00079... +INFO:tensorflow:global_step/sec: 1.32907 +INFO:tensorflow:step = 30901, loss = 0.613382, precision = 0.929688 (75.241 sec) +INFO:tensorflow:global_step/sec: 1.37281 +INFO:tensorflow:step = 31001, loss = 0.674589, precision = 0.882812 (72.843 sec) +INFO:tensorflow:global_step/sec: 1.3734 +INFO:tensorflow:step = 31101, loss = 0.603141, precision = 0.90625 (72.812 sec) +INFO:tensorflow:global_step/sec: 1.37204 +INFO:tensorflow:step = 31201, loss = 0.755368, precision = 0.84375 (72.885 sec) +Saved checkpoint after 80 epoch(s) to ../data/resnet164/checkpoints/00080... +INFO:tensorflow:global_step/sec: 1.32094 +INFO:tensorflow:step = 31301, loss = 0.625406, precision = 0.875 (75.704 sec) +INFO:tensorflow:global_step/sec: 1.37168 +INFO:tensorflow:step = 31401, loss = 0.684148, precision = 0.875 (72.903 sec) +INFO:tensorflow:global_step/sec: 1.3728 +INFO:tensorflow:step = 31501, loss = 0.605712, precision = 0.914062 (72.844 sec) +INFO:tensorflow:global_step/sec: 1.37183 +INFO:tensorflow:step = 31601, loss = 0.692505, precision = 0.859375 (72.896 sec) +Saved checkpoint after 81 epoch(s) to ../data/resnet164/checkpoints/00081... +INFO:tensorflow:global_step/sec: 1.32414 +INFO:tensorflow:step = 31701, loss = 0.833368, precision = 0.804688 (75.520 sec) +INFO:tensorflow:global_step/sec: 1.37121 +INFO:tensorflow:step = 31801, loss = 0.739295, precision = 0.882812 (72.928 sec) +INFO:tensorflow:global_step/sec: 1.37244 +INFO:tensorflow:step = 31901, loss = 0.63602, precision = 0.859375 (72.863 sec) +INFO:tensorflow:global_step/sec: 1.37245 +INFO:tensorflow:step = 32001, loss = 0.740549, precision = 0.828125 (72.863 sec) +Saved checkpoint after 82 epoch(s) to ../data/resnet164/checkpoints/00082... +INFO:tensorflow:global_step/sec: 1.32881 +INFO:tensorflow:step = 32101, loss = 0.595047, precision = 0.90625 (75.255 sec) +INFO:tensorflow:global_step/sec: 1.37143 +INFO:tensorflow:step = 32201, loss = 0.689707, precision = 0.882812 (72.917 sec) +INFO:tensorflow:global_step/sec: 1.37266 +INFO:tensorflow:step = 32301, loss = 0.587154, precision = 0.898438 (72.851 sec) +INFO:tensorflow:global_step/sec: 1.37158 +INFO:tensorflow:step = 32401, loss = 0.648937, precision = 0.867188 (72.909 sec) +Saved checkpoint after 83 epoch(s) to ../data/resnet164/checkpoints/00083... +INFO:tensorflow:global_step/sec: 1.33046 +INFO:tensorflow:step = 32501, loss = 0.796714, precision = 0.828125 (75.162 sec) +INFO:tensorflow:global_step/sec: 1.37186 +INFO:tensorflow:step = 32601, loss = 0.708111, precision = 0.875 (72.894 sec) +INFO:tensorflow:global_step/sec: 1.37256 +INFO:tensorflow:step = 32701, loss = 0.609998, precision = 0.882812 (72.856 sec) +INFO:tensorflow:global_step/sec: 1.37074 +INFO:tensorflow:step = 32801, loss = 0.670477, precision = 0.875 (72.954 sec) +Saved checkpoint after 84 epoch(s) to ../data/resnet164/checkpoints/00084... +INFO:tensorflow:global_step/sec: 1.32811 +INFO:tensorflow:step = 32901, loss = 0.631521, precision = 0.882812 (75.295 sec) +INFO:tensorflow:global_step/sec: 1.37129 +INFO:tensorflow:step = 33001, loss = 0.803135, precision = 0.859375 (72.924 sec) +INFO:tensorflow:global_step/sec: 1.37139 +INFO:tensorflow:step = 33101, loss = 0.602904, precision = 0.90625 (72.919 sec) +INFO:tensorflow:global_step/sec: 1.37062 +INFO:tensorflow:step = 33201, loss = 0.64675, precision = 0.898438 (72.960 sec) +Saved checkpoint after 85 epoch(s) to ../data/resnet164/checkpoints/00085... +INFO:tensorflow:global_step/sec: 1.32324 +INFO:tensorflow:step = 33301, loss = 0.745409, precision = 0.882812 (75.572 sec) +INFO:tensorflow:global_step/sec: 1.37122 +INFO:tensorflow:step = 33401, loss = 0.578977, precision = 0.929688 (72.928 sec) +INFO:tensorflow:global_step/sec: 1.37147 +INFO:tensorflow:step = 33501, loss = 0.674285, precision = 0.890625 (72.914 sec) +INFO:tensorflow:global_step/sec: 1.37149 +INFO:tensorflow:step = 33601, loss = 0.673798, precision = 0.867188 (72.914 sec) +Saved checkpoint after 86 epoch(s) to ../data/resnet164/checkpoints/00086... +INFO:tensorflow:global_step/sec: 1.3263 +INFO:tensorflow:step = 33701, loss = 0.648117, precision = 0.882812 (75.397 sec) +INFO:tensorflow:global_step/sec: 1.3717 +INFO:tensorflow:step = 33801, loss = 0.693877, precision = 0.859375 (72.902 sec) +INFO:tensorflow:global_step/sec: 1.37217 +INFO:tensorflow:step = 33901, loss = 0.620209, precision = 0.914062 (72.877 sec) +INFO:tensorflow:global_step/sec: 1.3712 +INFO:tensorflow:step = 34001, loss = 0.626663, precision = 0.890625 (72.929 sec) +Saved checkpoint after 87 epoch(s) to ../data/resnet164/checkpoints/00087... +INFO:tensorflow:global_step/sec: 1.32932 +INFO:tensorflow:step = 34101, loss = 0.65773, precision = 0.882812 (75.227 sec) +INFO:tensorflow:global_step/sec: 1.37324 +INFO:tensorflow:step = 34201, loss = 0.658738, precision = 0.898438 (72.820 sec) +INFO:tensorflow:global_step/sec: 1.37227 +INFO:tensorflow:step = 34301, loss = 0.616763, precision = 0.90625 (72.872 sec) +INFO:tensorflow:global_step/sec: 1.3723 +INFO:tensorflow:step = 34401, loss = 0.684392, precision = 0.84375 (72.871 sec) +Saved checkpoint after 88 epoch(s) to ../data/resnet164/checkpoints/00088... +INFO:tensorflow:global_step/sec: 1.32987 +INFO:tensorflow:step = 34501, loss = 0.67165, precision = 0.882812 (75.195 sec) +INFO:tensorflow:global_step/sec: 1.37343 +INFO:tensorflow:step = 34601, loss = 0.59089, precision = 0.921875 (72.810 sec) +INFO:tensorflow:global_step/sec: 1.37282 +INFO:tensorflow:step = 34701, loss = 0.726916, precision = 0.882812 (72.842 sec) +Saved checkpoint after 89 epoch(s) to ../data/resnet164/checkpoints/00089... +INFO:tensorflow:global_step/sec: 1.32508 +INFO:tensorflow:step = 34801, loss = 0.711008, precision = 0.875 (75.467 sec) +INFO:tensorflow:global_step/sec: 1.37225 +INFO:tensorflow:step = 34901, loss = 0.713092, precision = 0.851562 (72.873 sec) +INFO:tensorflow:global_step/sec: 1.37206 +INFO:tensorflow:step = 35001, loss = 0.589955, precision = 0.929688 (72.883 sec) +INFO:tensorflow:global_step/sec: 1.37155 +INFO:tensorflow:step = 35101, loss = 0.757659, precision = 0.835938 (72.910 sec) +Saved checkpoint after 90 epoch(s) to ../data/resnet164/checkpoints/00090... +INFO:tensorflow:global_step/sec: 1.31997 +INFO:tensorflow:step = 35201, loss = 0.678101, precision = 0.859375 (75.760 sec) +INFO:tensorflow:global_step/sec: 1.3718 +INFO:tensorflow:step = 35301, loss = 0.677211, precision = 0.882812 (72.897 sec) +INFO:tensorflow:global_step/sec: 1.37217 +INFO:tensorflow:step = 35401, loss = 0.708771, precision = 0.875 (72.877 sec) +INFO:tensorflow:global_step/sec: 1.37214 +INFO:tensorflow:step = 35501, loss = 0.717971, precision = 0.882812 (72.879 sec) +Saved checkpoint after 91 epoch(s) to ../data/resnet164/checkpoints/00091... +INFO:tensorflow:global_step/sec: 1.32926 +INFO:tensorflow:step = 35601, loss = 0.795168, precision = 0.84375 (75.230 sec) +INFO:tensorflow:global_step/sec: 1.3715 +INFO:tensorflow:step = 35701, loss = 0.533566, precision = 0.914062 (72.913 sec) +INFO:tensorflow:global_step/sec: 1.37208 +INFO:tensorflow:step = 35801, loss = 0.444001, precision = 0.953125 (72.882 sec) +INFO:tensorflow:global_step/sec: 1.3717 +INFO:tensorflow:step = 35901, loss = 0.424197, precision = 0.976562 (72.902 sec) +Saved checkpoint after 92 epoch(s) to ../data/resnet164/checkpoints/00092... +INFO:tensorflow:global_step/sec: 1.32346 +INFO:tensorflow:step = 36001, loss = 0.563998, precision = 0.90625 (75.560 sec) +INFO:tensorflow:global_step/sec: 1.37084 +INFO:tensorflow:step = 36101, loss = 0.44258, precision = 0.960938 (72.948 sec) +INFO:tensorflow:global_step/sec: 1.36995 +INFO:tensorflow:step = 36201, loss = 0.438873, precision = 0.96875 (72.995 sec) +INFO:tensorflow:global_step/sec: 1.37105 +INFO:tensorflow:step = 36301, loss = 0.458619, precision = 0.945312 (72.937 sec) +Saved checkpoint after 93 epoch(s) to ../data/resnet164/checkpoints/00093... +INFO:tensorflow:global_step/sec: 1.32886 +INFO:tensorflow:step = 36401, loss = 0.422357, precision = 0.945312 (75.253 sec) +INFO:tensorflow:global_step/sec: 1.37039 +INFO:tensorflow:step = 36501, loss = 0.403577, precision = 0.976562 (72.971 sec) +INFO:tensorflow:global_step/sec: 1.36992 +INFO:tensorflow:step = 36601, loss = 0.418099, precision = 0.96875 (72.997 sec) +INFO:tensorflow:global_step/sec: 1.36994 +INFO:tensorflow:step = 36701, loss = 0.419526, precision = 0.960938 (72.996 sec) +Saved checkpoint after 94 epoch(s) to ../data/resnet164/checkpoints/00094... +INFO:tensorflow:global_step/sec: 1.32601 +INFO:tensorflow:step = 36801, loss = 0.426, precision = 0.96875 (75.415 sec) +INFO:tensorflow:global_step/sec: 1.36967 +INFO:tensorflow:step = 36901, loss = 0.38412, precision = 0.976562 (73.010 sec) +INFO:tensorflow:global_step/sec: 1.37013 +INFO:tensorflow:step = 37001, loss = 0.354636, precision = 0.984375 (72.986 sec) +INFO:tensorflow:global_step/sec: 1.36969 +INFO:tensorflow:step = 37101, loss = 0.417852, precision = 0.953125 (73.009 sec) +Saved checkpoint after 95 epoch(s) to ../data/resnet164/checkpoints/00095... +INFO:tensorflow:global_step/sec: 1.32946 +INFO:tensorflow:step = 37201, loss = 0.426299, precision = 0.953125 (75.219 sec) +INFO:tensorflow:global_step/sec: 1.36944 +INFO:tensorflow:step = 37301, loss = 0.397222, precision = 0.960938 (73.022 sec) +INFO:tensorflow:global_step/sec: 1.36969 +INFO:tensorflow:step = 37401, loss = 0.338224, precision = 0.992188 (73.009 sec) +INFO:tensorflow:global_step/sec: 1.37273 +INFO:tensorflow:step = 37501, loss = 0.379208, precision = 0.984375 (72.847 sec) +Saved checkpoint after 96 epoch(s) to ../data/resnet164/checkpoints/00096... +INFO:tensorflow:global_step/sec: 1.32991 +INFO:tensorflow:step = 37601, loss = 0.40155, precision = 0.960938 (75.193 sec) +INFO:tensorflow:global_step/sec: 1.37587 +INFO:tensorflow:step = 37701, loss = 0.374061, precision = 0.96875 (72.681 sec) +INFO:tensorflow:global_step/sec: 1.37586 +INFO:tensorflow:step = 37801, loss = 0.379465, precision = 0.953125 (72.682 sec) +INFO:tensorflow:global_step/sec: 1.37411 +INFO:tensorflow:step = 37901, loss = 0.375295, precision = 0.96875 (72.774 sec) +Saved checkpoint after 97 epoch(s) to ../data/resnet164/checkpoints/00097... +INFO:tensorflow:global_step/sec: 1.33289 +INFO:tensorflow:step = 38001, loss = 0.330116, precision = 0.992188 (75.026 sec) +INFO:tensorflow:global_step/sec: 1.37575 +INFO:tensorflow:step = 38101, loss = 0.343554, precision = 0.984375 (72.687 sec) +INFO:tensorflow:global_step/sec: 1.37514 +INFO:tensorflow:step = 38201, loss = 0.37667, precision = 0.960938 (72.720 sec) +INFO:tensorflow:global_step/sec: 1.37575 +INFO:tensorflow:step = 38301, loss = 0.402784, precision = 0.953125 (72.688 sec) +Saved checkpoint after 98 epoch(s) to ../data/resnet164/checkpoints/00098... +INFO:tensorflow:global_step/sec: 1.3322 +INFO:tensorflow:step = 38401, loss = 0.314072, precision = 0.984375 (75.064 sec) +INFO:tensorflow:global_step/sec: 1.37574 +INFO:tensorflow:step = 38501, loss = 0.346755, precision = 0.96875 (72.688 sec) +INFO:tensorflow:global_step/sec: 1.3764 +INFO:tensorflow:step = 38601, loss = 0.381454, precision = 0.953125 (72.653 sec) +INFO:tensorflow:global_step/sec: 1.37572 +INFO:tensorflow:step = 38701, loss = 0.317723, precision = 0.976562 (72.690 sec) +Saved checkpoint after 99 epoch(s) to ../data/resnet164/checkpoints/00099... +INFO:tensorflow:global_step/sec: 1.33456 +INFO:tensorflow:step = 38801, loss = 0.368211, precision = 0.976562 (74.931 sec) +INFO:tensorflow:global_step/sec: 1.37397 +INFO:tensorflow:step = 38901, loss = 0.340032, precision = 0.953125 (72.782 sec) +INFO:tensorflow:global_step/sec: 1.37401 +INFO:tensorflow:step = 39001, loss = 0.364551, precision = 0.96875 (72.780 sec) +Saved checkpoint after 100 epoch(s) to ../data/resnet164/checkpoints/00100... +INFO:tensorflow:global_step/sec: 1.33497 +INFO:tensorflow:step = 39101, loss = 0.300982, precision = 0.992188 (74.908 sec) +INFO:tensorflow:global_step/sec: 1.37573 +INFO:tensorflow:step = 39201, loss = 0.356936, precision = 0.960938 (72.688 sec) +INFO:tensorflow:global_step/sec: 1.37554 +INFO:tensorflow:step = 39301, loss = 0.300981, precision = 0.96875 (72.698 sec) +INFO:tensorflow:global_step/sec: 1.37623 +INFO:tensorflow:step = 39401, loss = 0.362108, precision = 0.953125 (72.663 sec) +Saved checkpoint after 101 epoch(s) to ../data/resnet164/checkpoints/00101... +INFO:tensorflow:global_step/sec: 1.33293 +INFO:tensorflow:step = 39501, loss = 0.284121, precision = 0.992188 (75.023 sec) +INFO:tensorflow:global_step/sec: 1.37709 +INFO:tensorflow:step = 39601, loss = 0.341939, precision = 0.976562 (72.617 sec) +INFO:tensorflow:global_step/sec: 1.37512 +INFO:tensorflow:step = 39701, loss = 0.268707, precision = 1.0 (72.721 sec) +INFO:tensorflow:global_step/sec: 1.37474 +INFO:tensorflow:step = 39801, loss = 0.277791, precision = 0.992188 (72.741 sec) +Saved checkpoint after 102 epoch(s) to ../data/resnet164/checkpoints/00102... +INFO:tensorflow:global_step/sec: 1.33215 +INFO:tensorflow:step = 39901, loss = 0.283132, precision = 0.992188 (75.067 sec) +INFO:tensorflow:global_step/sec: 1.37596 +INFO:tensorflow:step = 40001, loss = 0.296474, precision = 0.984375 (72.676 sec) +INFO:tensorflow:global_step/sec: 1.37625 +INFO:tensorflow:step = 40101, loss = 0.320616, precision = 0.960938 (72.661 sec) +INFO:tensorflow:global_step/sec: 1.37474 +INFO:tensorflow:step = 40201, loss = 0.261075, precision = 0.992188 (72.741 sec) +Saved checkpoint after 103 epoch(s) to ../data/resnet164/checkpoints/00103... +INFO:tensorflow:global_step/sec: 1.328 +INFO:tensorflow:step = 40301, loss = 0.259838, precision = 0.992188 (75.301 sec) +INFO:tensorflow:global_step/sec: 1.37456 +INFO:tensorflow:step = 40401, loss = 0.285218, precision = 0.984375 (72.751 sec) +INFO:tensorflow:global_step/sec: 1.37578 +INFO:tensorflow:step = 40501, loss = 0.280434, precision = 0.984375 (72.686 sec) +INFO:tensorflow:global_step/sec: 1.37654 +INFO:tensorflow:step = 40601, loss = 0.27499, precision = 0.976562 (72.646 sec) +Saved checkpoint after 104 epoch(s) to ../data/resnet164/checkpoints/00104... +INFO:tensorflow:global_step/sec: 1.33392 +INFO:tensorflow:step = 40701, loss = 0.30732, precision = 0.96875 (74.967 sec) +INFO:tensorflow:global_step/sec: 1.37581 +INFO:tensorflow:step = 40801, loss = 0.280185, precision = 0.976562 (72.685 sec) +INFO:tensorflow:global_step/sec: 1.37607 +INFO:tensorflow:step = 40901, loss = 0.290653, precision = 0.984375 (72.671 sec) +INFO:tensorflow:global_step/sec: 1.37526 +INFO:tensorflow:step = 41001, loss = 0.303518, precision = 0.976562 (72.714 sec) +Saved checkpoint after 105 epoch(s) to ../data/resnet164/checkpoints/00105... +INFO:tensorflow:global_step/sec: 1.33586 +INFO:tensorflow:step = 41101, loss = 0.274048, precision = 0.96875 (74.857 sec) +INFO:tensorflow:global_step/sec: 1.3765 +INFO:tensorflow:step = 41201, loss = 0.25313, precision = 0.984375 (72.648 sec) +INFO:tensorflow:global_step/sec: 1.37632 +INFO:tensorflow:step = 41301, loss = 0.271363, precision = 0.976562 (72.657 sec) +INFO:tensorflow:global_step/sec: 1.37568 +INFO:tensorflow:step = 41401, loss = 0.268606, precision = 0.976562 (72.692 sec) +Saved checkpoint after 106 epoch(s) to ../data/resnet164/checkpoints/00106... +INFO:tensorflow:global_step/sec: 1.32996 +INFO:tensorflow:step = 41501, loss = 0.257374, precision = 0.992188 (75.190 sec) +INFO:tensorflow:global_step/sec: 1.37447 +INFO:tensorflow:step = 41601, loss = 0.232236, precision = 0.992188 (72.756 sec) +INFO:tensorflow:global_step/sec: 1.37544 +INFO:tensorflow:step = 41701, loss = 0.253331, precision = 0.992188 (72.704 sec) +INFO:tensorflow:global_step/sec: 1.37607 +INFO:tensorflow:step = 41801, loss = 0.217525, precision = 1.0 (72.671 sec) +Saved checkpoint after 107 epoch(s) to ../data/resnet164/checkpoints/00107... +INFO:tensorflow:global_step/sec: 1.33082 +INFO:tensorflow:step = 41901, loss = 0.262633, precision = 0.984375 (75.142 sec) +INFO:tensorflow:global_step/sec: 1.37456 +INFO:tensorflow:step = 42001, loss = 0.241527, precision = 0.992188 (72.751 sec) +INFO:tensorflow:global_step/sec: 1.37415 +INFO:tensorflow:step = 42101, loss = 0.253634, precision = 0.984375 (72.772 sec) +INFO:tensorflow:global_step/sec: 1.37658 +INFO:tensorflow:step = 42201, loss = 0.268976, precision = 0.976562 (72.644 sec) +Saved checkpoint after 108 epoch(s) to ../data/resnet164/checkpoints/00108... +INFO:tensorflow:global_step/sec: 1.33408 +INFO:tensorflow:step = 42301, loss = 0.237522, precision = 0.992188 (74.958 sec) +INFO:tensorflow:global_step/sec: 1.37494 +INFO:tensorflow:step = 42401, loss = 0.259983, precision = 0.96875 (72.730 sec) +INFO:tensorflow:global_step/sec: 1.37597 +INFO:tensorflow:step = 42501, loss = 0.248893, precision = 0.992188 (72.676 sec) +INFO:tensorflow:global_step/sec: 1.37602 +INFO:tensorflow:step = 42601, loss = 0.244052, precision = 0.992188 (72.674 sec) +Saved checkpoint after 109 epoch(s) to ../data/resnet164/checkpoints/00109... +INFO:tensorflow:global_step/sec: 1.33671 +INFO:tensorflow:step = 42701, loss = 0.295629, precision = 0.96875 (74.812 sec) +INFO:tensorflow:global_step/sec: 1.37555 +INFO:tensorflow:step = 42801, loss = 0.22094, precision = 1.0 (72.696 sec) +INFO:tensorflow:global_step/sec: 1.37428 +INFO:tensorflow:step = 42901, loss = 0.231686, precision = 0.984375 (72.766 sec) +INFO:tensorflow:global_step/sec: 1.3748 +INFO:tensorflow:step = 43001, loss = 0.238953, precision = 0.992188 (72.738 sec) +Saved checkpoint after 110 epoch(s) to ../data/resnet164/checkpoints/00110... +INFO:tensorflow:global_step/sec: 1.33609 +INFO:tensorflow:step = 43101, loss = 0.29896, precision = 0.96875 (74.845 sec) +INFO:tensorflow:global_step/sec: 1.37596 +INFO:tensorflow:step = 43201, loss = 0.246066, precision = 0.984375 (72.676 sec) +INFO:tensorflow:global_step/sec: 1.3767 +INFO:tensorflow:step = 43301, loss = 0.277592, precision = 0.960938 (72.637 sec) +Saved checkpoint after 111 epoch(s) to ../data/resnet164/checkpoints/00111... +INFO:tensorflow:global_step/sec: 1.32863 +INFO:tensorflow:step = 43401, loss = 0.230422, precision = 0.992188 (75.265 sec) +INFO:tensorflow:global_step/sec: 1.37595 +INFO:tensorflow:step = 43501, loss = 0.229203, precision = 0.984375 (72.677 sec) +INFO:tensorflow:global_step/sec: 1.37654 +INFO:tensorflow:step = 43601, loss = 0.246125, precision = 0.976562 (72.646 sec) +INFO:tensorflow:global_step/sec: 1.37567 +INFO:tensorflow:step = 43701, loss = 0.253518, precision = 0.984375 (72.692 sec) +Saved checkpoint after 112 epoch(s) to ../data/resnet164/checkpoints/00112... +INFO:tensorflow:global_step/sec: 1.33506 +INFO:tensorflow:step = 43801, loss = 0.251241, precision = 0.976562 (74.903 sec) +INFO:tensorflow:global_step/sec: 1.37621 +INFO:tensorflow:step = 43901, loss = 0.220936, precision = 0.984375 (72.663 sec) +INFO:tensorflow:global_step/sec: 1.37638 +INFO:tensorflow:step = 44001, loss = 0.209757, precision = 0.992188 (72.655 sec) +INFO:tensorflow:global_step/sec: 1.37586 +INFO:tensorflow:step = 44101, loss = 0.239398, precision = 0.976562 (72.682 sec) +Saved checkpoint after 113 epoch(s) to ../data/resnet164/checkpoints/00113... +INFO:tensorflow:global_step/sec: 1.33094 +INFO:tensorflow:step = 44201, loss = 0.205319, precision = 1.0 (75.135 sec) +INFO:tensorflow:global_step/sec: 1.37498 +INFO:tensorflow:step = 44301, loss = 0.195147, precision = 1.0 (72.728 sec) +INFO:tensorflow:global_step/sec: 1.37474 +INFO:tensorflow:step = 44401, loss = 0.238269, precision = 0.976562 (72.741 sec) +INFO:tensorflow:global_step/sec: 1.37464 +INFO:tensorflow:step = 44501, loss = 0.247042, precision = 0.96875 (72.746 sec) +Saved checkpoint after 114 epoch(s) to ../data/resnet164/checkpoints/00114... +INFO:tensorflow:global_step/sec: 1.33348 +INFO:tensorflow:step = 44601, loss = 0.200843, precision = 1.0 (74.992 sec) +INFO:tensorflow:global_step/sec: 1.37579 +INFO:tensorflow:step = 44701, loss = 0.205595, precision = 1.0 (72.685 sec) +INFO:tensorflow:global_step/sec: 1.37609 +INFO:tensorflow:step = 44801, loss = 0.219409, precision = 0.976562 (72.670 sec) +INFO:tensorflow:global_step/sec: 1.37637 +INFO:tensorflow:step = 44901, loss = 0.221233, precision = 0.984375 (72.655 sec) +Saved checkpoint after 115 epoch(s) to ../data/resnet164/checkpoints/00115... +INFO:tensorflow:global_step/sec: 1.33293 +INFO:tensorflow:step = 45001, loss = 0.234406, precision = 0.992188 (75.023 sec) +INFO:tensorflow:global_step/sec: 1.37604 +INFO:tensorflow:step = 45101, loss = 0.277791, precision = 0.953125 (72.672 sec) +INFO:tensorflow:global_step/sec: 1.37568 +INFO:tensorflow:step = 45201, loss = 0.20173, precision = 0.992188 (72.691 sec) +INFO:tensorflow:global_step/sec: 1.37657 +INFO:tensorflow:step = 45301, loss = 0.245424, precision = 0.960938 (72.644 sec) +Saved checkpoint after 116 epoch(s) to ../data/resnet164/checkpoints/00116... +INFO:tensorflow:global_step/sec: 1.33776 +INFO:tensorflow:step = 45401, loss = 0.234475, precision = 0.984375 (74.752 sec) +INFO:tensorflow:global_step/sec: 1.37573 +INFO:tensorflow:step = 45501, loss = 0.217055, precision = 0.984375 (72.689 sec) +INFO:tensorflow:global_step/sec: 1.3741 +INFO:tensorflow:step = 45601, loss = 0.201397, precision = 0.992188 (72.775 sec) +INFO:tensorflow:global_step/sec: 1.37411 +INFO:tensorflow:step = 45701, loss = 0.207424, precision = 0.992188 (72.774 sec) +Saved checkpoint after 117 epoch(s) to ../data/resnet164/checkpoints/00117... +INFO:tensorflow:global_step/sec: 1.33054 +INFO:tensorflow:step = 45801, loss = 0.201582, precision = 0.992188 (75.157 sec) +INFO:tensorflow:global_step/sec: 1.37562 +INFO:tensorflow:step = 45901, loss = 0.211636, precision = 0.984375 (72.695 sec) +INFO:tensorflow:global_step/sec: 1.37517 +INFO:tensorflow:step = 46001, loss = 0.239352, precision = 0.960938 (72.718 sec) +INFO:tensorflow:global_step/sec: 1.37601 +INFO:tensorflow:step = 46101, loss = 0.229453, precision = 0.96875 (72.673 sec) +Saved checkpoint after 118 epoch(s) to ../data/resnet164/checkpoints/00118... +INFO:tensorflow:global_step/sec: 1.33666 +INFO:tensorflow:step = 46201, loss = 0.221328, precision = 0.984375 (74.813 sec) +INFO:tensorflow:global_step/sec: 1.37547 +INFO:tensorflow:step = 46301, loss = 0.272667, precision = 0.96875 (72.702 sec) +INFO:tensorflow:global_step/sec: 1.37565 +INFO:tensorflow:step = 46401, loss = 0.248312, precision = 0.976562 (72.693 sec) +INFO:tensorflow:global_step/sec: 1.3766 +INFO:tensorflow:step = 46501, loss = 0.230535, precision = 0.976562 (72.643 sec) +Saved checkpoint after 119 epoch(s) to ../data/resnet164/checkpoints/00119... +INFO:tensorflow:global_step/sec: 1.33299 +INFO:tensorflow:step = 46601, loss = 0.188552, precision = 0.992188 (75.019 sec) +INFO:tensorflow:global_step/sec: 1.37538 +INFO:tensorflow:step = 46701, loss = 0.230677, precision = 0.976562 (72.707 sec) +INFO:tensorflow:global_step/sec: 1.37656 +INFO:tensorflow:step = 46801, loss = 0.223248, precision = 0.96875 (72.645 sec) +INFO:tensorflow:global_step/sec: 1.37465 +INFO:tensorflow:step = 46901, loss = 0.193288, precision = 0.992188 (72.746 sec) +Saved checkpoint after 120 epoch(s) to ../data/resnet164/checkpoints/00120... +INFO:tensorflow:global_step/sec: 1.33085 +INFO:tensorflow:step = 47001, loss = 0.204425, precision = 0.992188 (75.140 sec) +INFO:tensorflow:global_step/sec: 1.37551 +INFO:tensorflow:step = 47101, loss = 0.275734, precision = 0.96875 (72.700 sec) +INFO:tensorflow:global_step/sec: 1.37698 +INFO:tensorflow:step = 47201, loss = 0.235977, precision = 0.960938 (72.623 sec) +INFO:tensorflow:global_step/sec: 1.37615 +INFO:tensorflow:step = 47301, loss = 0.25154, precision = 0.976562 (72.666 sec) +Saved checkpoint after 121 epoch(s) to ../data/resnet164/checkpoints/00121... +INFO:tensorflow:global_step/sec: 1.32963 +INFO:tensorflow:step = 47401, loss = 0.172139, precision = 1.0 (75.209 sec) +INFO:tensorflow:global_step/sec: 1.37533 +INFO:tensorflow:step = 47501, loss = 0.181039, precision = 1.0 (72.710 sec) +INFO:tensorflow:global_step/sec: 1.37664 +INFO:tensorflow:step = 47601, loss = 0.191769, precision = 0.992188 (72.641 sec) +INFO:tensorflow:global_step/sec: 1.37438 +INFO:tensorflow:step = 47701, loss = 0.18515, precision = 0.992188 (72.760 sec) +Saved checkpoint after 122 epoch(s) to ../data/resnet164/checkpoints/00122... +INFO:tensorflow:global_step/sec: 1.33141 +INFO:tensorflow:step = 47801, loss = 0.227922, precision = 0.976562 (75.108 sec) +INFO:tensorflow:global_step/sec: 1.37542 +INFO:tensorflow:step = 47901, loss = 0.193229, precision = 0.992188 (72.705 sec) +INFO:tensorflow:global_step/sec: 1.37543 +INFO:tensorflow:step = 48001, loss = 0.221737, precision = 0.984375 (72.705 sec) +Saved checkpoint after 123 epoch(s) to ../data/resnet164/checkpoints/00123... +INFO:tensorflow:global_step/sec: 1.33462 +INFO:tensorflow:step = 48101, loss = 0.179847, precision = 0.992188 (74.928 sec) +INFO:tensorflow:global_step/sec: 1.37441 +INFO:tensorflow:step = 48201, loss = 0.224352, precision = 0.984375 (72.759 sec) +INFO:tensorflow:global_step/sec: 1.37458 +INFO:tensorflow:step = 48301, loss = 0.197054, precision = 0.984375 (72.749 sec) +INFO:tensorflow:global_step/sec: 1.37477 +INFO:tensorflow:step = 48401, loss = 0.277586, precision = 0.960938 (72.740 sec) +Saved checkpoint after 124 epoch(s) to ../data/resnet164/checkpoints/00124... +INFO:tensorflow:global_step/sec: 1.33312 +INFO:tensorflow:step = 48501, loss = 0.175522, precision = 1.0 (75.012 sec) +INFO:tensorflow:global_step/sec: 1.37538 +INFO:tensorflow:step = 48601, loss = 0.260654, precision = 0.960938 (72.707 sec) +INFO:tensorflow:global_step/sec: 1.37665 +INFO:tensorflow:step = 48701, loss = 0.239347, precision = 0.96875 (72.640 sec) +INFO:tensorflow:global_step/sec: 1.37532 +INFO:tensorflow:step = 48801, loss = 0.220022, precision = 0.976562 (72.711 sec) +Saved checkpoint after 125 epoch(s) to ../data/resnet164/checkpoints/00125... +INFO:tensorflow:global_step/sec: 1.33606 +INFO:tensorflow:step = 48901, loss = 0.195232, precision = 0.992188 (74.847 sec) +INFO:tensorflow:global_step/sec: 1.37618 +INFO:tensorflow:step = 49001, loss = 0.227587, precision = 0.953125 (72.665 sec) +INFO:tensorflow:global_step/sec: 1.37688 +INFO:tensorflow:step = 49101, loss = 0.213443, precision = 0.96875 (72.628 sec) +INFO:tensorflow:global_step/sec: 1.37647 +INFO:tensorflow:step = 49201, loss = 0.17703, precision = 0.984375 (72.650 sec) +Saved checkpoint after 126 epoch(s) to ../data/resnet164/checkpoints/00126... +INFO:tensorflow:global_step/sec: 1.33115 +INFO:tensorflow:step = 49301, loss = 0.195522, precision = 0.976562 (75.123 sec) +INFO:tensorflow:global_step/sec: 1.37676 +INFO:tensorflow:step = 49401, loss = 0.221453, precision = 0.960938 (72.635 sec) +INFO:tensorflow:global_step/sec: 1.37535 +INFO:tensorflow:step = 49501, loss = 0.252095, precision = 0.960938 (72.708 sec) +INFO:tensorflow:global_step/sec: 1.37637 +INFO:tensorflow:step = 49601, loss = 0.16859, precision = 1.0 (72.655 sec) +Saved checkpoint after 127 epoch(s) to ../data/resnet164/checkpoints/00127... +INFO:tensorflow:global_step/sec: 1.33579 +INFO:tensorflow:step = 49701, loss = 0.220294, precision = 0.976562 (74.862 sec) +INFO:tensorflow:global_step/sec: 1.37593 +INFO:tensorflow:step = 49801, loss = 0.240632, precision = 0.960938 (72.678 sec) +INFO:tensorflow:global_step/sec: 1.37614 +INFO:tensorflow:step = 49901, loss = 0.197596, precision = 0.992188 (72.667 sec) +INFO:tensorflow:global_step/sec: 1.37657 +INFO:tensorflow:step = 50001, loss = 0.203103, precision = 0.976562 (72.645 sec) +Saved checkpoint after 128 epoch(s) to ../data/resnet164/checkpoints/00128... +INFO:tensorflow:global_step/sec: 1.33503 +INFO:tensorflow:step = 50101, loss = 0.210604, precision = 0.976562 (74.905 sec) +INFO:tensorflow:global_step/sec: 1.37659 +INFO:tensorflow:step = 50201, loss = 0.21301, precision = 0.976562 (72.644 sec) +INFO:tensorflow:global_step/sec: 1.37698 +INFO:tensorflow:step = 50301, loss = 0.235742, precision = 0.96875 (72.622 sec) +INFO:tensorflow:global_step/sec: 1.37674 +INFO:tensorflow:step = 50401, loss = 0.25704, precision = 0.976562 (72.635 sec) +Saved checkpoint after 129 epoch(s) to ../data/resnet164/checkpoints/00129... +INFO:tensorflow:global_step/sec: 1.33625 +INFO:tensorflow:step = 50501, loss = 0.20138, precision = 0.976562 (74.836 sec) +INFO:tensorflow:global_step/sec: 1.37653 +INFO:tensorflow:step = 50601, loss = 0.242423, precision = 0.953125 (72.647 sec) +INFO:tensorflow:global_step/sec: 1.37541 +INFO:tensorflow:step = 50701, loss = 0.170311, precision = 1.0 (72.705 sec) +INFO:tensorflow:global_step/sec: 1.37517 +INFO:tensorflow:step = 50801, loss = 0.181703, precision = 0.984375 (72.719 sec) +Saved checkpoint after 130 epoch(s) to ../data/resnet164/checkpoints/00130... +INFO:tensorflow:global_step/sec: 1.33404 +INFO:tensorflow:step = 50901, loss = 0.228196, precision = 0.984375 (74.960 sec) +INFO:tensorflow:global_step/sec: 1.37476 +INFO:tensorflow:step = 51001, loss = 0.196903, precision = 0.992188 (72.740 sec) +INFO:tensorflow:global_step/sec: 1.37532 +INFO:tensorflow:step = 51101, loss = 0.206214, precision = 0.976562 (72.710 sec) +INFO:tensorflow:global_step/sec: 1.37684 +INFO:tensorflow:step = 51201, loss = 0.223961, precision = 0.976562 (72.630 sec) +Saved checkpoint after 131 epoch(s) to ../data/resnet164/checkpoints/00131... +INFO:tensorflow:global_step/sec: 1.33755 +INFO:tensorflow:step = 51301, loss = 0.184607, precision = 0.984375 (74.763 sec) +INFO:tensorflow:global_step/sec: 1.37562 +INFO:tensorflow:step = 51401, loss = 0.228426, precision = 0.96875 (72.695 sec) +INFO:tensorflow:global_step/sec: 1.37711 +INFO:tensorflow:step = 51501, loss = 0.210701, precision = 0.96875 (72.616 sec) +INFO:tensorflow:global_step/sec: 1.37619 +INFO:tensorflow:step = 51601, loss = 0.23392, precision = 0.96875 (72.664 sec) +Saved checkpoint after 132 epoch(s) to ../data/resnet164/checkpoints/00132... +INFO:tensorflow:global_step/sec: 1.32987 +INFO:tensorflow:step = 51701, loss = 0.202458, precision = 0.976562 (75.195 sec) +INFO:tensorflow:global_step/sec: 1.37644 +INFO:tensorflow:step = 51801, loss = 0.173381, precision = 1.0 (72.652 sec) +INFO:tensorflow:global_step/sec: 1.37563 +INFO:tensorflow:step = 51901, loss = 0.214372, precision = 0.984375 (72.693 sec) +INFO:tensorflow:global_step/sec: 1.37625 +INFO:tensorflow:step = 52001, loss = 0.207286, precision = 0.96875 (72.661 sec) +Saved checkpoint after 133 epoch(s) to ../data/resnet164/checkpoints/00133... +INFO:tensorflow:global_step/sec: 1.33449 +INFO:tensorflow:step = 52101, loss = 0.15874, precision = 1.0 (74.935 sec) +INFO:tensorflow:global_step/sec: 1.37681 +INFO:tensorflow:step = 52201, loss = 0.160716, precision = 1.0 (72.632 sec) +INFO:tensorflow:global_step/sec: 1.37525 +INFO:tensorflow:step = 52301, loss = 0.223477, precision = 0.96875 (72.715 sec) +Saved checkpoint after 134 epoch(s) to ../data/resnet164/checkpoints/00134... +INFO:tensorflow:global_step/sec: 1.33619 +INFO:tensorflow:step = 52401, loss = 0.202668, precision = 0.984375 (74.840 sec) +INFO:tensorflow:global_step/sec: 1.37718 +INFO:tensorflow:step = 52501, loss = 0.172361, precision = 0.992188 (72.612 sec) +INFO:tensorflow:global_step/sec: 1.37598 +INFO:tensorflow:step = 52601, loss = 0.172799, precision = 0.992188 (72.676 sec) +INFO:tensorflow:global_step/sec: 1.37586 +INFO:tensorflow:step = 52701, loss = 0.178416, precision = 0.976562 (72.682 sec) +Saved checkpoint after 135 epoch(s) to ../data/resnet164/checkpoints/00135... +INFO:tensorflow:global_step/sec: 1.33528 +INFO:tensorflow:step = 52801, loss = 0.165431, precision = 0.992188 (74.891 sec) +INFO:tensorflow:global_step/sec: 1.37653 +INFO:tensorflow:step = 52901, loss = 0.185213, precision = 0.992188 (72.646 sec) +INFO:tensorflow:global_step/sec: 1.37649 +INFO:tensorflow:step = 53001, loss = 0.220356, precision = 0.976562 (72.648 sec) +INFO:tensorflow:global_step/sec: 1.37618 +INFO:tensorflow:step = 53101, loss = 0.198655, precision = 0.984375 (72.665 sec) +Saved checkpoint after 136 epoch(s) to ../data/resnet164/checkpoints/00136... +INFO:tensorflow:global_step/sec: 1.33511 +INFO:tensorflow:step = 53201, loss = 0.168323, precision = 1.0 (74.900 sec) +INFO:tensorflow:global_step/sec: 1.3759 +INFO:tensorflow:step = 53301, loss = 0.203759, precision = 0.976562 (72.680 sec) +INFO:tensorflow:global_step/sec: 1.37626 +INFO:tensorflow:step = 53401, loss = 0.164593, precision = 0.992188 (72.661 sec) +INFO:tensorflow:global_step/sec: 1.37706 +INFO:tensorflow:step = 53501, loss = 0.156677, precision = 1.0 (72.619 sec) +Saved checkpoint after 137 epoch(s) to ../data/resnet164/checkpoints/00137... +INFO:tensorflow:global_step/sec: 1.33585 +INFO:tensorflow:step = 53601, loss = 0.163951, precision = 1.0 (74.859 sec) +INFO:tensorflow:global_step/sec: 1.37615 +INFO:tensorflow:step = 53701, loss = 0.15642, precision = 1.0 (72.666 sec) +INFO:tensorflow:global_step/sec: 1.37554 +INFO:tensorflow:step = 53801, loss = 0.151564, precision = 1.0 (72.699 sec) +INFO:tensorflow:global_step/sec: 1.37437 +INFO:tensorflow:step = 53901, loss = 0.154357, precision = 1.0 (72.760 sec) +Saved checkpoint after 138 epoch(s) to ../data/resnet164/checkpoints/00138... +INFO:tensorflow:global_step/sec: 1.33736 +INFO:tensorflow:step = 54001, loss = 0.162756, precision = 0.992188 (74.774 sec) +INFO:tensorflow:global_step/sec: 1.37638 +INFO:tensorflow:step = 54101, loss = 0.153207, precision = 1.0 (72.655 sec) +INFO:tensorflow:global_step/sec: 1.37541 +INFO:tensorflow:step = 54201, loss = 0.150896, precision = 1.0 (72.706 sec) +INFO:tensorflow:global_step/sec: 1.3756 +INFO:tensorflow:step = 54301, loss = 0.149978, precision = 1.0 (72.696 sec) +Saved checkpoint after 139 epoch(s) to ../data/resnet164/checkpoints/00139... +INFO:tensorflow:global_step/sec: 1.32854 +INFO:tensorflow:step = 54401, loss = 0.158126, precision = 0.992188 (75.271 sec) +INFO:tensorflow:global_step/sec: 1.37447 +INFO:tensorflow:step = 54501, loss = 0.155425, precision = 1.0 (72.755 sec) +INFO:tensorflow:global_step/sec: 1.37342 +INFO:tensorflow:step = 54601, loss = 0.183816, precision = 0.984375 (72.811 sec) +INFO:tensorflow:global_step/sec: 1.37072 +INFO:tensorflow:step = 54701, loss = 0.153194, precision = 1.0 (72.954 sec) +Saved checkpoint after 140 epoch(s) to ../data/resnet164/checkpoints/00140... +INFO:tensorflow:global_step/sec: 1.3258 +INFO:tensorflow:step = 54801, loss = 0.171877, precision = 1.0 (75.426 sec) +INFO:tensorflow:global_step/sec: 1.37064 +INFO:tensorflow:step = 54901, loss = 0.150814, precision = 1.0 (72.959 sec) +INFO:tensorflow:global_step/sec: 1.37162 +INFO:tensorflow:step = 55001, loss = 0.153731, precision = 1.0 (72.907 sec) +INFO:tensorflow:global_step/sec: 1.37156 +INFO:tensorflow:step = 55101, loss = 0.1504, precision = 1.0 (72.910 sec) +Saved checkpoint after 141 epoch(s) to ../data/resnet164/checkpoints/00141... +INFO:tensorflow:global_step/sec: 1.32759 +INFO:tensorflow:step = 55201, loss = 0.15786, precision = 1.0 (75.325 sec) +INFO:tensorflow:global_step/sec: 1.3714 +INFO:tensorflow:step = 55301, loss = 0.182799, precision = 0.984375 (72.918 sec) +INFO:tensorflow:global_step/sec: 1.37104 +INFO:tensorflow:step = 55401, loss = 0.146483, precision = 1.0 (72.937 sec) +INFO:tensorflow:global_step/sec: 1.37042 +INFO:tensorflow:step = 55501, loss = 0.16491, precision = 0.984375 (72.970 sec) +Saved checkpoint after 142 epoch(s) to ../data/resnet164/checkpoints/00142... +INFO:tensorflow:global_step/sec: 1.32075 +INFO:tensorflow:step = 55601, loss = 0.151376, precision = 1.0 (75.714 sec) +INFO:tensorflow:global_step/sec: 1.37098 +INFO:tensorflow:step = 55701, loss = 0.150965, precision = 1.0 (72.941 sec) +INFO:tensorflow:global_step/sec: 1.37021 +INFO:tensorflow:step = 55801, loss = 0.147649, precision = 1.0 (72.982 sec) +INFO:tensorflow:global_step/sec: 1.37044 +INFO:tensorflow:step = 55901, loss = 0.144713, precision = 1.0 (72.970 sec) +Saved checkpoint after 143 epoch(s) to ../data/resnet164/checkpoints/00143... +INFO:tensorflow:global_step/sec: 1.32426 +INFO:tensorflow:step = 56001, loss = 0.142099, precision = 1.0 (75.513 sec) +INFO:tensorflow:global_step/sec: 1.37106 +INFO:tensorflow:step = 56101, loss = 0.143424, precision = 1.0 (72.937 sec) +INFO:tensorflow:global_step/sec: 1.37025 +INFO:tensorflow:step = 56201, loss = 0.143486, precision = 1.0 (72.980 sec) +INFO:tensorflow:global_step/sec: 1.37154 +INFO:tensorflow:step = 56301, loss = 0.141217, precision = 1.0 (72.910 sec) +Saved checkpoint after 144 epoch(s) to ../data/resnet164/checkpoints/00144... +INFO:tensorflow:global_step/sec: 1.32198 +INFO:tensorflow:step = 56401, loss = 0.143601, precision = 1.0 (75.644 sec) +INFO:tensorflow:global_step/sec: 1.37109 +INFO:tensorflow:step = 56501, loss = 0.142957, precision = 1.0 (72.934 sec) +INFO:tensorflow:global_step/sec: 1.37239 +INFO:tensorflow:step = 56601, loss = 0.14233, precision = 1.0 (72.866 sec) +Saved checkpoint after 145 epoch(s) to ../data/resnet164/checkpoints/00145... +INFO:tensorflow:global_step/sec: 1.3222 +INFO:tensorflow:step = 56701, loss = 0.144581, precision = 1.0 (75.631 sec) +INFO:tensorflow:global_step/sec: 1.37239 +INFO:tensorflow:step = 56801, loss = 0.173838, precision = 0.984375 (72.866 sec) +INFO:tensorflow:global_step/sec: 1.37147 +INFO:tensorflow:step = 56901, loss = 0.140901, precision = 1.0 (72.914 sec) +INFO:tensorflow:global_step/sec: 1.37109 +INFO:tensorflow:step = 57001, loss = 0.147701, precision = 1.0 (72.935 sec) +Saved checkpoint after 146 epoch(s) to ../data/resnet164/checkpoints/00146... +INFO:tensorflow:global_step/sec: 1.32365 +INFO:tensorflow:step = 57101, loss = 0.142454, precision = 1.0 (75.549 sec) +INFO:tensorflow:global_step/sec: 1.37062 +INFO:tensorflow:step = 57201, loss = 0.148876, precision = 1.0 (72.960 sec) +INFO:tensorflow:global_step/sec: 1.37119 +INFO:tensorflow:step = 57301, loss = 0.140081, precision = 1.0 (72.929 sec) +INFO:tensorflow:global_step/sec: 1.37075 +INFO:tensorflow:step = 57401, loss = 0.140242, precision = 1.0 (72.953 sec) +Saved checkpoint after 147 epoch(s) to ../data/resnet164/checkpoints/00147... +INFO:tensorflow:global_step/sec: 1.3235 +INFO:tensorflow:step = 57501, loss = 0.140508, precision = 1.0 (75.557 sec) +INFO:tensorflow:global_step/sec: 1.3711 +INFO:tensorflow:step = 57601, loss = 0.143952, precision = 1.0 (72.934 sec) +INFO:tensorflow:global_step/sec: 1.37067 +INFO:tensorflow:step = 57701, loss = 0.141609, precision = 1.0 (72.957 sec) +INFO:tensorflow:global_step/sec: 1.37068 +INFO:tensorflow:step = 57801, loss = 0.164796, precision = 0.992188 (72.957 sec) +Saved checkpoint after 148 epoch(s) to ../data/resnet164/checkpoints/00148... +INFO:tensorflow:global_step/sec: 1.32545 +INFO:tensorflow:step = 57901, loss = 0.139433, precision = 1.0 (75.446 sec) +INFO:tensorflow:global_step/sec: 1.37316 +INFO:tensorflow:step = 58001, loss = 0.13974, precision = 1.0 (72.825 sec) +INFO:tensorflow:global_step/sec: 1.37318 +INFO:tensorflow:step = 58101, loss = 0.140267, precision = 1.0 (72.823 sec) +INFO:tensorflow:global_step/sec: 1.37157 +INFO:tensorflow:step = 58201, loss = 0.139249, precision = 1.0 (72.909 sec) +Saved checkpoint after 149 epoch(s) to ../data/resnet164/checkpoints/00149... +INFO:tensorflow:global_step/sec: 1.32677 +INFO:tensorflow:step = 58301, loss = 0.142581, precision = 1.0 (75.371 sec) +INFO:tensorflow:global_step/sec: 1.37387 +INFO:tensorflow:step = 58401, loss = 0.139182, precision = 1.0 (72.787 sec) +INFO:tensorflow:global_step/sec: 1.37411 +INFO:tensorflow:step = 58501, loss = 0.142754, precision = 1.0 (72.774 sec) +INFO:tensorflow:global_step/sec: 1.37433 +INFO:tensorflow:step = 58601, loss = 0.139259, precision = 1.0 (72.763 sec) +Saved checkpoint after 150 epoch(s) to ../data/resnet164/checkpoints/00150... +INFO:tensorflow:global_step/sec: 1.33385 +INFO:tensorflow:step = 58701, loss = 0.147538, precision = 1.0 (74.971 sec) +INFO:tensorflow:global_step/sec: 1.37328 +INFO:tensorflow:step = 58801, loss = 0.137755, precision = 1.0 (72.818 sec) +INFO:tensorflow:global_step/sec: 1.37499 +INFO:tensorflow:step = 58901, loss = 0.14146, precision = 1.0 (72.728 sec) +INFO:tensorflow:global_step/sec: 1.37425 +INFO:tensorflow:step = 59001, loss = 0.141565, precision = 1.0 (72.766 sec) +Saved checkpoint after 151 epoch(s) to ../data/resnet164/checkpoints/00151... +INFO:tensorflow:global_step/sec: 1.33326 +INFO:tensorflow:step = 59101, loss = 0.138848, precision = 1.0 (75.004 sec) +INFO:tensorflow:global_step/sec: 1.37516 +INFO:tensorflow:step = 59201, loss = 0.138342, precision = 1.0 (72.719 sec) +INFO:tensorflow:global_step/sec: 1.37469 +INFO:tensorflow:step = 59301, loss = 0.141717, precision = 1.0 (72.744 sec) +INFO:tensorflow:global_step/sec: 1.37434 +INFO:tensorflow:step = 59401, loss = 0.142786, precision = 0.992188 (72.762 sec) +Saved checkpoint after 152 epoch(s) to ../data/resnet164/checkpoints/00152... +INFO:tensorflow:global_step/sec: 1.32189 +INFO:tensorflow:step = 59501, loss = 0.141649, precision = 1.0 (75.649 sec) +INFO:tensorflow:global_step/sec: 1.37363 +INFO:tensorflow:step = 59601, loss = 0.138406, precision = 1.0 (72.800 sec) +INFO:tensorflow:global_step/sec: 1.37294 +INFO:tensorflow:step = 59701, loss = 0.138563, precision = 1.0 (72.836 sec) +INFO:tensorflow:global_step/sec: 1.37312 +INFO:tensorflow:step = 59801, loss = 0.141515, precision = 1.0 (72.827 sec) +Saved checkpoint after 153 epoch(s) to ../data/resnet164/checkpoints/00153... +INFO:tensorflow:global_step/sec: 1.32856 +INFO:tensorflow:step = 59901, loss = 0.13736, precision = 1.0 (75.269 sec) +INFO:tensorflow:global_step/sec: 1.37247 +INFO:tensorflow:step = 60001, loss = 0.137421, precision = 1.0 (72.861 sec) +INFO:tensorflow:global_step/sec: 1.37278 +INFO:tensorflow:step = 60101, loss = 0.140505, precision = 1.0 (72.845 sec) +INFO:tensorflow:global_step/sec: 1.37268 +INFO:tensorflow:step = 60201, loss = 0.137316, precision = 1.0 (72.850 sec) +Saved checkpoint after 154 epoch(s) to ../data/resnet164/checkpoints/00154... +INFO:tensorflow:global_step/sec: 1.32727 +INFO:tensorflow:step = 60301, loss = 0.137673, precision = 1.0 (75.343 sec) +INFO:tensorflow:global_step/sec: 1.37313 +INFO:tensorflow:step = 60401, loss = 0.135725, precision = 1.0 (72.826 sec) +INFO:tensorflow:global_step/sec: 1.37328 +INFO:tensorflow:step = 60501, loss = 0.137388, precision = 1.0 (72.819 sec) +INFO:tensorflow:global_step/sec: 1.37242 +INFO:tensorflow:step = 60601, loss = 0.139099, precision = 1.0 (72.864 sec) +Saved checkpoint after 155 epoch(s) to ../data/resnet164/checkpoints/00155... +INFO:tensorflow:global_step/sec: 1.32666 +INFO:tensorflow:step = 60701, loss = 0.138227, precision = 1.0 (75.377 sec) +INFO:tensorflow:global_step/sec: 1.37168 +INFO:tensorflow:step = 60801, loss = 0.139211, precision = 1.0 (72.903 sec) +INFO:tensorflow:global_step/sec: 1.37296 +INFO:tensorflow:step = 60901, loss = 0.137272, precision = 1.0 (72.835 sec) +Saved checkpoint after 156 epoch(s) to ../data/resnet164/checkpoints/00156... +INFO:tensorflow:global_step/sec: 1.32279 +INFO:tensorflow:step = 61001, loss = 0.138848, precision = 1.0 (75.598 sec) +INFO:tensorflow:global_step/sec: 1.37257 +INFO:tensorflow:step = 61101, loss = 0.135212, precision = 1.0 (72.856 sec) +INFO:tensorflow:global_step/sec: 1.371 +INFO:tensorflow:step = 61201, loss = 0.135522, precision = 1.0 (72.939 sec) +INFO:tensorflow:global_step/sec: 1.37129 +INFO:tensorflow:step = 61301, loss = 0.134207, precision = 1.0 (72.924 sec) +Saved checkpoint after 157 epoch(s) to ../data/resnet164/checkpoints/00157... +INFO:tensorflow:global_step/sec: 1.32398 +INFO:tensorflow:step = 61401, loss = 0.13362, precision = 1.0 (75.530 sec) +INFO:tensorflow:global_step/sec: 1.37086 +INFO:tensorflow:step = 61501, loss = 0.135371, precision = 1.0 (72.947 sec) +INFO:tensorflow:global_step/sec: 1.37096 +INFO:tensorflow:step = 61601, loss = 0.133694, precision = 1.0 (72.942 sec) +INFO:tensorflow:global_step/sec: 1.37106 +INFO:tensorflow:step = 61701, loss = 0.134887, precision = 1.0 (72.936 sec) +Saved checkpoint after 158 epoch(s) to ../data/resnet164/checkpoints/00158... +INFO:tensorflow:global_step/sec: 1.32586 +INFO:tensorflow:step = 61801, loss = 0.134319, precision = 1.0 (75.423 sec) +INFO:tensorflow:global_step/sec: 1.37011 +INFO:tensorflow:step = 61901, loss = 0.136354, precision = 1.0 (72.987 sec) +INFO:tensorflow:global_step/sec: 1.37091 +INFO:tensorflow:step = 62001, loss = 0.13573, precision = 1.0 (72.945 sec) +INFO:tensorflow:global_step/sec: 1.37045 +INFO:tensorflow:step = 62101, loss = 0.134486, precision = 1.0 (72.968 sec) +Saved checkpoint after 159 epoch(s) to ../data/resnet164/checkpoints/00159... +INFO:tensorflow:global_step/sec: 1.32462 +INFO:tensorflow:step = 62201, loss = 0.133601, precision = 1.0 (75.493 sec) +INFO:tensorflow:global_step/sec: 1.37096 +INFO:tensorflow:step = 62301, loss = 0.134537, precision = 1.0 (72.942 sec) +INFO:tensorflow:global_step/sec: 1.37051 +INFO:tensorflow:step = 62401, loss = 0.133469, precision = 1.0 (72.966 sec) +INFO:tensorflow:global_step/sec: 1.37112 +INFO:tensorflow:step = 62501, loss = 0.137857, precision = 1.0 (72.933 sec) +Saved checkpoint after 160 epoch(s) to ../data/resnet164/checkpoints/00160... +INFO:tensorflow:global_step/sec: 1.32435 +INFO:tensorflow:step = 62601, loss = 0.134111, precision = 1.0 (75.509 sec) +INFO:tensorflow:global_step/sec: 1.37055 +INFO:tensorflow:step = 62701, loss = 0.133459, precision = 1.0 (72.963 sec) +INFO:tensorflow:global_step/sec: 1.37024 +INFO:tensorflow:step = 62801, loss = 0.137493, precision = 1.0 (72.980 sec) +INFO:tensorflow:global_step/sec: 1.37026 +INFO:tensorflow:step = 62901, loss = 0.137201, precision = 1.0 (72.979 sec) +Saved checkpoint after 161 epoch(s) to ../data/resnet164/checkpoints/00161... +INFO:tensorflow:global_step/sec: 1.32427 +INFO:tensorflow:step = 63001, loss = 0.132884, precision = 1.0 (75.513 sec) +INFO:tensorflow:global_step/sec: 1.37053 +INFO:tensorflow:step = 63101, loss = 0.143888, precision = 0.992188 (72.964 sec) +INFO:tensorflow:global_step/sec: 1.37011 +INFO:tensorflow:step = 63201, loss = 0.131535, precision = 1.0 (72.987 sec) +INFO:tensorflow:global_step/sec: 1.37047 +INFO:tensorflow:step = 63301, loss = 0.132539, precision = 1.0 (72.968 sec) +Saved checkpoint after 162 epoch(s) to ../data/resnet164/checkpoints/00162... +INFO:tensorflow:global_step/sec: 1.31718 +INFO:tensorflow:step = 63401, loss = 0.134075, precision = 1.0 (75.920 sec) +INFO:tensorflow:global_step/sec: 1.37047 +INFO:tensorflow:step = 63501, loss = 0.133518, precision = 1.0 (72.968 sec) +INFO:tensorflow:global_step/sec: 1.37009 +INFO:tensorflow:step = 63601, loss = 0.134949, precision = 1.0 (72.988 sec) +INFO:tensorflow:global_step/sec: 1.3711 +INFO:tensorflow:step = 63701, loss = 0.131021, precision = 1.0 (72.934 sec) +Saved checkpoint after 163 epoch(s) to ../data/resnet164/checkpoints/00163... +INFO:tensorflow:global_step/sec: 1.32606 +INFO:tensorflow:step = 63801, loss = 0.132371, precision = 1.0 (75.412 sec) +INFO:tensorflow:global_step/sec: 1.37047 +INFO:tensorflow:step = 63901, loss = 0.133143, precision = 1.0 (72.967 sec) +INFO:tensorflow:global_step/sec: 1.37164 +INFO:tensorflow:step = 64001, loss = 0.132044, precision = 1.0 (72.906 sec) +INFO:tensorflow:global_step/sec: 1.3701 +INFO:tensorflow:step = 64101, loss = 0.132153, precision = 1.0 (72.987 sec) +Saved checkpoint after 164 epoch(s) to ../data/resnet164/checkpoints/00164... +INFO:tensorflow:global_step/sec: 1.33008 +INFO:tensorflow:step = 64201, loss = 0.131443, precision = 1.0 (75.184 sec) +INFO:tensorflow:global_step/sec: 1.37037 +INFO:tensorflow:step = 64301, loss = 0.142675, precision = 1.0 (72.973 sec) +INFO:tensorflow:global_step/sec: 1.36941 +INFO:tensorflow:step = 64401, loss = 0.132099, precision = 1.0 (73.024 sec) +INFO:tensorflow:global_step/sec: 1.37012 +INFO:tensorflow:step = 64501, loss = 0.130072, precision = 1.0 (72.987 sec) +Saved checkpoint after 165 epoch(s) to ../data/resnet164/checkpoints/00165... +INFO:tensorflow:global_step/sec: 1.3237 +INFO:tensorflow:step = 64601, loss = 0.133413, precision = 1.0 (75.545 sec) +INFO:tensorflow:global_step/sec: 1.36957 +INFO:tensorflow:step = 64701, loss = 0.130218, precision = 1.0 (73.016 sec) +INFO:tensorflow:global_step/sec: 1.36993 +INFO:tensorflow:step = 64801, loss = 0.131524, precision = 1.0 (72.997 sec) +INFO:tensorflow:global_step/sec: 1.37007 +INFO:tensorflow:step = 64901, loss = 0.130337, precision = 1.0 (72.989 sec) +Saved checkpoint after 166 epoch(s) to ../data/resnet164/checkpoints/00166... +INFO:tensorflow:global_step/sec: 1.32449 +INFO:tensorflow:step = 65001, loss = 0.129754, precision = 1.0 (75.501 sec) +INFO:tensorflow:global_step/sec: 1.37006 +INFO:tensorflow:step = 65101, loss = 0.129856, precision = 1.0 (72.989 sec) +INFO:tensorflow:global_step/sec: 1.37221 +INFO:tensorflow:step = 65201, loss = 0.130564, precision = 1.0 (72.875 sec) +Saved checkpoint after 167 epoch(s) to ../data/resnet164/checkpoints/00167... +INFO:tensorflow:global_step/sec: 1.32376 +INFO:tensorflow:step = 65301, loss = 0.1345, precision = 1.0 (75.542 sec) +INFO:tensorflow:global_step/sec: 1.37096 +INFO:tensorflow:step = 65401, loss = 0.130159, precision = 1.0 (72.941 sec) +INFO:tensorflow:global_step/sec: 1.37053 +INFO:tensorflow:step = 65501, loss = 0.128931, precision = 1.0 (72.964 sec) +INFO:tensorflow:global_step/sec: 1.37052 +INFO:tensorflow:step = 65601, loss = 0.129526, precision = 1.0 (72.965 sec) +Saved checkpoint after 168 epoch(s) to ../data/resnet164/checkpoints/00168... +INFO:tensorflow:global_step/sec: 1.32186 +INFO:tensorflow:step = 65701, loss = 0.1284, precision = 1.0 (75.651 sec) +INFO:tensorflow:global_step/sec: 1.37111 +INFO:tensorflow:step = 65801, loss = 0.129839, precision = 1.0 (72.933 sec) +INFO:tensorflow:global_step/sec: 1.37145 +INFO:tensorflow:step = 65901, loss = 0.132518, precision = 1.0 (72.916 sec) +INFO:tensorflow:global_step/sec: 1.3718 +INFO:tensorflow:step = 66001, loss = 0.130273, precision = 1.0 (72.897 sec) +Saved checkpoint after 169 epoch(s) to ../data/resnet164/checkpoints/00169... +INFO:tensorflow:global_step/sec: 1.33102 +INFO:tensorflow:step = 66101, loss = 0.128927, precision = 1.0 (75.130 sec) +INFO:tensorflow:global_step/sec: 1.37434 +INFO:tensorflow:step = 66201, loss = 0.131662, precision = 1.0 (72.762 sec) +INFO:tensorflow:global_step/sec: 1.37367 +INFO:tensorflow:step = 66301, loss = 0.129682, precision = 1.0 (72.798 sec) +INFO:tensorflow:global_step/sec: 1.37311 +INFO:tensorflow:step = 66401, loss = 0.130323, precision = 1.0 (72.827 sec) +Saved checkpoint after 170 epoch(s) to ../data/resnet164/checkpoints/00170... +INFO:tensorflow:global_step/sec: 1.32837 +INFO:tensorflow:step = 66501, loss = 0.129022, precision = 1.0 (75.280 sec) +INFO:tensorflow:global_step/sec: 1.37321 +INFO:tensorflow:step = 66601, loss = 0.126942, precision = 1.0 (72.822 sec) +INFO:tensorflow:global_step/sec: 1.37384 +INFO:tensorflow:step = 66701, loss = 0.128382, precision = 1.0 (72.789 sec) +INFO:tensorflow:global_step/sec: 1.37332 +INFO:tensorflow:step = 66801, loss = 0.130771, precision = 1.0 (72.816 sec) +Saved checkpoint after 171 epoch(s) to ../data/resnet164/checkpoints/00171... +INFO:tensorflow:global_step/sec: 1.32974 +INFO:tensorflow:step = 66901, loss = 0.127779, precision = 1.0 (75.203 sec) +INFO:tensorflow:global_step/sec: 1.37139 +INFO:tensorflow:step = 67001, loss = 0.128616, precision = 1.0 (72.919 sec) +INFO:tensorflow:global_step/sec: 1.37072 +INFO:tensorflow:step = 67101, loss = 0.126759, precision = 1.0 (72.954 sec) +INFO:tensorflow:global_step/sec: 1.37085 +INFO:tensorflow:step = 67201, loss = 0.127235, precision = 1.0 (72.947 sec) +Saved checkpoint after 172 epoch(s) to ../data/resnet164/checkpoints/00172... +INFO:tensorflow:global_step/sec: 1.32158 +INFO:tensorflow:step = 67301, loss = 0.129113, precision = 1.0 (75.667 sec) +INFO:tensorflow:global_step/sec: 1.37098 +INFO:tensorflow:step = 67401, loss = 0.139109, precision = 0.992188 (72.941 sec) +INFO:tensorflow:global_step/sec: 1.37069 +INFO:tensorflow:step = 67501, loss = 0.12652, precision = 1.0 (72.956 sec) +INFO:tensorflow:global_step/sec: 1.37087 +INFO:tensorflow:step = 67601, loss = 0.127681, precision = 1.0 (72.946 sec) +Saved checkpoint after 173 epoch(s) to ../data/resnet164/checkpoints/00173... +INFO:tensorflow:global_step/sec: 1.32391 +INFO:tensorflow:step = 67701, loss = 0.126293, precision = 1.0 (75.534 sec) +INFO:tensorflow:global_step/sec: 1.3699 +INFO:tensorflow:step = 67801, loss = 0.125801, precision = 1.0 (72.998 sec) +INFO:tensorflow:global_step/sec: 1.37084 +INFO:tensorflow:step = 67901, loss = 0.127118, precision = 1.0 (72.948 sec) +INFO:tensorflow:global_step/sec: 1.37105 +INFO:tensorflow:step = 68001, loss = 0.126082, precision = 1.0 (72.937 sec) +Saved checkpoint after 174 epoch(s) to ../data/resnet164/checkpoints/00174... +INFO:tensorflow:global_step/sec: 1.32363 +INFO:tensorflow:step = 68101, loss = 0.12545, precision = 1.0 (75.550 sec) +INFO:tensorflow:global_step/sec: 1.37105 +INFO:tensorflow:step = 68201, loss = 0.125747, precision = 1.0 (72.937 sec) +INFO:tensorflow:global_step/sec: 1.37148 +INFO:tensorflow:step = 68301, loss = 0.125237, precision = 1.0 (72.914 sec) +INFO:tensorflow:global_step/sec: 1.37016 +INFO:tensorflow:step = 68401, loss = 0.12716, precision = 1.0 (72.984 sec) +Saved checkpoint after 175 epoch(s) to ../data/resnet164/checkpoints/00175... +INFO:tensorflow:global_step/sec: 1.32119 +INFO:tensorflow:step = 68501, loss = 0.133327, precision = 1.0 (75.690 sec) +INFO:tensorflow:global_step/sec: 1.36955 +INFO:tensorflow:step = 68601, loss = 0.125252, precision = 1.0 (73.017 sec) +INFO:tensorflow:global_step/sec: 1.37167 +INFO:tensorflow:step = 68701, loss = 0.12557, precision = 1.0 (72.904 sec) +INFO:tensorflow:global_step/sec: 1.36944 +INFO:tensorflow:step = 68801, loss = 0.125778, precision = 1.0 (73.023 sec) +Saved checkpoint after 176 epoch(s) to ../data/resnet164/checkpoints/00176... +INFO:tensorflow:global_step/sec: 1.32362 +INFO:tensorflow:step = 68901, loss = 0.124342, precision = 1.0 (75.550 sec) +INFO:tensorflow:global_step/sec: 1.3699 +INFO:tensorflow:step = 69001, loss = 0.13291, precision = 1.0 (72.998 sec) +INFO:tensorflow:global_step/sec: 1.37074 +INFO:tensorflow:step = 69101, loss = 0.124675, precision = 1.0 (72.953 sec) +INFO:tensorflow:global_step/sec: 1.37035 +INFO:tensorflow:step = 69201, loss = 0.125703, precision = 1.0 (72.974 sec) +Saved checkpoint after 177 epoch(s) to ../data/resnet164/checkpoints/00177... +INFO:tensorflow:global_step/sec: 1.32607 +INFO:tensorflow:step = 69301, loss = 0.12395, precision = 1.0 (75.411 sec) +INFO:tensorflow:global_step/sec: 1.3713 +INFO:tensorflow:step = 69401, loss = 0.12452, precision = 1.0 (72.924 sec) +INFO:tensorflow:global_step/sec: 1.37301 +INFO:tensorflow:step = 69501, loss = 0.124592, precision = 1.0 (72.832 sec) +Saved checkpoint after 178 epoch(s) to ../data/resnet164/checkpoints/00178... +INFO:tensorflow:global_step/sec: 1.32337 +INFO:tensorflow:step = 69601, loss = 0.12426, precision = 1.0 (75.565 sec) +INFO:tensorflow:global_step/sec: 1.37249 +INFO:tensorflow:step = 69701, loss = 0.124209, precision = 1.0 (72.860 sec) +INFO:tensorflow:global_step/sec: 1.37175 +INFO:tensorflow:step = 69801, loss = 0.124242, precision = 1.0 (72.900 sec) +INFO:tensorflow:global_step/sec: 1.37391 +INFO:tensorflow:step = 69901, loss = 0.123283, precision = 1.0 (72.785 sec) +Saved checkpoint after 179 epoch(s) to ../data/resnet164/checkpoints/00179... +INFO:tensorflow:global_step/sec: 1.33072 +INFO:tensorflow:step = 70001, loss = 0.124132, precision = 1.0 (75.148 sec) +INFO:tensorflow:global_step/sec: 1.37216 +INFO:tensorflow:step = 70101, loss = 0.124292, precision = 1.0 (72.877 sec) +INFO:tensorflow:global_step/sec: 1.37333 +INFO:tensorflow:step = 70201, loss = 0.128087, precision = 1.0 (72.816 sec) +INFO:tensorflow:global_step/sec: 1.37423 +INFO:tensorflow:step = 70301, loss = 0.122918, precision = 1.0 (72.768 sec) +Saved checkpoint after 180 epoch(s) to ../data/resnet164/checkpoints/00180... +INFO:tensorflow:global_step/sec: 1.32761 +INFO:tensorflow:step = 70401, loss = 0.123988, precision = 1.0 (75.324 sec) +INFO:tensorflow:global_step/sec: 1.37443 +INFO:tensorflow:step = 70501, loss = 0.122924, precision = 1.0 (72.757 sec) +INFO:tensorflow:global_step/sec: 1.3735 +INFO:tensorflow:step = 70601, loss = 0.124969, precision = 1.0 (72.807 sec) +INFO:tensorflow:global_step/sec: 1.37463 +INFO:tensorflow:step = 70701, loss = 0.125923, precision = 1.0 (72.747 sec) +Saved checkpoint after 181 epoch(s) to ../data/resnet164/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1k80_gc/resnet164_nb_train.log b/tensorflow/CIFAR10/logs/1k80_gc/resnet164_nb_train.log new file mode 100644 index 0000000..0024cbd --- /dev/null +++ b/tensorflow/CIFAR10/logs/1k80_gc/resnet164_nb_train.log @@ -0,0 +1,2162 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/2.60m params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_10/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/sub2/bn2/beta (16, 16/16 params) + unit_1_10/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_11/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/sub2/bn2/beta (16, 16/16 params) + unit_1_11/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_12/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/sub2/bn2/beta (16, 16/16 params) + unit_1_12/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_13/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/sub2/bn2/beta (16, 16/16 params) + unit_1_13/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_14/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/sub2/bn2/beta (16, 16/16 params) + unit_1_14/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_15/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/sub2/bn2/beta (16, 16/16 params) + unit_1_15/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_16/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/sub2/bn2/beta (16, 16/16 params) + unit_1_16/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_17/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/sub2/bn2/beta (16, 16/16 params) + unit_1_17/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_18/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_18/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_18/sub2/bn2/beta (16, 16/16 params) + unit_1_18/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_19/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_19/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_19/sub2/bn2/beta (16, 16/16 params) + unit_1_19/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_20/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_20/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_20/sub2/bn2/beta (16, 16/16 params) + unit_1_20/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_21/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_21/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_21/sub2/bn2/beta (16, 16/16 params) + unit_1_21/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_22/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_22/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_22/sub2/bn2/beta (16, 16/16 params) + unit_1_22/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_23/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_23/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_23/sub2/bn2/beta (16, 16/16 params) + unit_1_23/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_24/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_24/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_24/sub2/bn2/beta (16, 16/16 params) + unit_1_24/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_25/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_25/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_25/sub2/bn2/beta (16, 16/16 params) + unit_1_25/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_26/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_26/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_26/sub2/bn2/beta (16, 16/16 params) + unit_1_26/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_3/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_4/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_5/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_6/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_7/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_8/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_9/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/sub2/bn2/beta (16, 16/16 params) + unit_1_9/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_10/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/sub2/bn2/beta (32, 32/32 params) + unit_2_10/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_11/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/sub2/bn2/beta (32, 32/32 params) + unit_2_11/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_12/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/sub2/bn2/beta (32, 32/32 params) + unit_2_12/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_13/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/sub2/bn2/beta (32, 32/32 params) + unit_2_13/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_14/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/sub2/bn2/beta (32, 32/32 params) + unit_2_14/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_15/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/sub2/bn2/beta (32, 32/32 params) + unit_2_15/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_16/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/sub2/bn2/beta (32, 32/32 params) + unit_2_16/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_17/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/sub2/bn2/beta (32, 32/32 params) + unit_2_17/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_18/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_18/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_18/sub2/bn2/beta (32, 32/32 params) + unit_2_18/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_19/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_19/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_19/sub2/bn2/beta (32, 32/32 params) + unit_2_19/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_20/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_20/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_20/sub2/bn2/beta (32, 32/32 params) + unit_2_20/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_21/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_21/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_21/sub2/bn2/beta (32, 32/32 params) + unit_2_21/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_22/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_22/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_22/sub2/bn2/beta (32, 32/32 params) + unit_2_22/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_23/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_23/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_23/sub2/bn2/beta (32, 32/32 params) + unit_2_23/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_24/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_24/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_24/sub2/bn2/beta (32, 32/32 params) + unit_2_24/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_25/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_25/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_25/sub2/bn2/beta (32, 32/32 params) + unit_2_25/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_26/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_26/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_26/sub2/bn2/beta (32, 32/32 params) + unit_2_26/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_3/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_4/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_5/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_6/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_7/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_8/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_9/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/sub2/bn2/beta (32, 32/32 params) + unit_2_9/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_10/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/sub2/bn2/beta (64, 64/64 params) + unit_3_10/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_11/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/sub2/bn2/beta (64, 64/64 params) + unit_3_11/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_12/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/sub2/bn2/beta (64, 64/64 params) + unit_3_12/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_13/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/sub2/bn2/beta (64, 64/64 params) + unit_3_13/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_14/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/sub2/bn2/beta (64, 64/64 params) + unit_3_14/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_15/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/sub2/bn2/beta (64, 64/64 params) + unit_3_15/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_16/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/sub2/bn2/beta (64, 64/64 params) + unit_3_16/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_17/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/sub2/bn2/beta (64, 64/64 params) + unit_3_17/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_18/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_18/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_18/sub2/bn2/beta (64, 64/64 params) + unit_3_18/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_19/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_19/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_19/sub2/bn2/beta (64, 64/64 params) + unit_3_19/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_20/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_20/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_20/sub2/bn2/beta (64, 64/64 params) + unit_3_20/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_21/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_21/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_21/sub2/bn2/beta (64, 64/64 params) + unit_3_21/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_22/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_22/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_22/sub2/bn2/beta (64, 64/64 params) + unit_3_22/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_23/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_23/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_23/sub2/bn2/beta (64, 64/64 params) + unit_3_23/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_24/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_24/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_24/sub2/bn2/beta (64, 64/64 params) + unit_3_24/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_25/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_25/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_25/sub2/bn2/beta (64, 64/64 params) + unit_3_25/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_26/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_26/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_26/sub2/bn2/beta (64, 64/64 params) + unit_3_26/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_3/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_4/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_5/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_6/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_7/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_8/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_9/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/sub2/bn2/beta (64, 64/64 params) + unit_3_9/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/97.35b flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + +======================End of Report========================== +2017-07-30 00:28:29.144415: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero +2017-07-30 00:28:29.145195: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla K80 +major: 3 minor: 7 memoryClockRate (GHz) 0.8235 +pciBusID 0000:00:04.0 +Total memory: 11.17GiB +Free memory: 11.09GiB +2017-07-30 00:28:29.145226: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-07-30 00:28:29.145238: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-07-30 00:28:29.145249: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:00:04.0) +2017-07-30 00:28:30.604724: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 00:28:30.604774: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 8 visible devices +2017-07-30 00:28:30.607775: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0xb0cc220 executing computations on platform Host. Devices: +2017-07-30 00:28:30.607801: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-07-30 00:28:30.608613: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 00:28:30.608639: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 8 visible devices +2017-07-30 00:28:30.609244: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x6e089a0 executing computations on platform CUDA. Devices: +2017-07-30 00:28:30.609267: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla K80, Compute Capability 3.7 +2017-07-30 00:28:31.673700: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:247] PoolAllocator: After 1146 get requests, put_count=1100 evicted_count=1000 eviction_rate=0.909091 and unsatisfied allocation rate=1 +2017-07-30 00:28:31.673803: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:259] Raising pool_size_limit_ from 100 to 110 +INFO:tensorflow:step = 1, loss = 5.42529, precision = 0.101562 +INFO:tensorflow:global_step/sec: 1.52998 +INFO:tensorflow:step = 101, loss = 4.70481, precision = 0.398438 (65.361 sec) +INFO:tensorflow:global_step/sec: 1.55454 +INFO:tensorflow:step = 201, loss = 4.44921, precision = 0.445312 (64.328 sec) +INFO:tensorflow:global_step/sec: 1.55422 +INFO:tensorflow:step = 301, loss = 4.40707, precision = 0.40625 (64.341 sec) +total_params: 2596842 +Saved checkpoint after 1 epoch(s) to ../data/resnet164/checkpoints/00001... +INFO:tensorflow:global_step/sec: 1.49712 +INFO:tensorflow:step = 401, loss = 4.45006, precision = 0.359375 (66.795 sec) +INFO:tensorflow:global_step/sec: 1.55408 +INFO:tensorflow:step = 501, loss = 3.7671, precision = 0.617188 (64.346 sec) +INFO:tensorflow:global_step/sec: 1.55268 +INFO:tensorflow:step = 601, loss = 3.57693, precision = 0.546875 (64.405 sec) +INFO:tensorflow:global_step/sec: 1.55152 +INFO:tensorflow:step = 701, loss = 3.20708, precision = 0.648438 (64.453 sec) +Saved checkpoint after 2 epoch(s) to ../data/resnet164/checkpoints/00002... +INFO:tensorflow:global_step/sec: 1.50092 +INFO:tensorflow:step = 801, loss = 2.98232, precision = 0.640625 (66.626 sec) +INFO:tensorflow:global_step/sec: 1.55239 +INFO:tensorflow:step = 901, loss = 2.84318, precision = 0.59375 (64.417 sec) +INFO:tensorflow:global_step/sec: 1.55254 +INFO:tensorflow:step = 1001, loss = 2.64989, precision = 0.671875 (64.411 sec) +INFO:tensorflow:global_step/sec: 1.55253 +INFO:tensorflow:step = 1101, loss = 2.33754, precision = 0.742188 (64.411 sec) +Saved checkpoint after 3 epoch(s) to ../data/resnet164/checkpoints/00003... +INFO:tensorflow:global_step/sec: 1.50473 +INFO:tensorflow:step = 1201, loss = 2.12816, precision = 0.742188 (66.458 sec) +INFO:tensorflow:global_step/sec: 1.55226 +INFO:tensorflow:step = 1301, loss = 2.12693, precision = 0.726562 (64.422 sec) +INFO:tensorflow:global_step/sec: 1.55279 +INFO:tensorflow:step = 1401, loss = 1.90838, precision = 0.734375 (64.400 sec) +INFO:tensorflow:global_step/sec: 1.55198 +INFO:tensorflow:step = 1501, loss = 1.77818, precision = 0.765625 (64.434 sec) +Saved checkpoint after 4 epoch(s) to ../data/resnet164/checkpoints/00004... +INFO:tensorflow:global_step/sec: 1.50354 +INFO:tensorflow:step = 1601, loss = 1.50308, precision = 0.84375 (66.510 sec) +INFO:tensorflow:global_step/sec: 1.55314 +INFO:tensorflow:step = 1701, loss = 1.66468, precision = 0.773438 (64.385 sec) +INFO:tensorflow:global_step/sec: 1.5513 +INFO:tensorflow:step = 1801, loss = 1.48698, precision = 0.8125 (64.462 sec) +INFO:tensorflow:global_step/sec: 1.5527 +INFO:tensorflow:step = 1901, loss = 1.45161, precision = 0.804688 (64.404 sec) +Saved checkpoint after 5 epoch(s) to ../data/resnet164/checkpoints/00005... +INFO:tensorflow:global_step/sec: 1.50271 +INFO:tensorflow:step = 2001, loss = 1.42672, precision = 0.78125 (66.547 sec) +INFO:tensorflow:global_step/sec: 1.55314 +INFO:tensorflow:step = 2101, loss = 1.23342, precision = 0.835938 (64.386 sec) +INFO:tensorflow:global_step/sec: 1.55227 +INFO:tensorflow:step = 2201, loss = 1.29977, precision = 0.742188 (64.422 sec) +INFO:tensorflow:global_step/sec: 1.5506 +INFO:tensorflow:step = 2301, loss = 1.47773, precision = 0.726562 (64.491 sec) +Saved checkpoint after 6 epoch(s) to ../data/resnet164/checkpoints/00006... +INFO:tensorflow:global_step/sec: 1.50356 +INFO:tensorflow:step = 2401, loss = 1.19154, precision = 0.804688 (66.509 sec) +INFO:tensorflow:global_step/sec: 1.55232 +INFO:tensorflow:step = 2501, loss = 1.00647, precision = 0.851562 (64.419 sec) +INFO:tensorflow:global_step/sec: 1.55459 +INFO:tensorflow:step = 2601, loss = 1.16307, precision = 0.765625 (64.326 sec) +INFO:tensorflow:global_step/sec: 1.55509 +INFO:tensorflow:step = 2701, loss = 1.05025, precision = 0.8125 (64.305 sec) +Saved checkpoint after 7 epoch(s) to ../data/resnet164/checkpoints/00007... +INFO:tensorflow:global_step/sec: 1.50444 +INFO:tensorflow:step = 2801, loss = 0.972763, precision = 0.820312 (66.470 sec) +INFO:tensorflow:global_step/sec: 1.55358 +INFO:tensorflow:step = 2901, loss = 1.11477, precision = 0.8125 (64.367 sec) +INFO:tensorflow:global_step/sec: 1.55293 +INFO:tensorflow:step = 3001, loss = 0.872044, precision = 0.867188 (64.394 sec) +INFO:tensorflow:global_step/sec: 1.55377 +INFO:tensorflow:step = 3101, loss = 0.876646, precision = 0.851562 (64.360 sec) +Saved checkpoint after 8 epoch(s) to ../data/resnet164/checkpoints/00008... +INFO:tensorflow:global_step/sec: 1.49955 +INFO:tensorflow:step = 3201, loss = 0.924231, precision = 0.828125 (66.687 sec) +INFO:tensorflow:global_step/sec: 1.55322 +INFO:tensorflow:step = 3301, loss = 0.768556, precision = 0.898438 (64.382 sec) +INFO:tensorflow:global_step/sec: 1.55366 +INFO:tensorflow:step = 3401, loss = 1.09788, precision = 0.78125 (64.364 sec) +INFO:tensorflow:global_step/sec: 1.55585 +INFO:tensorflow:step = 3501, loss = 0.980874, precision = 0.78125 (64.274 sec) +Saved checkpoint after 9 epoch(s) to ../data/resnet164/checkpoints/00009... +INFO:tensorflow:global_step/sec: 1.50448 +INFO:tensorflow:step = 3601, loss = 0.852686, precision = 0.84375 (66.468 sec) +INFO:tensorflow:global_step/sec: 1.55633 +INFO:tensorflow:step = 3701, loss = 0.988412, precision = 0.78125 (64.253 sec) +INFO:tensorflow:global_step/sec: 1.55556 +INFO:tensorflow:step = 3801, loss = 0.746781, precision = 0.867188 (64.286 sec) +INFO:tensorflow:global_step/sec: 1.55596 +INFO:tensorflow:step = 3901, loss = 0.768117, precision = 0.851562 (64.269 sec) +Saved checkpoint after 10 epoch(s) to ../data/resnet164/checkpoints/00010... +INFO:tensorflow:global_step/sec: 1.50639 +INFO:tensorflow:step = 4001, loss = 0.865338, precision = 0.8125 (66.384 sec) +INFO:tensorflow:global_step/sec: 1.55666 +INFO:tensorflow:step = 4101, loss = 0.903653, precision = 0.804688 (64.240 sec) +INFO:tensorflow:global_step/sec: 1.55493 +INFO:tensorflow:step = 4201, loss = 0.845447, precision = 0.84375 (64.311 sec) +Saved checkpoint after 11 epoch(s) to ../data/resnet164/checkpoints/00011... +INFO:tensorflow:global_step/sec: 1.50585 +INFO:tensorflow:step = 4301, loss = 0.75678, precision = 0.859375 (66.408 sec) +INFO:tensorflow:global_step/sec: 1.55539 +INFO:tensorflow:step = 4401, loss = 0.885439, precision = 0.804688 (64.292 sec) +INFO:tensorflow:global_step/sec: 1.55521 +INFO:tensorflow:step = 4501, loss = 0.677788, precision = 0.890625 (64.300 sec) +INFO:tensorflow:global_step/sec: 1.55545 +INFO:tensorflow:step = 4601, loss = 0.7357, precision = 0.867188 (64.290 sec) +Saved checkpoint after 12 epoch(s) to ../data/resnet164/checkpoints/00012... +INFO:tensorflow:global_step/sec: 1.50571 +INFO:tensorflow:step = 4701, loss = 0.802703, precision = 0.828125 (66.414 sec) +INFO:tensorflow:global_step/sec: 1.55547 +INFO:tensorflow:step = 4801, loss = 0.743804, precision = 0.828125 (64.289 sec) +INFO:tensorflow:global_step/sec: 1.55582 +INFO:tensorflow:step = 4901, loss = 0.806177, precision = 0.8125 (64.275 sec) +INFO:tensorflow:global_step/sec: 1.55474 +INFO:tensorflow:step = 5001, loss = 0.80863, precision = 0.828125 (64.320 sec) +Saved checkpoint after 13 epoch(s) to ../data/resnet164/checkpoints/00013... +INFO:tensorflow:global_step/sec: 1.50663 +INFO:tensorflow:step = 5101, loss = 0.775837, precision = 0.84375 (66.373 sec) +INFO:tensorflow:global_step/sec: 1.55525 +INFO:tensorflow:step = 5201, loss = 0.758603, precision = 0.820312 (64.298 sec) +INFO:tensorflow:global_step/sec: 1.55541 +INFO:tensorflow:step = 5301, loss = 0.696905, precision = 0.851562 (64.292 sec) +INFO:tensorflow:global_step/sec: 1.55542 +INFO:tensorflow:step = 5401, loss = 0.754018, precision = 0.875 (64.291 sec) +Saved checkpoint after 14 epoch(s) to ../data/resnet164/checkpoints/00014... +INFO:tensorflow:global_step/sec: 1.50546 +INFO:tensorflow:step = 5501, loss = 0.714807, precision = 0.859375 (66.425 sec) +INFO:tensorflow:global_step/sec: 1.55512 +INFO:tensorflow:step = 5601, loss = 0.849768, precision = 0.820312 (64.304 sec) +INFO:tensorflow:global_step/sec: 1.55534 +INFO:tensorflow:step = 5701, loss = 0.677599, precision = 0.882812 (64.294 sec) +INFO:tensorflow:global_step/sec: 1.55534 +INFO:tensorflow:step = 5801, loss = 0.723746, precision = 0.875 (64.295 sec) +Saved checkpoint after 15 epoch(s) to ../data/resnet164/checkpoints/00015... +INFO:tensorflow:global_step/sec: 1.50614 +INFO:tensorflow:step = 5901, loss = 0.710693, precision = 0.84375 (66.395 sec) +INFO:tensorflow:global_step/sec: 1.55427 +INFO:tensorflow:step = 6001, loss = 0.750072, precision = 0.835938 (64.339 sec) +INFO:tensorflow:global_step/sec: 1.55446 +INFO:tensorflow:step = 6101, loss = 0.869223, precision = 0.820312 (64.331 sec) +INFO:tensorflow:global_step/sec: 1.55495 +INFO:tensorflow:step = 6201, loss = 0.726417, precision = 0.835938 (64.311 sec) +Saved checkpoint after 16 epoch(s) to ../data/resnet164/checkpoints/00016... +INFO:tensorflow:global_step/sec: 1.50317 +INFO:tensorflow:step = 6301, loss = 0.671229, precision = 0.859375 (66.526 sec) +INFO:tensorflow:global_step/sec: 1.55479 +INFO:tensorflow:step = 6401, loss = 0.752852, precision = 0.820312 (64.317 sec) +INFO:tensorflow:global_step/sec: 1.55403 +INFO:tensorflow:step = 6501, loss = 0.721325, precision = 0.84375 (64.348 sec) +INFO:tensorflow:global_step/sec: 1.55507 +INFO:tensorflow:step = 6601, loss = 0.658973, precision = 0.875 (64.306 sec) +Saved checkpoint after 17 epoch(s) to ../data/resnet164/checkpoints/00017... +INFO:tensorflow:global_step/sec: 1.50626 +INFO:tensorflow:step = 6701, loss = 0.907062, precision = 0.789062 (66.390 sec) +INFO:tensorflow:global_step/sec: 1.55464 +INFO:tensorflow:step = 6801, loss = 0.775092, precision = 0.828125 (64.324 sec) +INFO:tensorflow:global_step/sec: 1.55317 +INFO:tensorflow:step = 6901, loss = 0.709242, precision = 0.867188 (64.385 sec) +INFO:tensorflow:global_step/sec: 1.55491 +INFO:tensorflow:step = 7001, loss = 0.787196, precision = 0.828125 (64.313 sec) +Saved checkpoint after 18 epoch(s) to ../data/resnet164/checkpoints/00018... +INFO:tensorflow:global_step/sec: 1.50157 +INFO:tensorflow:step = 7101, loss = 0.710007, precision = 0.867188 (66.597 sec) +INFO:tensorflow:global_step/sec: 1.55577 +INFO:tensorflow:step = 7201, loss = 0.797034, precision = 0.828125 (64.277 sec) +INFO:tensorflow:global_step/sec: 1.55378 +INFO:tensorflow:step = 7301, loss = 0.726048, precision = 0.835938 (64.359 sec) +INFO:tensorflow:global_step/sec: 1.55399 +INFO:tensorflow:step = 7401, loss = 0.546889, precision = 0.890625 (64.351 sec) +Saved checkpoint after 19 epoch(s) to ../data/resnet164/checkpoints/00019... +INFO:tensorflow:global_step/sec: 1.50289 +INFO:tensorflow:step = 7501, loss = 0.756644, precision = 0.8125 (66.539 sec) +INFO:tensorflow:global_step/sec: 1.55358 +INFO:tensorflow:step = 7601, loss = 0.805381, precision = 0.851562 (64.368 sec) +INFO:tensorflow:global_step/sec: 1.55455 +INFO:tensorflow:step = 7701, loss = 0.615294, precision = 0.867188 (64.327 sec) +INFO:tensorflow:global_step/sec: 1.55348 +INFO:tensorflow:step = 7801, loss = 0.696616, precision = 0.84375 (64.372 sec) +Saved checkpoint after 20 epoch(s) to ../data/resnet164/checkpoints/00020... +INFO:tensorflow:global_step/sec: 1.50627 +INFO:tensorflow:step = 7901, loss = 0.611861, precision = 0.898438 (66.389 sec) +INFO:tensorflow:global_step/sec: 1.55482 +INFO:tensorflow:step = 8001, loss = 0.675021, precision = 0.859375 (64.316 sec) +INFO:tensorflow:global_step/sec: 1.5548 +INFO:tensorflow:step = 8101, loss = 0.710037, precision = 0.835938 (64.317 sec) +INFO:tensorflow:global_step/sec: 1.55365 +INFO:tensorflow:step = 8201, loss = 0.765332, precision = 0.882812 (64.365 sec) +Saved checkpoint after 21 epoch(s) to ../data/resnet164/checkpoints/00021... +INFO:tensorflow:global_step/sec: 1.506 +INFO:tensorflow:step = 8301, loss = 0.824056, precision = 0.820312 (66.401 sec) +INFO:tensorflow:global_step/sec: 1.55407 +INFO:tensorflow:step = 8401, loss = 0.616776, precision = 0.898438 (64.347 sec) +INFO:tensorflow:global_step/sec: 1.55406 +INFO:tensorflow:step = 8501, loss = 0.738488, precision = 0.84375 (64.348 sec) +INFO:tensorflow:global_step/sec: 1.55529 +INFO:tensorflow:step = 8601, loss = 0.803024, precision = 0.804688 (64.297 sec) +Saved checkpoint after 22 epoch(s) to ../data/resnet164/checkpoints/00022... +INFO:tensorflow:global_step/sec: 1.50707 +INFO:tensorflow:step = 8701, loss = 0.682789, precision = 0.828125 (66.354 sec) +INFO:tensorflow:global_step/sec: 1.55441 +INFO:tensorflow:step = 8801, loss = 0.678241, precision = 0.851562 (64.333 sec) +INFO:tensorflow:global_step/sec: 1.5541 +INFO:tensorflow:step = 8901, loss = 0.80119, precision = 0.835938 (64.346 sec) +Saved checkpoint after 23 epoch(s) to ../data/resnet164/checkpoints/00023... +INFO:tensorflow:global_step/sec: 1.50589 +INFO:tensorflow:step = 9001, loss = 0.587193, precision = 0.882812 (66.406 sec) +INFO:tensorflow:global_step/sec: 1.55513 +INFO:tensorflow:step = 9101, loss = 0.703558, precision = 0.859375 (64.303 sec) +INFO:tensorflow:global_step/sec: 1.55547 +INFO:tensorflow:step = 9201, loss = 0.573236, precision = 0.914062 (64.289 sec) +INFO:tensorflow:global_step/sec: 1.55363 +INFO:tensorflow:step = 9301, loss = 0.721798, precision = 0.820312 (64.365 sec) +Saved checkpoint after 24 epoch(s) to ../data/resnet164/checkpoints/00024... +INFO:tensorflow:global_step/sec: 1.50563 +INFO:tensorflow:step = 9401, loss = 0.692114, precision = 0.890625 (66.417 sec) +INFO:tensorflow:global_step/sec: 1.55494 +INFO:tensorflow:step = 9501, loss = 0.65609, precision = 0.875 (64.311 sec) +INFO:tensorflow:global_step/sec: 1.55497 +INFO:tensorflow:step = 9601, loss = 0.64168, precision = 0.882812 (64.310 sec) +INFO:tensorflow:global_step/sec: 1.55424 +INFO:tensorflow:step = 9701, loss = 0.748816, precision = 0.859375 (64.340 sec) +Saved checkpoint after 25 epoch(s) to ../data/resnet164/checkpoints/00025... +INFO:tensorflow:global_step/sec: 1.50657 +INFO:tensorflow:step = 9801, loss = 0.616768, precision = 0.882812 (66.376 sec) +INFO:tensorflow:global_step/sec: 1.55477 +INFO:tensorflow:step = 9901, loss = 0.774183, precision = 0.851562 (64.318 sec) +INFO:tensorflow:global_step/sec: 1.5566 +INFO:tensorflow:step = 10001, loss = 0.56189, precision = 0.914062 (64.243 sec) +INFO:tensorflow:global_step/sec: 1.55444 +INFO:tensorflow:step = 10101, loss = 0.655237, precision = 0.867188 (64.332 sec) +Saved checkpoint after 26 epoch(s) to ../data/resnet164/checkpoints/00026... +INFO:tensorflow:global_step/sec: 1.50654 +INFO:tensorflow:step = 10201, loss = 0.692798, precision = 0.867188 (66.378 sec) +INFO:tensorflow:global_step/sec: 1.5548 +INFO:tensorflow:step = 10301, loss = 0.677958, precision = 0.890625 (64.317 sec) +INFO:tensorflow:global_step/sec: 1.55517 +INFO:tensorflow:step = 10401, loss = 0.792972, precision = 0.820312 (64.302 sec) +INFO:tensorflow:global_step/sec: 1.55573 +INFO:tensorflow:step = 10501, loss = 0.656706, precision = 0.875 (64.279 sec) +Saved checkpoint after 27 epoch(s) to ../data/resnet164/checkpoints/00027... +INFO:tensorflow:global_step/sec: 1.50556 +INFO:tensorflow:step = 10601, loss = 0.741266, precision = 0.859375 (66.421 sec) +INFO:tensorflow:global_step/sec: 1.555 +INFO:tensorflow:step = 10701, loss = 0.692083, precision = 0.835938 (64.308 sec) +INFO:tensorflow:global_step/sec: 1.55373 +INFO:tensorflow:step = 10801, loss = 0.710944, precision = 0.851562 (64.361 sec) +INFO:tensorflow:global_step/sec: 1.55437 +INFO:tensorflow:step = 10901, loss = 0.643551, precision = 0.882812 (64.335 sec) +Saved checkpoint after 28 epoch(s) to ../data/resnet164/checkpoints/00028... +INFO:tensorflow:global_step/sec: 1.50573 +INFO:tensorflow:step = 11001, loss = 0.754302, precision = 0.859375 (66.413 sec) +INFO:tensorflow:global_step/sec: 1.55512 +INFO:tensorflow:step = 11101, loss = 0.676264, precision = 0.90625 (64.304 sec) +INFO:tensorflow:global_step/sec: 1.55471 +INFO:tensorflow:step = 11201, loss = 0.642456, precision = 0.882812 (64.321 sec) +INFO:tensorflow:global_step/sec: 1.55547 +INFO:tensorflow:step = 11301, loss = 0.634338, precision = 0.90625 (64.289 sec) +Saved checkpoint after 29 epoch(s) to ../data/resnet164/checkpoints/00029... +INFO:tensorflow:global_step/sec: 1.50241 +INFO:tensorflow:step = 11401, loss = 0.76565, precision = 0.84375 (66.560 sec) +INFO:tensorflow:global_step/sec: 1.55691 +INFO:tensorflow:step = 11501, loss = 0.649574, precision = 0.859375 (64.230 sec) +INFO:tensorflow:global_step/sec: 1.55585 +INFO:tensorflow:step = 11601, loss = 0.663617, precision = 0.859375 (64.273 sec) +INFO:tensorflow:global_step/sec: 1.55619 +INFO:tensorflow:step = 11701, loss = 0.654718, precision = 0.859375 (64.260 sec) +Saved checkpoint after 30 epoch(s) to ../data/resnet164/checkpoints/00030... +INFO:tensorflow:global_step/sec: 1.50677 +INFO:tensorflow:step = 11801, loss = 0.767095, precision = 0.828125 (66.367 sec) +INFO:tensorflow:global_step/sec: 1.55582 +INFO:tensorflow:step = 11901, loss = 0.673761, precision = 0.882812 (64.274 sec) +INFO:tensorflow:global_step/sec: 1.55552 +INFO:tensorflow:step = 12001, loss = 0.699715, precision = 0.851562 (64.287 sec) +INFO:tensorflow:global_step/sec: 1.55495 +INFO:tensorflow:step = 12101, loss = 0.816914, precision = 0.78125 (64.311 sec) +Saved checkpoint after 31 epoch(s) to ../data/resnet164/checkpoints/00031... +INFO:tensorflow:global_step/sec: 1.50688 +INFO:tensorflow:step = 12201, loss = 0.771614, precision = 0.851562 (66.363 sec) +INFO:tensorflow:global_step/sec: 1.5565 +INFO:tensorflow:step = 12301, loss = 0.608724, precision = 0.890625 (64.247 sec) +INFO:tensorflow:global_step/sec: 1.55584 +INFO:tensorflow:step = 12401, loss = 0.567717, precision = 0.90625 (64.274 sec) +INFO:tensorflow:global_step/sec: 1.5561 +INFO:tensorflow:step = 12501, loss = 0.693034, precision = 0.898438 (64.263 sec) +Saved checkpoint after 32 epoch(s) to ../data/resnet164/checkpoints/00032... +INFO:tensorflow:global_step/sec: 1.50247 +INFO:tensorflow:step = 12601, loss = 0.805572, precision = 0.84375 (66.557 sec) +INFO:tensorflow:global_step/sec: 1.55457 +INFO:tensorflow:step = 12701, loss = 0.627928, precision = 0.867188 (64.326 sec) +INFO:tensorflow:global_step/sec: 1.5559 +INFO:tensorflow:step = 12801, loss = 0.599362, precision = 0.875 (64.271 sec) +INFO:tensorflow:global_step/sec: 1.55541 +INFO:tensorflow:step = 12901, loss = 0.762054, precision = 0.804688 (64.291 sec) +Saved checkpoint after 33 epoch(s) to ../data/resnet164/checkpoints/00033... +INFO:tensorflow:global_step/sec: 1.50649 +INFO:tensorflow:step = 13001, loss = 0.596839, precision = 0.914062 (66.380 sec) +INFO:tensorflow:global_step/sec: 1.55653 +INFO:tensorflow:step = 13101, loss = 0.707306, precision = 0.84375 (64.245 sec) +INFO:tensorflow:global_step/sec: 1.55629 +INFO:tensorflow:step = 13201, loss = 0.59261, precision = 0.914062 (64.256 sec) +Saved checkpoint after 34 epoch(s) to ../data/resnet164/checkpoints/00034... +INFO:tensorflow:global_step/sec: 1.50587 +INFO:tensorflow:step = 13301, loss = 0.562554, precision = 0.90625 (66.406 sec) +INFO:tensorflow:global_step/sec: 1.55643 +INFO:tensorflow:step = 13401, loss = 0.554142, precision = 0.914062 (64.250 sec) +INFO:tensorflow:global_step/sec: 1.55639 +INFO:tensorflow:step = 13501, loss = 0.639325, precision = 0.890625 (64.251 sec) +INFO:tensorflow:global_step/sec: 1.55671 +INFO:tensorflow:step = 13601, loss = 0.688261, precision = 0.898438 (64.238 sec) +Saved checkpoint after 35 epoch(s) to ../data/resnet164/checkpoints/00035... +INFO:tensorflow:global_step/sec: 1.50792 +INFO:tensorflow:step = 13701, loss = 0.692296, precision = 0.875 (66.317 sec) +INFO:tensorflow:global_step/sec: 1.55759 +INFO:tensorflow:step = 13801, loss = 0.663815, precision = 0.890625 (64.202 sec) +INFO:tensorflow:global_step/sec: 1.55641 +INFO:tensorflow:step = 13901, loss = 0.63251, precision = 0.882812 (64.250 sec) +INFO:tensorflow:global_step/sec: 1.55606 +INFO:tensorflow:step = 14001, loss = 0.58911, precision = 0.890625 (64.265 sec) +Saved checkpoint after 36 epoch(s) to ../data/resnet164/checkpoints/00036... +INFO:tensorflow:global_step/sec: 1.50632 +INFO:tensorflow:step = 14101, loss = 0.644102, precision = 0.851562 (66.387 sec) +INFO:tensorflow:global_step/sec: 1.55641 +INFO:tensorflow:step = 14201, loss = 0.627923, precision = 0.898438 (64.251 sec) +INFO:tensorflow:global_step/sec: 1.55651 +INFO:tensorflow:step = 14301, loss = 0.666114, precision = 0.859375 (64.246 sec) +INFO:tensorflow:global_step/sec: 1.5559 +INFO:tensorflow:step = 14401, loss = 0.55923, precision = 0.929688 (64.271 sec) +Saved checkpoint after 37 epoch(s) to ../data/resnet164/checkpoints/00037... +INFO:tensorflow:global_step/sec: 1.50707 +INFO:tensorflow:step = 14501, loss = 0.634285, precision = 0.867188 (66.354 sec) +INFO:tensorflow:global_step/sec: 1.55981 +INFO:tensorflow:step = 14601, loss = 0.655628, precision = 0.90625 (64.110 sec) +INFO:tensorflow:global_step/sec: 1.557 +INFO:tensorflow:step = 14701, loss = 0.657244, precision = 0.90625 (64.226 sec) +INFO:tensorflow:global_step/sec: 1.557 +INFO:tensorflow:step = 14801, loss = 0.763827, precision = 0.851562 (64.226 sec) +Saved checkpoint after 38 epoch(s) to ../data/resnet164/checkpoints/00038... +INFO:tensorflow:global_step/sec: 1.50568 +INFO:tensorflow:step = 14901, loss = 0.614476, precision = 0.875 (66.415 sec) +INFO:tensorflow:global_step/sec: 1.55679 +INFO:tensorflow:step = 15001, loss = 0.575544, precision = 0.882812 (64.235 sec) +INFO:tensorflow:global_step/sec: 1.55714 +INFO:tensorflow:step = 15101, loss = 0.545328, precision = 0.90625 (64.220 sec) +INFO:tensorflow:global_step/sec: 1.55688 +INFO:tensorflow:step = 15201, loss = 0.544854, precision = 0.9375 (64.231 sec) +Saved checkpoint after 39 epoch(s) to ../data/resnet164/checkpoints/00039... +INFO:tensorflow:global_step/sec: 1.49789 +INFO:tensorflow:step = 15301, loss = 0.607293, precision = 0.875 (66.760 sec) +INFO:tensorflow:global_step/sec: 1.55698 +INFO:tensorflow:step = 15401, loss = 0.733651, precision = 0.867188 (64.227 sec) +INFO:tensorflow:global_step/sec: 1.55734 +INFO:tensorflow:step = 15501, loss = 0.694392, precision = 0.828125 (64.212 sec) +INFO:tensorflow:global_step/sec: 1.55766 +INFO:tensorflow:step = 15601, loss = 0.61747, precision = 0.890625 (64.199 sec) +Saved checkpoint after 40 epoch(s) to ../data/resnet164/checkpoints/00040... +INFO:tensorflow:global_step/sec: 1.50734 +INFO:tensorflow:step = 15701, loss = 0.596015, precision = 0.898438 (66.342 sec) +INFO:tensorflow:global_step/sec: 1.55727 +INFO:tensorflow:step = 15801, loss = 0.886764, precision = 0.804688 (64.215 sec) +INFO:tensorflow:global_step/sec: 1.55709 +INFO:tensorflow:step = 15901, loss = 0.693558, precision = 0.867188 (64.222 sec) +INFO:tensorflow:global_step/sec: 1.55772 +INFO:tensorflow:step = 16001, loss = 0.878001, precision = 0.804688 (64.197 sec) +Saved checkpoint after 41 epoch(s) to ../data/resnet164/checkpoints/00041... +INFO:tensorflow:global_step/sec: 1.50674 +INFO:tensorflow:step = 16101, loss = 0.755829, precision = 0.820312 (66.369 sec) +INFO:tensorflow:global_step/sec: 1.55602 +INFO:tensorflow:step = 16201, loss = 0.565037, precision = 0.898438 (64.267 sec) +INFO:tensorflow:global_step/sec: 1.55594 +INFO:tensorflow:step = 16301, loss = 0.592363, precision = 0.898438 (64.270 sec) +INFO:tensorflow:global_step/sec: 1.55631 +INFO:tensorflow:step = 16401, loss = 0.550342, precision = 0.914062 (64.255 sec) +Saved checkpoint after 42 epoch(s) to ../data/resnet164/checkpoints/00042... +INFO:tensorflow:global_step/sec: 1.50723 +INFO:tensorflow:step = 16501, loss = 0.651547, precision = 0.875 (66.347 sec) +INFO:tensorflow:global_step/sec: 1.55612 +INFO:tensorflow:step = 16601, loss = 0.611668, precision = 0.898438 (64.262 sec) +INFO:tensorflow:global_step/sec: 1.55592 +INFO:tensorflow:step = 16701, loss = 0.621329, precision = 0.898438 (64.271 sec) +INFO:tensorflow:global_step/sec: 1.557 +INFO:tensorflow:step = 16801, loss = 0.744931, precision = 0.835938 (64.226 sec) +Saved checkpoint after 43 epoch(s) to ../data/resnet164/checkpoints/00043... +INFO:tensorflow:global_step/sec: 1.50772 +INFO:tensorflow:step = 16901, loss = 0.630523, precision = 0.882812 (66.326 sec) +INFO:tensorflow:global_step/sec: 1.55559 +INFO:tensorflow:step = 17001, loss = 0.54846, precision = 0.921875 (64.284 sec) +INFO:tensorflow:global_step/sec: 1.55566 +INFO:tensorflow:step = 17101, loss = 0.731216, precision = 0.851562 (64.281 sec) +INFO:tensorflow:global_step/sec: 1.5561 +INFO:tensorflow:step = 17201, loss = 0.710436, precision = 0.851562 (64.263 sec) +Saved checkpoint after 44 epoch(s) to ../data/resnet164/checkpoints/00044... +INFO:tensorflow:global_step/sec: 1.50546 +INFO:tensorflow:step = 17301, loss = 0.575276, precision = 0.890625 (66.425 sec) +INFO:tensorflow:global_step/sec: 1.55699 +INFO:tensorflow:step = 17401, loss = 0.716483, precision = 0.84375 (64.226 sec) +INFO:tensorflow:global_step/sec: 1.55598 +INFO:tensorflow:step = 17501, loss = 0.795177, precision = 0.84375 (64.268 sec) +Saved checkpoint after 45 epoch(s) to ../data/resnet164/checkpoints/00045... +INFO:tensorflow:global_step/sec: 1.50739 +INFO:tensorflow:step = 17601, loss = 0.830768, precision = 0.84375 (66.340 sec) +INFO:tensorflow:global_step/sec: 1.55503 +INFO:tensorflow:step = 17701, loss = 0.70319, precision = 0.882812 (64.307 sec) +INFO:tensorflow:global_step/sec: 1.55524 +INFO:tensorflow:step = 17801, loss = 0.740993, precision = 0.859375 (64.299 sec) +INFO:tensorflow:global_step/sec: 1.55532 +INFO:tensorflow:step = 17901, loss = 0.637983, precision = 0.882812 (64.296 sec) +Saved checkpoint after 46 epoch(s) to ../data/resnet164/checkpoints/00046... +INFO:tensorflow:global_step/sec: 1.50665 +INFO:tensorflow:step = 18001, loss = 0.642179, precision = 0.867188 (66.373 sec) +INFO:tensorflow:global_step/sec: 1.5553 +INFO:tensorflow:step = 18101, loss = 0.595496, precision = 0.890625 (64.296 sec) +INFO:tensorflow:global_step/sec: 1.55534 +INFO:tensorflow:step = 18201, loss = 0.701327, precision = 0.859375 (64.295 sec) +INFO:tensorflow:global_step/sec: 1.55644 +INFO:tensorflow:step = 18301, loss = 0.608165, precision = 0.898438 (64.249 sec) +Saved checkpoint after 47 epoch(s) to ../data/resnet164/checkpoints/00047... +INFO:tensorflow:global_step/sec: 1.50641 +INFO:tensorflow:step = 18401, loss = 0.509867, precision = 0.953125 (66.383 sec) +INFO:tensorflow:global_step/sec: 1.55561 +INFO:tensorflow:step = 18501, loss = 0.528672, precision = 0.90625 (64.283 sec) +INFO:tensorflow:global_step/sec: 1.55488 +INFO:tensorflow:step = 18601, loss = 0.565785, precision = 0.921875 (64.314 sec) +INFO:tensorflow:global_step/sec: 1.55585 +INFO:tensorflow:step = 18701, loss = 0.903027, precision = 0.828125 (64.273 sec) +Saved checkpoint after 48 epoch(s) to ../data/resnet164/checkpoints/00048... +INFO:tensorflow:global_step/sec: 1.50675 +INFO:tensorflow:step = 18801, loss = 0.624131, precision = 0.90625 (66.368 sec) +INFO:tensorflow:global_step/sec: 1.5546 +INFO:tensorflow:step = 18901, loss = 0.623021, precision = 0.898438 (64.325 sec) +INFO:tensorflow:global_step/sec: 1.55469 +INFO:tensorflow:step = 19001, loss = 0.52305, precision = 0.921875 (64.322 sec) +INFO:tensorflow:global_step/sec: 1.55428 +INFO:tensorflow:step = 19101, loss = 0.638209, precision = 0.898438 (64.338 sec) +Saved checkpoint after 49 epoch(s) to ../data/resnet164/checkpoints/00049... +INFO:tensorflow:global_step/sec: 1.50163 +INFO:tensorflow:step = 19201, loss = 0.690031, precision = 0.851562 (66.594 sec) +INFO:tensorflow:global_step/sec: 1.55509 +INFO:tensorflow:step = 19301, loss = 0.659574, precision = 0.890625 (64.305 sec) +INFO:tensorflow:global_step/sec: 1.55416 +INFO:tensorflow:step = 19401, loss = 0.643722, precision = 0.898438 (64.344 sec) +INFO:tensorflow:global_step/sec: 1.55383 +INFO:tensorflow:step = 19501, loss = 0.613904, precision = 0.921875 (64.357 sec) +Saved checkpoint after 50 epoch(s) to ../data/resnet164/checkpoints/00050... +INFO:tensorflow:global_step/sec: 1.50127 +INFO:tensorflow:step = 19601, loss = 0.70531, precision = 0.867188 (66.611 sec) +INFO:tensorflow:global_step/sec: 1.55621 +INFO:tensorflow:step = 19701, loss = 0.672662, precision = 0.867188 (64.259 sec) +INFO:tensorflow:global_step/sec: 1.55502 +INFO:tensorflow:step = 19801, loss = 0.602554, precision = 0.898438 (64.308 sec) +INFO:tensorflow:global_step/sec: 1.55541 +INFO:tensorflow:step = 19901, loss = 0.71891, precision = 0.851562 (64.292 sec) +Saved checkpoint after 51 epoch(s) to ../data/resnet164/checkpoints/00051... +INFO:tensorflow:global_step/sec: 1.50643 +INFO:tensorflow:step = 20001, loss = 0.715232, precision = 0.859375 (66.382 sec) +INFO:tensorflow:global_step/sec: 1.55599 +INFO:tensorflow:step = 20101, loss = 0.803541, precision = 0.835938 (64.268 sec) +INFO:tensorflow:global_step/sec: 1.5561 +INFO:tensorflow:step = 20201, loss = 0.728295, precision = 0.867188 (64.263 sec) +INFO:tensorflow:global_step/sec: 1.55468 +INFO:tensorflow:step = 20301, loss = 0.727328, precision = 0.84375 (64.322 sec) +Saved checkpoint after 52 epoch(s) to ../data/resnet164/checkpoints/00052... +INFO:tensorflow:global_step/sec: 1.50615 +INFO:tensorflow:step = 20401, loss = 0.698256, precision = 0.828125 (66.394 sec) +INFO:tensorflow:global_step/sec: 1.55519 +INFO:tensorflow:step = 20501, loss = 0.738047, precision = 0.828125 (64.301 sec) +INFO:tensorflow:global_step/sec: 1.55623 +INFO:tensorflow:step = 20601, loss = 0.563412, precision = 0.898438 (64.258 sec) +INFO:tensorflow:global_step/sec: 1.55673 +INFO:tensorflow:step = 20701, loss = 0.58688, precision = 0.921875 (64.237 sec) +Saved checkpoint after 53 epoch(s) to ../data/resnet164/checkpoints/00053... +INFO:tensorflow:global_step/sec: 1.50855 +INFO:tensorflow:step = 20801, loss = 0.74191, precision = 0.867188 (66.289 sec) +INFO:tensorflow:global_step/sec: 1.55589 +INFO:tensorflow:step = 20901, loss = 0.550981, precision = 0.921875 (64.272 sec) +INFO:tensorflow:global_step/sec: 1.55573 +INFO:tensorflow:step = 21001, loss = 0.530965, precision = 0.921875 (64.279 sec) +INFO:tensorflow:global_step/sec: 1.55602 +INFO:tensorflow:step = 21101, loss = 0.580255, precision = 0.9375 (64.267 sec) +Saved checkpoint after 54 epoch(s) to ../data/resnet164/checkpoints/00054... +INFO:tensorflow:global_step/sec: 1.50826 +INFO:tensorflow:step = 21201, loss = 0.578652, precision = 0.898438 (66.301 sec) +INFO:tensorflow:global_step/sec: 1.55599 +INFO:tensorflow:step = 21301, loss = 0.71297, precision = 0.875 (64.268 sec) +INFO:tensorflow:global_step/sec: 1.55687 +INFO:tensorflow:step = 21401, loss = 0.718809, precision = 0.835938 (64.232 sec) +INFO:tensorflow:global_step/sec: 1.55587 +INFO:tensorflow:step = 21501, loss = 0.588553, precision = 0.90625 (64.273 sec) +Saved checkpoint after 55 epoch(s) to ../data/resnet164/checkpoints/00055... +INFO:tensorflow:global_step/sec: 1.50676 +INFO:tensorflow:step = 21601, loss = 0.65827, precision = 0.890625 (66.368 sec) +INFO:tensorflow:global_step/sec: 1.55548 +INFO:tensorflow:step = 21701, loss = 0.769404, precision = 0.835938 (64.289 sec) +INFO:tensorflow:global_step/sec: 1.55684 +INFO:tensorflow:step = 21801, loss = 0.601255, precision = 0.90625 (64.233 sec) +Saved checkpoint after 56 epoch(s) to ../data/resnet164/checkpoints/00056... +INFO:tensorflow:global_step/sec: 1.5079 +INFO:tensorflow:step = 21901, loss = 0.572716, precision = 0.898438 (66.317 sec) +INFO:tensorflow:global_step/sec: 1.55687 +INFO:tensorflow:step = 22001, loss = 0.725251, precision = 0.820312 (64.231 sec) +INFO:tensorflow:global_step/sec: 1.55645 +INFO:tensorflow:step = 22101, loss = 0.56911, precision = 0.921875 (64.249 sec) +INFO:tensorflow:global_step/sec: 1.55573 +INFO:tensorflow:step = 22201, loss = 0.650998, precision = 0.914062 (64.279 sec) +Saved checkpoint after 57 epoch(s) to ../data/resnet164/checkpoints/00057... +INFO:tensorflow:global_step/sec: 1.50631 +INFO:tensorflow:step = 22301, loss = 0.571344, precision = 0.914062 (66.387 sec) +INFO:tensorflow:global_step/sec: 1.55635 +INFO:tensorflow:step = 22401, loss = 0.661008, precision = 0.882812 (64.253 sec) +INFO:tensorflow:global_step/sec: 1.55632 +INFO:tensorflow:step = 22501, loss = 0.662948, precision = 0.890625 (64.254 sec) +INFO:tensorflow:global_step/sec: 1.55601 +INFO:tensorflow:step = 22601, loss = 0.664441, precision = 0.90625 (64.267 sec) +Saved checkpoint after 58 epoch(s) to ../data/resnet164/checkpoints/00058... +INFO:tensorflow:global_step/sec: 1.50708 +INFO:tensorflow:step = 22701, loss = 0.746279, precision = 0.835938 (66.353 sec) +INFO:tensorflow:global_step/sec: 1.5556 +INFO:tensorflow:step = 22801, loss = 0.672426, precision = 0.882812 (64.284 sec) +INFO:tensorflow:global_step/sec: 1.55595 +INFO:tensorflow:step = 22901, loss = 0.586522, precision = 0.914062 (64.269 sec) +INFO:tensorflow:global_step/sec: 1.55698 +INFO:tensorflow:step = 23001, loss = 0.563922, precision = 0.914062 (64.227 sec) +Saved checkpoint after 59 epoch(s) to ../data/resnet164/checkpoints/00059... +INFO:tensorflow:global_step/sec: 1.50392 +INFO:tensorflow:step = 23101, loss = 0.627094, precision = 0.90625 (66.493 sec) +INFO:tensorflow:global_step/sec: 1.55513 +INFO:tensorflow:step = 23201, loss = 0.586493, precision = 0.898438 (64.304 sec) +INFO:tensorflow:global_step/sec: 1.55663 +INFO:tensorflow:step = 23301, loss = 0.668713, precision = 0.859375 (64.241 sec) +INFO:tensorflow:global_step/sec: 1.55675 +INFO:tensorflow:step = 23401, loss = 0.592107, precision = 0.90625 (64.236 sec) +Saved checkpoint after 60 epoch(s) to ../data/resnet164/checkpoints/00060... +INFO:tensorflow:global_step/sec: 1.50845 +INFO:tensorflow:step = 23501, loss = 0.640908, precision = 0.90625 (66.294 sec) +INFO:tensorflow:global_step/sec: 1.55505 +INFO:tensorflow:step = 23601, loss = 0.846159, precision = 0.820312 (64.306 sec) +INFO:tensorflow:global_step/sec: 1.55622 +INFO:tensorflow:step = 23701, loss = 0.743729, precision = 0.875 (64.258 sec) +INFO:tensorflow:global_step/sec: 1.55511 +INFO:tensorflow:step = 23801, loss = 0.594708, precision = 0.882812 (64.304 sec) +Saved checkpoint after 61 epoch(s) to ../data/resnet164/checkpoints/00061... +INFO:tensorflow:global_step/sec: 1.50836 +INFO:tensorflow:step = 23901, loss = 0.603558, precision = 0.90625 (66.297 sec) +INFO:tensorflow:global_step/sec: 1.55508 +INFO:tensorflow:step = 24001, loss = 0.676485, precision = 0.898438 (64.306 sec) +INFO:tensorflow:global_step/sec: 1.55621 +INFO:tensorflow:step = 24101, loss = 0.529737, precision = 0.929688 (64.258 sec) +INFO:tensorflow:global_step/sec: 1.55635 +INFO:tensorflow:step = 24201, loss = 0.565949, precision = 0.890625 (64.253 sec) +Saved checkpoint after 62 epoch(s) to ../data/resnet164/checkpoints/00062... +INFO:tensorflow:global_step/sec: 1.5095 +INFO:tensorflow:step = 24301, loss = 0.619603, precision = 0.867188 (66.247 sec) +INFO:tensorflow:global_step/sec: 1.55656 +INFO:tensorflow:step = 24401, loss = 0.657148, precision = 0.882812 (64.244 sec) +INFO:tensorflow:global_step/sec: 1.55545 +INFO:tensorflow:step = 24501, loss = 0.635658, precision = 0.898438 (64.290 sec) +INFO:tensorflow:global_step/sec: 1.55651 +INFO:tensorflow:step = 24601, loss = 0.570276, precision = 0.90625 (64.246 sec) +Saved checkpoint after 63 epoch(s) to ../data/resnet164/checkpoints/00063... +INFO:tensorflow:global_step/sec: 1.50852 +INFO:tensorflow:step = 24701, loss = 0.63613, precision = 0.890625 (66.291 sec) +INFO:tensorflow:global_step/sec: 1.55714 +INFO:tensorflow:step = 24801, loss = 0.772151, precision = 0.84375 (64.220 sec) +INFO:tensorflow:global_step/sec: 1.55656 +INFO:tensorflow:step = 24901, loss = 0.606431, precision = 0.914062 (64.244 sec) +INFO:tensorflow:global_step/sec: 1.55622 +INFO:tensorflow:step = 25001, loss = 0.539717, precision = 0.90625 (64.258 sec) +Saved checkpoint after 64 epoch(s) to ../data/resnet164/checkpoints/00064... +INFO:tensorflow:global_step/sec: 1.50968 +INFO:tensorflow:step = 25101, loss = 0.769457, precision = 0.835938 (66.239 sec) +INFO:tensorflow:global_step/sec: 1.55856 +INFO:tensorflow:step = 25201, loss = 0.704265, precision = 0.859375 (64.162 sec) +INFO:tensorflow:global_step/sec: 1.55735 +INFO:tensorflow:step = 25301, loss = 0.603789, precision = 0.882812 (64.212 sec) +INFO:tensorflow:global_step/sec: 1.55589 +INFO:tensorflow:step = 25401, loss = 0.544825, precision = 0.914062 (64.272 sec) +Saved checkpoint after 65 epoch(s) to ../data/resnet164/checkpoints/00065... +INFO:tensorflow:global_step/sec: 1.50862 +INFO:tensorflow:step = 25501, loss = 0.659881, precision = 0.898438 (66.286 sec) +INFO:tensorflow:global_step/sec: 1.55628 +INFO:tensorflow:step = 25601, loss = 0.676015, precision = 0.875 (64.256 sec) +INFO:tensorflow:global_step/sec: 1.55675 +INFO:tensorflow:step = 25701, loss = 0.70976, precision = 0.851562 (64.236 sec) +INFO:tensorflow:global_step/sec: 1.55564 +INFO:tensorflow:step = 25801, loss = 0.690276, precision = 0.882812 (64.282 sec) +Saved checkpoint after 66 epoch(s) to ../data/resnet164/checkpoints/00066... +INFO:tensorflow:global_step/sec: 1.50963 +INFO:tensorflow:step = 25901, loss = 0.637583, precision = 0.898438 (66.242 sec) +INFO:tensorflow:global_step/sec: 1.55515 +INFO:tensorflow:step = 26001, loss = 0.553813, precision = 0.929688 (64.303 sec) +INFO:tensorflow:global_step/sec: 1.55588 +INFO:tensorflow:step = 26101, loss = 0.60947, precision = 0.90625 (64.272 sec) +Saved checkpoint after 67 epoch(s) to ../data/resnet164/checkpoints/00067... +INFO:tensorflow:global_step/sec: 1.50834 +INFO:tensorflow:step = 26201, loss = 0.688387, precision = 0.875 (66.298 sec) +INFO:tensorflow:global_step/sec: 1.55726 +INFO:tensorflow:step = 26301, loss = 0.593381, precision = 0.890625 (64.215 sec) +INFO:tensorflow:global_step/sec: 1.55815 +INFO:tensorflow:step = 26401, loss = 0.612315, precision = 0.90625 (64.179 sec) +INFO:tensorflow:global_step/sec: 1.55654 +INFO:tensorflow:step = 26501, loss = 0.522839, precision = 0.960938 (64.245 sec) +Saved checkpoint after 68 epoch(s) to ../data/resnet164/checkpoints/00068... +INFO:tensorflow:global_step/sec: 1.50985 +INFO:tensorflow:step = 26601, loss = 0.49599, precision = 0.953125 (66.232 sec) +INFO:tensorflow:global_step/sec: 1.55777 +INFO:tensorflow:step = 26701, loss = 0.558221, precision = 0.945312 (64.194 sec) +INFO:tensorflow:global_step/sec: 1.55687 +INFO:tensorflow:step = 26801, loss = 0.658984, precision = 0.867188 (64.232 sec) +INFO:tensorflow:global_step/sec: 1.55618 +INFO:tensorflow:step = 26901, loss = 0.470458, precision = 0.945312 (64.260 sec) +Saved checkpoint after 69 epoch(s) to ../data/resnet164/checkpoints/00069... +INFO:tensorflow:global_step/sec: 1.5041 +INFO:tensorflow:step = 27001, loss = 0.625759, precision = 0.890625 (66.485 sec) +INFO:tensorflow:global_step/sec: 1.55837 +INFO:tensorflow:step = 27101, loss = 0.596275, precision = 0.914062 (64.169 sec) +INFO:tensorflow:global_step/sec: 1.55701 +INFO:tensorflow:step = 27201, loss = 0.589078, precision = 0.890625 (64.226 sec) +INFO:tensorflow:global_step/sec: 1.55727 +INFO:tensorflow:step = 27301, loss = 0.686679, precision = 0.875 (64.215 sec) +Saved checkpoint after 70 epoch(s) to ../data/resnet164/checkpoints/00070... +INFO:tensorflow:global_step/sec: 1.50846 +INFO:tensorflow:step = 27401, loss = 0.55169, precision = 0.914062 (66.293 sec) +INFO:tensorflow:global_step/sec: 1.55786 +INFO:tensorflow:step = 27501, loss = 0.543307, precision = 0.929688 (64.190 sec) +INFO:tensorflow:global_step/sec: 1.55754 +INFO:tensorflow:step = 27601, loss = 0.746455, precision = 0.835938 (64.204 sec) +INFO:tensorflow:global_step/sec: 1.5562 +INFO:tensorflow:step = 27701, loss = 0.603678, precision = 0.9375 (64.259 sec) +Saved checkpoint after 71 epoch(s) to ../data/resnet164/checkpoints/00071... +INFO:tensorflow:global_step/sec: 1.50865 +INFO:tensorflow:step = 27801, loss = 0.660553, precision = 0.898438 (66.285 sec) +INFO:tensorflow:global_step/sec: 1.55675 +INFO:tensorflow:step = 27901, loss = 0.572108, precision = 0.929688 (64.236 sec) +INFO:tensorflow:global_step/sec: 1.55619 +INFO:tensorflow:step = 28001, loss = 0.617018, precision = 0.882812 (64.260 sec) +INFO:tensorflow:global_step/sec: 1.55781 +INFO:tensorflow:step = 28101, loss = 0.622763, precision = 0.929688 (64.193 sec) +Saved checkpoint after 72 epoch(s) to ../data/resnet164/checkpoints/00072... +INFO:tensorflow:global_step/sec: 1.50494 +INFO:tensorflow:step = 28201, loss = 0.627321, precision = 0.898438 (66.448 sec) +INFO:tensorflow:global_step/sec: 1.5569 +INFO:tensorflow:step = 28301, loss = 0.724568, precision = 0.90625 (64.230 sec) +INFO:tensorflow:global_step/sec: 1.55609 +INFO:tensorflow:step = 28401, loss = 0.634337, precision = 0.898438 (64.264 sec) +INFO:tensorflow:global_step/sec: 1.55727 +INFO:tensorflow:step = 28501, loss = 0.686451, precision = 0.867188 (64.215 sec) +Saved checkpoint after 73 epoch(s) to ../data/resnet164/checkpoints/00073... +INFO:tensorflow:global_step/sec: 1.50915 +INFO:tensorflow:step = 28601, loss = 0.686224, precision = 0.875 (66.263 sec) +INFO:tensorflow:global_step/sec: 1.55641 +INFO:tensorflow:step = 28701, loss = 0.624439, precision = 0.890625 (64.250 sec) +INFO:tensorflow:global_step/sec: 1.55574 +INFO:tensorflow:step = 28801, loss = 0.58289, precision = 0.90625 (64.278 sec) +INFO:tensorflow:global_step/sec: 1.55633 +INFO:tensorflow:step = 28901, loss = 0.635351, precision = 0.882812 (64.254 sec) +Saved checkpoint after 74 epoch(s) to ../data/resnet164/checkpoints/00074... +INFO:tensorflow:global_step/sec: 1.50897 +INFO:tensorflow:step = 29001, loss = 0.579458, precision = 0.914062 (66.270 sec) +INFO:tensorflow:global_step/sec: 1.55577 +INFO:tensorflow:step = 29101, loss = 0.539899, precision = 0.898438 (64.277 sec) +INFO:tensorflow:global_step/sec: 1.55508 +INFO:tensorflow:step = 29201, loss = 0.583151, precision = 0.890625 (64.305 sec) +INFO:tensorflow:global_step/sec: 1.55554 +INFO:tensorflow:step = 29301, loss = 0.568712, precision = 0.90625 (64.286 sec) +Saved checkpoint after 75 epoch(s) to ../data/resnet164/checkpoints/00075... +INFO:tensorflow:global_step/sec: 1.5094 +INFO:tensorflow:step = 29401, loss = 0.652383, precision = 0.890625 (66.251 sec) +INFO:tensorflow:global_step/sec: 1.55555 +INFO:tensorflow:step = 29501, loss = 0.721786, precision = 0.859375 (64.286 sec) +INFO:tensorflow:global_step/sec: 1.55581 +INFO:tensorflow:step = 29601, loss = 0.661622, precision = 0.882812 (64.276 sec) +INFO:tensorflow:global_step/sec: 1.556 +INFO:tensorflow:step = 29701, loss = 0.641508, precision = 0.890625 (64.267 sec) +Saved checkpoint after 76 epoch(s) to ../data/resnet164/checkpoints/00076... +INFO:tensorflow:global_step/sec: 1.50882 +INFO:tensorflow:step = 29801, loss = 0.526703, precision = 0.945312 (66.277 sec) +INFO:tensorflow:global_step/sec: 1.55761 +INFO:tensorflow:step = 29901, loss = 0.642686, precision = 0.921875 (64.201 sec) +INFO:tensorflow:global_step/sec: 1.5555 +INFO:tensorflow:step = 30001, loss = 0.680753, precision = 0.851562 (64.288 sec) +INFO:tensorflow:global_step/sec: 1.555 +INFO:tensorflow:step = 30101, loss = 0.740496, precision = 0.867188 (64.309 sec) +Saved checkpoint after 77 epoch(s) to ../data/resnet164/checkpoints/00077... +INFO:tensorflow:global_step/sec: 1.50941 +INFO:tensorflow:step = 30201, loss = 0.813719, precision = 0.8125 (66.251 sec) +INFO:tensorflow:global_step/sec: 1.5553 +INFO:tensorflow:step = 30301, loss = 0.565451, precision = 0.898438 (64.297 sec) +INFO:tensorflow:global_step/sec: 1.55654 +INFO:tensorflow:step = 30401, loss = 0.809616, precision = 0.8125 (64.245 sec) +Saved checkpoint after 78 epoch(s) to ../data/resnet164/checkpoints/00078... +INFO:tensorflow:global_step/sec: 1.50808 +INFO:tensorflow:step = 30501, loss = 0.663812, precision = 0.90625 (66.309 sec) +INFO:tensorflow:global_step/sec: 1.5556 +INFO:tensorflow:step = 30601, loss = 0.655035, precision = 0.882812 (64.284 sec) +INFO:tensorflow:global_step/sec: 1.55551 +INFO:tensorflow:step = 30701, loss = 0.793376, precision = 0.835938 (64.287 sec) +INFO:tensorflow:global_step/sec: 1.55642 +INFO:tensorflow:step = 30801, loss = 0.766505, precision = 0.859375 (64.250 sec) +Saved checkpoint after 79 epoch(s) to ../data/resnet164/checkpoints/00079... +INFO:tensorflow:global_step/sec: 1.50406 +INFO:tensorflow:step = 30901, loss = 0.598427, precision = 0.875 (66.486 sec) +INFO:tensorflow:global_step/sec: 1.55439 +INFO:tensorflow:step = 31001, loss = 0.570438, precision = 0.914062 (64.334 sec) +INFO:tensorflow:global_step/sec: 1.55604 +INFO:tensorflow:step = 31101, loss = 0.73061, precision = 0.867188 (64.265 sec) +INFO:tensorflow:global_step/sec: 1.55537 +INFO:tensorflow:step = 31201, loss = 0.694766, precision = 0.875 (64.293 sec) +Saved checkpoint after 80 epoch(s) to ../data/resnet164/checkpoints/00080... +INFO:tensorflow:global_step/sec: 1.50996 +INFO:tensorflow:step = 31301, loss = 0.724486, precision = 0.859375 (66.227 sec) +INFO:tensorflow:global_step/sec: 1.55538 +INFO:tensorflow:step = 31401, loss = 0.656408, precision = 0.898438 (64.293 sec) +INFO:tensorflow:global_step/sec: 1.55464 +INFO:tensorflow:step = 31501, loss = 0.646483, precision = 0.875 (64.323 sec) +INFO:tensorflow:global_step/sec: 1.5552 +INFO:tensorflow:step = 31601, loss = 0.719265, precision = 0.867188 (64.300 sec) +Saved checkpoint after 81 epoch(s) to ../data/resnet164/checkpoints/00081... +INFO:tensorflow:global_step/sec: 1.50827 +INFO:tensorflow:step = 31701, loss = 0.606796, precision = 0.898438 (66.301 sec) +INFO:tensorflow:global_step/sec: 1.55617 +INFO:tensorflow:step = 31801, loss = 0.681703, precision = 0.875 (64.260 sec) +INFO:tensorflow:global_step/sec: 1.55474 +INFO:tensorflow:step = 31901, loss = 0.686805, precision = 0.890625 (64.319 sec) +INFO:tensorflow:global_step/sec: 1.55668 +INFO:tensorflow:step = 32001, loss = 0.644127, precision = 0.867188 (64.240 sec) +Saved checkpoint after 82 epoch(s) to ../data/resnet164/checkpoints/00082... +INFO:tensorflow:global_step/sec: 1.50875 +INFO:tensorflow:step = 32101, loss = 0.542369, precision = 0.914062 (66.280 sec) +INFO:tensorflow:global_step/sec: 1.55678 +INFO:tensorflow:step = 32201, loss = 0.678288, precision = 0.882812 (64.235 sec) +INFO:tensorflow:global_step/sec: 1.55584 +INFO:tensorflow:step = 32301, loss = 0.680477, precision = 0.875 (64.274 sec) +INFO:tensorflow:global_step/sec: 1.55529 +INFO:tensorflow:step = 32401, loss = 0.609346, precision = 0.898438 (64.297 sec) +Saved checkpoint after 83 epoch(s) to ../data/resnet164/checkpoints/00083... +INFO:tensorflow:global_step/sec: 1.50748 +INFO:tensorflow:step = 32501, loss = 0.757153, precision = 0.882812 (66.336 sec) +INFO:tensorflow:global_step/sec: 1.55533 +INFO:tensorflow:step = 32601, loss = 0.656563, precision = 0.898438 (64.295 sec) +INFO:tensorflow:global_step/sec: 1.55691 +INFO:tensorflow:step = 32701, loss = 0.662174, precision = 0.882812 (64.230 sec) +INFO:tensorflow:global_step/sec: 1.5556 +INFO:tensorflow:step = 32801, loss = 0.577156, precision = 0.898438 (64.284 sec) +Saved checkpoint after 84 epoch(s) to ../data/resnet164/checkpoints/00084... +INFO:tensorflow:global_step/sec: 1.50774 +INFO:tensorflow:step = 32901, loss = 0.647345, precision = 0.890625 (66.324 sec) +INFO:tensorflow:global_step/sec: 1.55601 +INFO:tensorflow:step = 33001, loss = 0.553805, precision = 0.921875 (64.267 sec) +INFO:tensorflow:global_step/sec: 1.55834 +INFO:tensorflow:step = 33101, loss = 0.864977, precision = 0.859375 (64.171 sec) +INFO:tensorflow:global_step/sec: 1.55712 +INFO:tensorflow:step = 33201, loss = 0.716414, precision = 0.867188 (64.221 sec) +Saved checkpoint after 85 epoch(s) to ../data/resnet164/checkpoints/00085... +INFO:tensorflow:global_step/sec: 1.51056 +INFO:tensorflow:step = 33301, loss = 0.584244, precision = 0.914062 (66.201 sec) +INFO:tensorflow:global_step/sec: 1.55801 +INFO:tensorflow:step = 33401, loss = 0.654099, precision = 0.90625 (64.184 sec) +INFO:tensorflow:global_step/sec: 1.55718 +INFO:tensorflow:step = 33501, loss = 0.668259, precision = 0.929688 (64.219 sec) +INFO:tensorflow:global_step/sec: 1.55758 +INFO:tensorflow:step = 33601, loss = 0.685741, precision = 0.882812 (64.202 sec) +Saved checkpoint after 86 epoch(s) to ../data/resnet164/checkpoints/00086... +INFO:tensorflow:global_step/sec: 1.50875 +INFO:tensorflow:step = 33701, loss = 0.543028, precision = 0.914062 (66.280 sec) +INFO:tensorflow:global_step/sec: 1.55512 +INFO:tensorflow:step = 33801, loss = 0.65903, precision = 0.859375 (64.304 sec) +INFO:tensorflow:global_step/sec: 1.55613 +INFO:tensorflow:step = 33901, loss = 0.542076, precision = 0.90625 (64.262 sec) +INFO:tensorflow:global_step/sec: 1.55625 +INFO:tensorflow:step = 34001, loss = 0.506551, precision = 0.945312 (64.257 sec) +Saved checkpoint after 87 epoch(s) to ../data/resnet164/checkpoints/00087... +INFO:tensorflow:global_step/sec: 1.50869 +INFO:tensorflow:step = 34101, loss = 0.54473, precision = 0.921875 (66.283 sec) +INFO:tensorflow:global_step/sec: 1.5559 +INFO:tensorflow:step = 34201, loss = 0.61333, precision = 0.9375 (64.271 sec) +INFO:tensorflow:global_step/sec: 1.55611 +INFO:tensorflow:step = 34301, loss = 0.639026, precision = 0.890625 (64.263 sec) +INFO:tensorflow:global_step/sec: 1.55632 +INFO:tensorflow:step = 34401, loss = 0.540375, precision = 0.945312 (64.254 sec) +Saved checkpoint after 88 epoch(s) to ../data/resnet164/checkpoints/00088... +INFO:tensorflow:global_step/sec: 1.51061 +INFO:tensorflow:step = 34501, loss = 0.698986, precision = 0.875 (66.198 sec) +INFO:tensorflow:global_step/sec: 1.55642 +INFO:tensorflow:step = 34601, loss = 0.681435, precision = 0.835938 (64.250 sec) +INFO:tensorflow:global_step/sec: 1.55604 +INFO:tensorflow:step = 34701, loss = 0.670666, precision = 0.875 (64.266 sec) +Saved checkpoint after 89 epoch(s) to ../data/resnet164/checkpoints/00089... +INFO:tensorflow:global_step/sec: 1.50503 +INFO:tensorflow:step = 34801, loss = 0.600994, precision = 0.929688 (66.444 sec) +INFO:tensorflow:global_step/sec: 1.55716 +INFO:tensorflow:step = 34901, loss = 0.69292, precision = 0.882812 (64.220 sec) +INFO:tensorflow:global_step/sec: 1.55782 +INFO:tensorflow:step = 35001, loss = 0.521913, precision = 0.9375 (64.192 sec) +INFO:tensorflow:global_step/sec: 1.55839 +INFO:tensorflow:step = 35101, loss = 0.580797, precision = 0.898438 (64.169 sec) +Saved checkpoint after 90 epoch(s) to ../data/resnet164/checkpoints/00090... +INFO:tensorflow:global_step/sec: 1.50998 +INFO:tensorflow:step = 35201, loss = 0.567569, precision = 0.90625 (66.226 sec) +INFO:tensorflow:global_step/sec: 1.5574 +INFO:tensorflow:step = 35301, loss = 0.630573, precision = 0.898438 (64.209 sec) +INFO:tensorflow:global_step/sec: 1.55841 +INFO:tensorflow:step = 35401, loss = 0.627736, precision = 0.898438 (64.168 sec) +INFO:tensorflow:global_step/sec: 1.55854 +INFO:tensorflow:step = 35501, loss = 0.649734, precision = 0.890625 (64.163 sec) +Saved checkpoint after 91 epoch(s) to ../data/resnet164/checkpoints/00091... +INFO:tensorflow:global_step/sec: 1.50967 +INFO:tensorflow:step = 35601, loss = 0.496075, precision = 0.960938 (66.240 sec) +INFO:tensorflow:global_step/sec: 1.55666 +INFO:tensorflow:step = 35701, loss = 0.588931, precision = 0.90625 (64.240 sec) +INFO:tensorflow:global_step/sec: 1.55818 +INFO:tensorflow:step = 35801, loss = 0.430073, precision = 0.984375 (64.177 sec) +INFO:tensorflow:global_step/sec: 1.55775 +INFO:tensorflow:step = 35901, loss = 0.44327, precision = 0.945312 (64.195 sec) +Saved checkpoint after 92 epoch(s) to ../data/resnet164/checkpoints/00092... +INFO:tensorflow:global_step/sec: 1.50933 +INFO:tensorflow:step = 36001, loss = 0.430657, precision = 0.945312 (66.255 sec) +INFO:tensorflow:global_step/sec: 1.55725 +INFO:tensorflow:step = 36101, loss = 0.394901, precision = 0.96875 (64.215 sec) +INFO:tensorflow:global_step/sec: 1.5564 +INFO:tensorflow:step = 36201, loss = 0.541164, precision = 0.929688 (64.251 sec) +INFO:tensorflow:global_step/sec: 1.55604 +INFO:tensorflow:step = 36301, loss = 0.601741, precision = 0.914062 (64.266 sec) +Saved checkpoint after 93 epoch(s) to ../data/resnet164/checkpoints/00093... +INFO:tensorflow:global_step/sec: 1.51106 +INFO:tensorflow:step = 36401, loss = 0.369239, precision = 0.984375 (66.179 sec) +INFO:tensorflow:global_step/sec: 1.5562 +INFO:tensorflow:step = 36501, loss = 0.361595, precision = 0.984375 (64.259 sec) +INFO:tensorflow:global_step/sec: 1.55581 +INFO:tensorflow:step = 36601, loss = 0.485816, precision = 0.945312 (64.275 sec) +INFO:tensorflow:global_step/sec: 1.55692 +INFO:tensorflow:step = 36701, loss = 0.332977, precision = 1.0 (64.229 sec) +Saved checkpoint after 94 epoch(s) to ../data/resnet164/checkpoints/00094... +INFO:tensorflow:global_step/sec: 1.50959 +INFO:tensorflow:step = 36801, loss = 0.358423, precision = 0.992188 (66.243 sec) +INFO:tensorflow:global_step/sec: 1.55719 +INFO:tensorflow:step = 36901, loss = 0.36747, precision = 0.976562 (64.218 sec) +INFO:tensorflow:global_step/sec: 1.55678 +INFO:tensorflow:step = 37001, loss = 0.358146, precision = 0.976562 (64.235 sec) +INFO:tensorflow:global_step/sec: 1.5576 +INFO:tensorflow:step = 37101, loss = 0.367659, precision = 0.984375 (64.201 sec) +Saved checkpoint after 95 epoch(s) to ../data/resnet164/checkpoints/00095... +INFO:tensorflow:global_step/sec: 1.5096 +INFO:tensorflow:step = 37201, loss = 0.355707, precision = 0.976562 (66.243 sec) +INFO:tensorflow:global_step/sec: 1.55737 +INFO:tensorflow:step = 37301, loss = 0.358915, precision = 0.960938 (64.210 sec) +INFO:tensorflow:global_step/sec: 1.55708 +INFO:tensorflow:step = 37401, loss = 0.344894, precision = 0.984375 (64.223 sec) +INFO:tensorflow:global_step/sec: 1.55527 +INFO:tensorflow:step = 37501, loss = 0.333874, precision = 0.976562 (64.297 sec) +Saved checkpoint after 96 epoch(s) to ../data/resnet164/checkpoints/00096... +INFO:tensorflow:global_step/sec: 1.5113 +INFO:tensorflow:step = 37601, loss = 0.386406, precision = 0.96875 (66.168 sec) +INFO:tensorflow:global_step/sec: 1.56067 +INFO:tensorflow:step = 37701, loss = 0.415048, precision = 0.945312 (64.075 sec) +INFO:tensorflow:global_step/sec: 1.56058 +INFO:tensorflow:step = 37801, loss = 0.331627, precision = 0.960938 (64.079 sec) +INFO:tensorflow:global_step/sec: 1.55918 +INFO:tensorflow:step = 37901, loss = 0.368064, precision = 0.96875 (64.136 sec) +Saved checkpoint after 97 epoch(s) to ../data/resnet164/checkpoints/00097... +INFO:tensorflow:global_step/sec: 1.51074 +INFO:tensorflow:step = 38001, loss = 0.329581, precision = 0.984375 (66.193 sec) +INFO:tensorflow:global_step/sec: 1.55903 +INFO:tensorflow:step = 38101, loss = 0.331151, precision = 0.984375 (64.142 sec) +INFO:tensorflow:global_step/sec: 1.55933 +INFO:tensorflow:step = 38201, loss = 0.411326, precision = 0.96875 (64.130 sec) +INFO:tensorflow:global_step/sec: 1.55882 +INFO:tensorflow:step = 38301, loss = 0.338947, precision = 0.976562 (64.151 sec) +Saved checkpoint after 98 epoch(s) to ../data/resnet164/checkpoints/00098... +INFO:tensorflow:global_step/sec: 1.50997 +INFO:tensorflow:step = 38401, loss = 0.316262, precision = 0.984375 (66.227 sec) +INFO:tensorflow:global_step/sec: 1.55976 +INFO:tensorflow:step = 38501, loss = 0.324199, precision = 0.96875 (64.112 sec) +INFO:tensorflow:global_step/sec: 1.55921 +INFO:tensorflow:step = 38601, loss = 0.283103, precision = 0.992188 (64.135 sec) +INFO:tensorflow:global_step/sec: 1.56013 +INFO:tensorflow:step = 38701, loss = 0.348006, precision = 0.960938 (64.097 sec) +Saved checkpoint after 99 epoch(s) to ../data/resnet164/checkpoints/00099... +INFO:tensorflow:global_step/sec: 1.50977 +INFO:tensorflow:step = 38801, loss = 0.309241, precision = 0.984375 (66.235 sec) +INFO:tensorflow:global_step/sec: 1.55976 +INFO:tensorflow:step = 38901, loss = 0.343737, precision = 0.976562 (64.112 sec) +INFO:tensorflow:global_step/sec: 1.55884 +INFO:tensorflow:step = 39001, loss = 0.34375, precision = 0.960938 (64.150 sec) +Saved checkpoint after 100 epoch(s) to ../data/resnet164/checkpoints/00100... +INFO:tensorflow:global_step/sec: 1.50562 +INFO:tensorflow:step = 39101, loss = 0.264023, precision = 1.0 (66.418 sec) +INFO:tensorflow:global_step/sec: 1.56047 +INFO:tensorflow:step = 39201, loss = 0.307587, precision = 0.984375 (64.084 sec) +INFO:tensorflow:global_step/sec: 1.55933 +INFO:tensorflow:step = 39301, loss = 0.321991, precision = 0.96875 (64.130 sec) +INFO:tensorflow:global_step/sec: 1.55937 +INFO:tensorflow:step = 39401, loss = 0.26543, precision = 0.984375 (64.129 sec) +Saved checkpoint after 101 epoch(s) to ../data/resnet164/checkpoints/00101... +INFO:tensorflow:global_step/sec: 1.51131 +INFO:tensorflow:step = 39501, loss = 0.271393, precision = 0.984375 (66.168 sec) +INFO:tensorflow:global_step/sec: 1.56065 +INFO:tensorflow:step = 39601, loss = 0.289209, precision = 0.976562 (64.076 sec) +INFO:tensorflow:global_step/sec: 1.559 +INFO:tensorflow:step = 39701, loss = 0.282073, precision = 0.984375 (64.144 sec) +INFO:tensorflow:global_step/sec: 1.55936 +INFO:tensorflow:step = 39801, loss = 0.31059, precision = 0.96875 (64.129 sec) +Saved checkpoint after 102 epoch(s) to ../data/resnet164/checkpoints/00102... +INFO:tensorflow:global_step/sec: 1.50985 +INFO:tensorflow:step = 39901, loss = 0.258152, precision = 0.992188 (66.232 sec) +INFO:tensorflow:global_step/sec: 1.55944 +INFO:tensorflow:step = 40001, loss = 0.283217, precision = 0.96875 (64.126 sec) +INFO:tensorflow:global_step/sec: 1.55983 +INFO:tensorflow:step = 40101, loss = 0.305152, precision = 0.984375 (64.110 sec) +INFO:tensorflow:global_step/sec: 1.55969 +INFO:tensorflow:step = 40201, loss = 0.270145, precision = 0.992188 (64.115 sec) +Saved checkpoint after 103 epoch(s) to ../data/resnet164/checkpoints/00103... +INFO:tensorflow:global_step/sec: 1.51015 +INFO:tensorflow:step = 40301, loss = 0.24524, precision = 0.992188 (66.218 sec) +INFO:tensorflow:global_step/sec: 1.55919 +INFO:tensorflow:step = 40401, loss = 0.278054, precision = 0.984375 (64.136 sec) +INFO:tensorflow:global_step/sec: 1.55886 +INFO:tensorflow:step = 40501, loss = 0.301363, precision = 0.960938 (64.149 sec) +INFO:tensorflow:global_step/sec: 1.55965 +INFO:tensorflow:step = 40601, loss = 0.296359, precision = 0.96875 (64.117 sec) +Saved checkpoint after 104 epoch(s) to ../data/resnet164/checkpoints/00104... +INFO:tensorflow:global_step/sec: 1.50617 +INFO:tensorflow:step = 40701, loss = 0.245113, precision = 0.976562 (66.394 sec) +INFO:tensorflow:global_step/sec: 1.55895 +INFO:tensorflow:step = 40801, loss = 0.243024, precision = 0.992188 (64.146 sec) +INFO:tensorflow:global_step/sec: 1.55884 +INFO:tensorflow:step = 40901, loss = 0.266032, precision = 0.976562 (64.150 sec) +INFO:tensorflow:global_step/sec: 1.55862 +INFO:tensorflow:step = 41001, loss = 0.259558, precision = 0.984375 (64.160 sec) +Saved checkpoint after 105 epoch(s) to ../data/resnet164/checkpoints/00105... +INFO:tensorflow:global_step/sec: 1.51071 +INFO:tensorflow:step = 41101, loss = 0.23381, precision = 0.992188 (66.194 sec) +INFO:tensorflow:global_step/sec: 1.55879 +INFO:tensorflow:step = 41201, loss = 0.24898, precision = 0.976562 (64.153 sec) +INFO:tensorflow:global_step/sec: 1.55865 +INFO:tensorflow:step = 41301, loss = 0.268608, precision = 0.976562 (64.158 sec) +INFO:tensorflow:global_step/sec: 1.55936 +INFO:tensorflow:step = 41401, loss = 0.24707, precision = 0.984375 (64.129 sec) +Saved checkpoint after 106 epoch(s) to ../data/resnet164/checkpoints/00106... +INFO:tensorflow:global_step/sec: 1.51131 +INFO:tensorflow:step = 41501, loss = 0.284456, precision = 0.984375 (66.168 sec) +INFO:tensorflow:global_step/sec: 1.55902 +INFO:tensorflow:step = 41601, loss = 0.265791, precision = 0.96875 (64.143 sec) +INFO:tensorflow:global_step/sec: 1.55857 +INFO:tensorflow:step = 41701, loss = 0.300427, precision = 0.976562 (64.161 sec) +INFO:tensorflow:global_step/sec: 1.5582 +INFO:tensorflow:step = 41801, loss = 0.237695, precision = 0.984375 (64.177 sec) +Saved checkpoint after 107 epoch(s) to ../data/resnet164/checkpoints/00107... +INFO:tensorflow:global_step/sec: 1.51017 +INFO:tensorflow:step = 41901, loss = 0.230931, precision = 0.984375 (66.218 sec) +INFO:tensorflow:global_step/sec: 1.55854 +INFO:tensorflow:step = 42001, loss = 0.232735, precision = 0.984375 (64.162 sec) +INFO:tensorflow:global_step/sec: 1.55902 +INFO:tensorflow:step = 42101, loss = 0.232785, precision = 0.984375 (64.143 sec) +INFO:tensorflow:global_step/sec: 1.55841 +INFO:tensorflow:step = 42201, loss = 0.219784, precision = 0.992188 (64.168 sec) +Saved checkpoint after 108 epoch(s) to ../data/resnet164/checkpoints/00108... +INFO:tensorflow:global_step/sec: 1.50895 +INFO:tensorflow:step = 42301, loss = 0.243665, precision = 0.976562 (66.271 sec) +INFO:tensorflow:global_step/sec: 1.55901 +INFO:tensorflow:step = 42401, loss = 0.232266, precision = 0.984375 (64.143 sec) +INFO:tensorflow:global_step/sec: 1.55744 +INFO:tensorflow:step = 42501, loss = 0.212406, precision = 0.992188 (64.208 sec) +INFO:tensorflow:global_step/sec: 1.55798 +INFO:tensorflow:step = 42601, loss = 0.21143, precision = 1.0 (64.186 sec) +Saved checkpoint after 109 epoch(s) to ../data/resnet164/checkpoints/00109... +INFO:tensorflow:global_step/sec: 1.51225 +INFO:tensorflow:step = 42701, loss = 0.255661, precision = 0.984375 (66.126 sec) +INFO:tensorflow:global_step/sec: 1.56006 +INFO:tensorflow:step = 42801, loss = 0.208834, precision = 0.984375 (64.100 sec) +INFO:tensorflow:global_step/sec: 1.56285 +INFO:tensorflow:step = 42901, loss = 0.204066, precision = 1.0 (63.985 sec) +INFO:tensorflow:global_step/sec: 1.56099 +INFO:tensorflow:step = 43001, loss = 0.212912, precision = 0.992188 (64.062 sec) +Saved checkpoint after 110 epoch(s) to ../data/resnet164/checkpoints/00110... +INFO:tensorflow:global_step/sec: 1.50625 +INFO:tensorflow:step = 43101, loss = 0.234989, precision = 0.976562 (66.390 sec) +INFO:tensorflow:global_step/sec: 1.55903 +INFO:tensorflow:step = 43201, loss = 0.23538, precision = 0.96875 (64.143 sec) +INFO:tensorflow:global_step/sec: 1.55957 +INFO:tensorflow:step = 43301, loss = 0.223121, precision = 0.976562 (64.120 sec) +Saved checkpoint after 111 epoch(s) to ../data/resnet164/checkpoints/00111... +INFO:tensorflow:global_step/sec: 1.50927 +INFO:tensorflow:step = 43401, loss = 0.242982, precision = 0.984375 (66.257 sec) +INFO:tensorflow:global_step/sec: 1.55999 +INFO:tensorflow:step = 43501, loss = 0.211297, precision = 0.992188 (64.103 sec) +INFO:tensorflow:global_step/sec: 1.55893 +INFO:tensorflow:step = 43601, loss = 0.241109, precision = 0.992188 (64.147 sec) +INFO:tensorflow:global_step/sec: 1.5598 +INFO:tensorflow:step = 43701, loss = 0.279842, precision = 0.960938 (64.111 sec) +Saved checkpoint after 112 epoch(s) to ../data/resnet164/checkpoints/00112... +INFO:tensorflow:global_step/sec: 1.51197 +INFO:tensorflow:step = 43801, loss = 0.254937, precision = 0.984375 (66.139 sec) +INFO:tensorflow:global_step/sec: 1.56022 +INFO:tensorflow:step = 43901, loss = 0.210464, precision = 1.0 (64.093 sec) +INFO:tensorflow:global_step/sec: 1.55977 +INFO:tensorflow:step = 44001, loss = 0.204109, precision = 0.992188 (64.112 sec) +INFO:tensorflow:global_step/sec: 1.55925 +INFO:tensorflow:step = 44101, loss = 0.24272, precision = 0.960938 (64.133 sec) +Saved checkpoint after 113 epoch(s) to ../data/resnet164/checkpoints/00113... +INFO:tensorflow:global_step/sec: 1.51008 +INFO:tensorflow:step = 44201, loss = 0.198875, precision = 0.992188 (66.222 sec) +INFO:tensorflow:global_step/sec: 1.5612 +INFO:tensorflow:step = 44301, loss = 0.204621, precision = 0.984375 (64.053 sec) +INFO:tensorflow:global_step/sec: 1.55954 +INFO:tensorflow:step = 44401, loss = 0.253383, precision = 0.976562 (64.122 sec) +INFO:tensorflow:global_step/sec: 1.55951 +INFO:tensorflow:step = 44501, loss = 0.188927, precision = 1.0 (64.123 sec) +Saved checkpoint after 114 epoch(s) to ../data/resnet164/checkpoints/00114... +INFO:tensorflow:global_step/sec: 1.51014 +INFO:tensorflow:step = 44601, loss = 0.18608, precision = 0.992188 (66.219 sec) +INFO:tensorflow:global_step/sec: 1.56075 +INFO:tensorflow:step = 44701, loss = 0.220923, precision = 0.984375 (64.072 sec) +INFO:tensorflow:global_step/sec: 1.55977 +INFO:tensorflow:step = 44801, loss = 0.209407, precision = 0.984375 (64.112 sec) +INFO:tensorflow:global_step/sec: 1.56017 +INFO:tensorflow:step = 44901, loss = 0.188773, precision = 1.0 (64.096 sec) +Saved checkpoint after 115 epoch(s) to ../data/resnet164/checkpoints/00115... +INFO:tensorflow:global_step/sec: 1.51009 +INFO:tensorflow:step = 45001, loss = 0.212436, precision = 0.984375 (66.221 sec) +INFO:tensorflow:global_step/sec: 1.55939 +INFO:tensorflow:step = 45101, loss = 0.225216, precision = 0.96875 (64.128 sec) +INFO:tensorflow:global_step/sec: 1.56071 +INFO:tensorflow:step = 45201, loss = 0.23178, precision = 0.96875 (64.073 sec) +INFO:tensorflow:global_step/sec: 1.55963 +INFO:tensorflow:step = 45301, loss = 0.204526, precision = 0.992188 (64.118 sec) +Saved checkpoint after 116 epoch(s) to ../data/resnet164/checkpoints/00116... +INFO:tensorflow:global_step/sec: 1.50993 +INFO:tensorflow:step = 45401, loss = 0.257109, precision = 0.953125 (66.228 sec) +INFO:tensorflow:global_step/sec: 1.56028 +INFO:tensorflow:step = 45501, loss = 0.275631, precision = 0.96875 (64.091 sec) +INFO:tensorflow:global_step/sec: 1.55994 +INFO:tensorflow:step = 45601, loss = 0.259229, precision = 0.953125 (64.105 sec) +INFO:tensorflow:global_step/sec: 1.56004 +INFO:tensorflow:step = 45701, loss = 0.21606, precision = 0.96875 (64.101 sec) +Saved checkpoint after 117 epoch(s) to ../data/resnet164/checkpoints/00117... +INFO:tensorflow:global_step/sec: 1.51075 +INFO:tensorflow:step = 45801, loss = 0.209175, precision = 0.992188 (66.192 sec) +INFO:tensorflow:global_step/sec: 1.55949 +INFO:tensorflow:step = 45901, loss = 0.244817, precision = 0.976562 (64.123 sec) +INFO:tensorflow:global_step/sec: 1.55944 +INFO:tensorflow:step = 46001, loss = 0.206592, precision = 0.984375 (64.126 sec) +INFO:tensorflow:global_step/sec: 1.56002 +INFO:tensorflow:step = 46101, loss = 0.190449, precision = 0.992188 (64.102 sec) +Saved checkpoint after 118 epoch(s) to ../data/resnet164/checkpoints/00118... +INFO:tensorflow:global_step/sec: 1.51058 +INFO:tensorflow:step = 46201, loss = 0.183692, precision = 1.0 (66.200 sec) +INFO:tensorflow:global_step/sec: 1.55905 +INFO:tensorflow:step = 46301, loss = 0.236168, precision = 0.96875 (64.141 sec) +INFO:tensorflow:global_step/sec: 1.5595 +INFO:tensorflow:step = 46401, loss = 0.227833, precision = 0.976562 (64.123 sec) +INFO:tensorflow:global_step/sec: 1.55985 +INFO:tensorflow:step = 46501, loss = 0.223616, precision = 0.976562 (64.109 sec) +Saved checkpoint after 119 epoch(s) to ../data/resnet164/checkpoints/00119... +INFO:tensorflow:global_step/sec: 1.51006 +INFO:tensorflow:step = 46601, loss = 0.186049, precision = 0.992188 (66.223 sec) +INFO:tensorflow:global_step/sec: 1.55957 +INFO:tensorflow:step = 46701, loss = 0.179404, precision = 1.0 (64.120 sec) +INFO:tensorflow:global_step/sec: 1.55964 +INFO:tensorflow:step = 46801, loss = 0.215352, precision = 0.984375 (64.117 sec) +INFO:tensorflow:global_step/sec: 1.55913 +INFO:tensorflow:step = 46901, loss = 0.21439, precision = 0.984375 (64.138 sec) +Saved checkpoint after 120 epoch(s) to ../data/resnet164/checkpoints/00120... +INFO:tensorflow:global_step/sec: 1.50552 +INFO:tensorflow:step = 47001, loss = 0.190554, precision = 0.984375 (66.422 sec) +INFO:tensorflow:global_step/sec: 1.56021 +INFO:tensorflow:step = 47101, loss = 0.184588, precision = 0.976562 (64.094 sec) +INFO:tensorflow:global_step/sec: 1.56 +INFO:tensorflow:step = 47201, loss = 0.198872, precision = 0.976562 (64.102 sec) +INFO:tensorflow:global_step/sec: 1.56005 +INFO:tensorflow:step = 47301, loss = 0.207008, precision = 0.992188 (64.100 sec) +Saved checkpoint after 121 epoch(s) to ../data/resnet164/checkpoints/00121... +INFO:tensorflow:global_step/sec: 1.51004 +INFO:tensorflow:step = 47401, loss = 0.177384, precision = 0.992188 (66.224 sec) +INFO:tensorflow:global_step/sec: 1.5605 +INFO:tensorflow:step = 47501, loss = 0.195025, precision = 0.992188 (64.082 sec) +INFO:tensorflow:global_step/sec: 1.55907 +INFO:tensorflow:step = 47601, loss = 0.198638, precision = 0.992188 (64.141 sec) +INFO:tensorflow:global_step/sec: 1.5591 +INFO:tensorflow:step = 47701, loss = 0.205832, precision = 0.992188 (64.139 sec) +Saved checkpoint after 122 epoch(s) to ../data/resnet164/checkpoints/00122... +INFO:tensorflow:global_step/sec: 1.51037 +INFO:tensorflow:step = 47801, loss = 0.229537, precision = 0.96875 (66.209 sec) +INFO:tensorflow:global_step/sec: 1.55975 +INFO:tensorflow:step = 47901, loss = 0.206509, precision = 0.984375 (64.113 sec) +INFO:tensorflow:global_step/sec: 1.56076 +INFO:tensorflow:step = 48001, loss = 0.25635, precision = 0.984375 (64.071 sec) +Saved checkpoint after 123 epoch(s) to ../data/resnet164/checkpoints/00123... +INFO:tensorflow:global_step/sec: 1.50917 +INFO:tensorflow:step = 48101, loss = 0.158144, precision = 1.0 (66.261 sec) +INFO:tensorflow:global_step/sec: 1.55903 +INFO:tensorflow:step = 48201, loss = 0.207726, precision = 0.960938 (64.142 sec) +INFO:tensorflow:global_step/sec: 1.5587 +INFO:tensorflow:step = 48301, loss = 0.1675, precision = 0.992188 (64.156 sec) +INFO:tensorflow:global_step/sec: 1.56004 +INFO:tensorflow:step = 48401, loss = 0.201936, precision = 0.984375 (64.101 sec) +Saved checkpoint after 124 epoch(s) to ../data/resnet164/checkpoints/00124... +INFO:tensorflow:global_step/sec: 1.51061 +INFO:tensorflow:step = 48501, loss = 0.174629, precision = 0.984375 (66.198 sec) +INFO:tensorflow:global_step/sec: 1.55964 +INFO:tensorflow:step = 48601, loss = 0.193887, precision = 0.984375 (64.117 sec) +INFO:tensorflow:global_step/sec: 1.55949 +INFO:tensorflow:step = 48701, loss = 0.16072, precision = 1.0 (64.124 sec) +INFO:tensorflow:global_step/sec: 1.55921 +INFO:tensorflow:step = 48801, loss = 0.204331, precision = 0.984375 (64.135 sec) +Saved checkpoint after 125 epoch(s) to ../data/resnet164/checkpoints/00125... +INFO:tensorflow:global_step/sec: 1.50982 +INFO:tensorflow:step = 48901, loss = 0.224227, precision = 0.960938 (66.233 sec) +INFO:tensorflow:global_step/sec: 1.55996 +INFO:tensorflow:step = 49001, loss = 0.166275, precision = 1.0 (64.104 sec) +INFO:tensorflow:global_step/sec: 1.55949 +INFO:tensorflow:step = 49101, loss = 0.251713, precision = 0.96875 (64.123 sec) +INFO:tensorflow:global_step/sec: 1.55942 +INFO:tensorflow:step = 49201, loss = 0.312352, precision = 0.945312 (64.126 sec) +Saved checkpoint after 126 epoch(s) to ../data/resnet164/checkpoints/00126... +INFO:tensorflow:global_step/sec: 1.5102 +INFO:tensorflow:step = 49301, loss = 0.187931, precision = 0.984375 (66.216 sec) +INFO:tensorflow:global_step/sec: 1.56026 +INFO:tensorflow:step = 49401, loss = 0.188361, precision = 0.976562 (64.092 sec) +INFO:tensorflow:global_step/sec: 1.56004 +INFO:tensorflow:step = 49501, loss = 0.233682, precision = 0.960938 (64.101 sec) +INFO:tensorflow:global_step/sec: 1.55897 +INFO:tensorflow:step = 49601, loss = 0.170785, precision = 0.984375 (64.145 sec) +Saved checkpoint after 127 epoch(s) to ../data/resnet164/checkpoints/00127... +INFO:tensorflow:global_step/sec: 1.50965 +INFO:tensorflow:step = 49701, loss = 0.16908, precision = 1.0 (66.241 sec) +INFO:tensorflow:global_step/sec: 1.55991 +INFO:tensorflow:step = 49801, loss = 0.174517, precision = 0.984375 (64.106 sec) +INFO:tensorflow:global_step/sec: 1.55828 +INFO:tensorflow:step = 49901, loss = 0.159227, precision = 1.0 (64.173 sec) +INFO:tensorflow:global_step/sec: 1.55984 +INFO:tensorflow:step = 50001, loss = 0.208618, precision = 0.960938 (64.109 sec) +Saved checkpoint after 128 epoch(s) to ../data/resnet164/checkpoints/00128... +INFO:tensorflow:global_step/sec: 1.50958 +INFO:tensorflow:step = 50101, loss = 0.154907, precision = 1.0 (66.243 sec) +INFO:tensorflow:global_step/sec: 1.55977 +INFO:tensorflow:step = 50201, loss = 0.163381, precision = 0.992188 (64.112 sec) +INFO:tensorflow:global_step/sec: 1.56023 +INFO:tensorflow:step = 50301, loss = 0.170725, precision = 0.992188 (64.093 sec) +INFO:tensorflow:global_step/sec: 1.55923 +INFO:tensorflow:step = 50401, loss = 0.213822, precision = 0.984375 (64.134 sec) +Saved checkpoint after 129 epoch(s) to ../data/resnet164/checkpoints/00129... +INFO:tensorflow:global_step/sec: 1.51024 +INFO:tensorflow:step = 50501, loss = 0.219297, precision = 0.96875 (66.215 sec) +INFO:tensorflow:global_step/sec: 1.55918 +INFO:tensorflow:step = 50601, loss = 0.28696, precision = 0.953125 (64.136 sec) +INFO:tensorflow:global_step/sec: 1.56021 +INFO:tensorflow:step = 50701, loss = 0.164396, precision = 0.992188 (64.094 sec) +INFO:tensorflow:global_step/sec: 1.56011 +INFO:tensorflow:step = 50801, loss = 0.280081, precision = 0.953125 (64.098 sec) +Saved checkpoint after 130 epoch(s) to ../data/resnet164/checkpoints/00130... +INFO:tensorflow:global_step/sec: 1.51 +INFO:tensorflow:step = 50901, loss = 0.174216, precision = 0.992188 (66.225 sec) +INFO:tensorflow:global_step/sec: 1.55919 +INFO:tensorflow:step = 51001, loss = 0.198369, precision = 0.976562 (64.136 sec) +INFO:tensorflow:global_step/sec: 1.56065 +INFO:tensorflow:step = 51101, loss = 0.195217, precision = 0.984375 (64.076 sec) +INFO:tensorflow:global_step/sec: 1.56012 +INFO:tensorflow:step = 51201, loss = 0.194907, precision = 0.976562 (64.098 sec) +Saved checkpoint after 131 epoch(s) to ../data/resnet164/checkpoints/00131... +INFO:tensorflow:global_step/sec: 1.50577 +INFO:tensorflow:step = 51301, loss = 0.182123, precision = 0.984375 (66.411 sec) +INFO:tensorflow:global_step/sec: 1.55951 +INFO:tensorflow:step = 51401, loss = 0.252125, precision = 0.960938 (64.123 sec) +INFO:tensorflow:global_step/sec: 1.55892 +INFO:tensorflow:step = 51501, loss = 0.236738, precision = 0.953125 (64.147 sec) +INFO:tensorflow:global_step/sec: 1.55956 +INFO:tensorflow:step = 51601, loss = 0.168293, precision = 0.992188 (64.121 sec) +Saved checkpoint after 132 epoch(s) to ../data/resnet164/checkpoints/00132... +INFO:tensorflow:global_step/sec: 1.51137 +INFO:tensorflow:step = 51701, loss = 0.2612, precision = 0.945312 (66.165 sec) +INFO:tensorflow:global_step/sec: 1.55988 +INFO:tensorflow:step = 51801, loss = 0.193644, precision = 0.984375 (64.108 sec) +INFO:tensorflow:global_step/sec: 1.55974 +INFO:tensorflow:step = 51901, loss = 0.186636, precision = 0.984375 (64.113 sec) +INFO:tensorflow:global_step/sec: 1.55957 +INFO:tensorflow:step = 52001, loss = 0.182819, precision = 0.976562 (64.120 sec) +Saved checkpoint after 133 epoch(s) to ../data/resnet164/checkpoints/00133... +INFO:tensorflow:global_step/sec: 1.50994 +INFO:tensorflow:step = 52101, loss = 0.197291, precision = 0.984375 (66.228 sec) +INFO:tensorflow:global_step/sec: 1.55947 +INFO:tensorflow:step = 52201, loss = 0.175568, precision = 0.984375 (64.125 sec) +INFO:tensorflow:global_step/sec: 1.55892 +INFO:tensorflow:step = 52301, loss = 0.178371, precision = 0.984375 (64.147 sec) +Saved checkpoint after 134 epoch(s) to ../data/resnet164/checkpoints/00134... +INFO:tensorflow:global_step/sec: 1.51099 +INFO:tensorflow:step = 52401, loss = 0.291029, precision = 0.953125 (66.182 sec) +INFO:tensorflow:global_step/sec: 1.55895 +INFO:tensorflow:step = 52501, loss = 0.173956, precision = 0.984375 (64.145 sec) +INFO:tensorflow:global_step/sec: 1.55931 +INFO:tensorflow:step = 52601, loss = 0.172647, precision = 0.992188 (64.131 sec) +INFO:tensorflow:global_step/sec: 1.55894 +INFO:tensorflow:step = 52701, loss = 0.182415, precision = 0.984375 (64.146 sec) +Saved checkpoint after 135 epoch(s) to ../data/resnet164/checkpoints/00135... +INFO:tensorflow:global_step/sec: 1.50992 +INFO:tensorflow:step = 52801, loss = 0.160527, precision = 0.992188 (66.229 sec) +INFO:tensorflow:global_step/sec: 1.55917 +INFO:tensorflow:step = 52901, loss = 0.19357, precision = 0.984375 (64.137 sec) +INFO:tensorflow:global_step/sec: 1.55945 +INFO:tensorflow:step = 53001, loss = 0.154943, precision = 1.0 (64.125 sec) +INFO:tensorflow:global_step/sec: 1.56032 +INFO:tensorflow:step = 53101, loss = 0.212342, precision = 0.96875 (64.089 sec) +Saved checkpoint after 136 epoch(s) to ../data/resnet164/checkpoints/00136... +INFO:tensorflow:global_step/sec: 1.50913 +INFO:tensorflow:step = 53201, loss = 0.147092, precision = 1.0 (66.264 sec) +INFO:tensorflow:global_step/sec: 1.55969 +INFO:tensorflow:step = 53301, loss = 0.156199, precision = 1.0 (64.115 sec) +INFO:tensorflow:global_step/sec: 1.55939 +INFO:tensorflow:step = 53401, loss = 0.154117, precision = 1.0 (64.128 sec) +INFO:tensorflow:global_step/sec: 1.55875 +INFO:tensorflow:step = 53501, loss = 0.156163, precision = 0.992188 (64.154 sec) +Saved checkpoint after 137 epoch(s) to ../data/resnet164/checkpoints/00137... +INFO:tensorflow:global_step/sec: 1.51017 +INFO:tensorflow:step = 53601, loss = 0.153425, precision = 1.0 (66.218 sec) +INFO:tensorflow:global_step/sec: 1.55959 +INFO:tensorflow:step = 53701, loss = 0.155571, precision = 0.992188 (64.119 sec) +INFO:tensorflow:global_step/sec: 1.55994 +INFO:tensorflow:step = 53801, loss = 0.141542, precision = 1.0 (64.105 sec) +INFO:tensorflow:global_step/sec: 1.55963 +INFO:tensorflow:step = 53901, loss = 0.141391, precision = 1.0 (64.118 sec) +Saved checkpoint after 138 epoch(s) to ../data/resnet164/checkpoints/00138... +INFO:tensorflow:global_step/sec: 1.51106 +INFO:tensorflow:step = 54001, loss = 0.149703, precision = 0.992188 (66.179 sec) +INFO:tensorflow:global_step/sec: 1.56033 +INFO:tensorflow:step = 54101, loss = 0.140171, precision = 1.0 (64.089 sec) +INFO:tensorflow:global_step/sec: 1.55944 +INFO:tensorflow:step = 54201, loss = 0.153272, precision = 1.0 (64.126 sec) +INFO:tensorflow:global_step/sec: 1.55955 +INFO:tensorflow:step = 54301, loss = 0.141548, precision = 1.0 (64.121 sec) +Saved checkpoint after 139 epoch(s) to ../data/resnet164/checkpoints/00139... +INFO:tensorflow:global_step/sec: 1.50998 +INFO:tensorflow:step = 54401, loss = 0.161917, precision = 0.992188 (66.226 sec) +INFO:tensorflow:global_step/sec: 1.56002 +INFO:tensorflow:step = 54501, loss = 0.146643, precision = 1.0 (64.102 sec) +INFO:tensorflow:global_step/sec: 1.55941 +INFO:tensorflow:step = 54601, loss = 0.14568, precision = 0.992188 (64.127 sec) +INFO:tensorflow:global_step/sec: 1.55987 +INFO:tensorflow:step = 54701, loss = 0.139196, precision = 1.0 (64.108 sec) +Saved checkpoint after 140 epoch(s) to ../data/resnet164/checkpoints/00140... +INFO:tensorflow:global_step/sec: 1.51104 +INFO:tensorflow:step = 54801, loss = 0.144777, precision = 1.0 (66.180 sec) +INFO:tensorflow:global_step/sec: 1.56058 +INFO:tensorflow:step = 54901, loss = 0.13897, precision = 1.0 (64.079 sec) +INFO:tensorflow:global_step/sec: 1.55962 +INFO:tensorflow:step = 55001, loss = 0.137707, precision = 1.0 (64.118 sec) +INFO:tensorflow:global_step/sec: 1.55964 +INFO:tensorflow:step = 55101, loss = 0.140332, precision = 1.0 (64.117 sec) +Saved checkpoint after 141 epoch(s) to ../data/resnet164/checkpoints/00141... +INFO:tensorflow:global_step/sec: 1.50619 +INFO:tensorflow:step = 55201, loss = 0.184275, precision = 0.984375 (66.393 sec) +INFO:tensorflow:global_step/sec: 1.55983 +INFO:tensorflow:step = 55301, loss = 0.136833, precision = 1.0 (64.109 sec) +INFO:tensorflow:global_step/sec: 1.56018 +INFO:tensorflow:step = 55401, loss = 0.14779, precision = 0.992188 (64.095 sec) +INFO:tensorflow:global_step/sec: 1.55929 +INFO:tensorflow:step = 55501, loss = 0.135586, precision = 1.0 (64.132 sec) +Saved checkpoint after 142 epoch(s) to ../data/resnet164/checkpoints/00142... +INFO:tensorflow:global_step/sec: 1.51148 +INFO:tensorflow:step = 55601, loss = 0.134854, precision = 1.0 (66.160 sec) +INFO:tensorflow:global_step/sec: 1.56007 +INFO:tensorflow:step = 55701, loss = 0.158486, precision = 0.984375 (64.099 sec) +INFO:tensorflow:global_step/sec: 1.55848 +INFO:tensorflow:step = 55801, loss = 0.135762, precision = 1.0 (64.165 sec) +INFO:tensorflow:global_step/sec: 1.55886 +INFO:tensorflow:step = 55901, loss = 0.139393, precision = 1.0 (64.149 sec) +Saved checkpoint after 143 epoch(s) to ../data/resnet164/checkpoints/00143... +INFO:tensorflow:global_step/sec: 1.50934 +INFO:tensorflow:step = 56001, loss = 0.151102, precision = 0.992188 (66.254 sec) +INFO:tensorflow:global_step/sec: 1.55959 +INFO:tensorflow:step = 56101, loss = 0.138876, precision = 1.0 (64.119 sec) +INFO:tensorflow:global_step/sec: 1.55981 +INFO:tensorflow:step = 56201, loss = 0.138353, precision = 1.0 (64.110 sec) +INFO:tensorflow:global_step/sec: 1.56063 +INFO:tensorflow:step = 56301, loss = 0.135595, precision = 1.0 (64.077 sec) +Saved checkpoint after 144 epoch(s) to ../data/resnet164/checkpoints/00144... +INFO:tensorflow:global_step/sec: 1.51073 +INFO:tensorflow:step = 56401, loss = 0.135065, precision = 1.0 (66.193 sec) +INFO:tensorflow:global_step/sec: 1.56018 +INFO:tensorflow:step = 56501, loss = 0.137083, precision = 1.0 (64.095 sec) +INFO:tensorflow:global_step/sec: 1.55941 +INFO:tensorflow:step = 56601, loss = 0.135313, precision = 1.0 (64.127 sec) +Saved checkpoint after 145 epoch(s) to ../data/resnet164/checkpoints/00145... +INFO:tensorflow:global_step/sec: 1.51083 +INFO:tensorflow:step = 56701, loss = 0.132914, precision = 1.0 (66.189 sec) +INFO:tensorflow:global_step/sec: 1.55936 +INFO:tensorflow:step = 56801, loss = 0.134527, precision = 1.0 (64.129 sec) +INFO:tensorflow:global_step/sec: 1.56055 +INFO:tensorflow:step = 56901, loss = 0.136173, precision = 1.0 (64.080 sec) +INFO:tensorflow:global_step/sec: 1.55982 +INFO:tensorflow:step = 57001, loss = 0.13415, precision = 1.0 (64.110 sec) +Saved checkpoint after 146 epoch(s) to ../data/resnet164/checkpoints/00146... +INFO:tensorflow:global_step/sec: 1.50999 +INFO:tensorflow:step = 57101, loss = 0.140558, precision = 0.992188 (66.225 sec) +INFO:tensorflow:global_step/sec: 1.55997 +INFO:tensorflow:step = 57201, loss = 0.13914, precision = 1.0 (64.104 sec) +INFO:tensorflow:global_step/sec: 1.56092 +INFO:tensorflow:step = 57301, loss = 0.138066, precision = 1.0 (64.065 sec) +INFO:tensorflow:global_step/sec: 1.56008 +INFO:tensorflow:step = 57401, loss = 0.136858, precision = 1.0 (64.100 sec) +Saved checkpoint after 147 epoch(s) to ../data/resnet164/checkpoints/00147... +INFO:tensorflow:global_step/sec: 1.50984 +INFO:tensorflow:step = 57501, loss = 0.137388, precision = 1.0 (66.232 sec) +INFO:tensorflow:global_step/sec: 1.55975 +INFO:tensorflow:step = 57601, loss = 0.165057, precision = 0.992188 (64.113 sec) +INFO:tensorflow:global_step/sec: 1.56031 +INFO:tensorflow:step = 57701, loss = 0.134281, precision = 1.0 (64.090 sec) +INFO:tensorflow:global_step/sec: 1.55959 +INFO:tensorflow:step = 57801, loss = 0.131956, precision = 1.0 (64.119 sec) +Saved checkpoint after 148 epoch(s) to ../data/resnet164/checkpoints/00148... +INFO:tensorflow:global_step/sec: 1.51004 +INFO:tensorflow:step = 57901, loss = 0.135757, precision = 1.0 (66.223 sec) +INFO:tensorflow:global_step/sec: 1.55903 +INFO:tensorflow:step = 58001, loss = 0.134123, precision = 1.0 (64.142 sec) +INFO:tensorflow:global_step/sec: 1.55992 +INFO:tensorflow:step = 58101, loss = 0.13383, precision = 1.0 (64.106 sec) +INFO:tensorflow:global_step/sec: 1.56016 +INFO:tensorflow:step = 58201, loss = 0.132442, precision = 1.0 (64.096 sec) +Saved checkpoint after 149 epoch(s) to ../data/resnet164/checkpoints/00149... +INFO:tensorflow:global_step/sec: 1.51014 +INFO:tensorflow:step = 58301, loss = 0.134616, precision = 1.0 (66.219 sec) +INFO:tensorflow:global_step/sec: 1.55985 +INFO:tensorflow:step = 58401, loss = 0.133238, precision = 1.0 (64.109 sec) +INFO:tensorflow:global_step/sec: 1.55994 +INFO:tensorflow:step = 58501, loss = 0.130403, precision = 1.0 (64.105 sec) +INFO:tensorflow:global_step/sec: 1.55929 +INFO:tensorflow:step = 58601, loss = 0.132455, precision = 1.0 (64.132 sec) +Saved checkpoint after 150 epoch(s) to ../data/resnet164/checkpoints/00150... +INFO:tensorflow:global_step/sec: 1.51102 +INFO:tensorflow:step = 58701, loss = 0.131274, precision = 1.0 (66.180 sec) +INFO:tensorflow:global_step/sec: 1.55906 +INFO:tensorflow:step = 58801, loss = 0.131284, precision = 1.0 (64.141 sec) +INFO:tensorflow:global_step/sec: 1.55953 +INFO:tensorflow:step = 58901, loss = 0.132098, precision = 1.0 (64.122 sec) +INFO:tensorflow:global_step/sec: 1.56017 +INFO:tensorflow:step = 59001, loss = 0.131052, precision = 1.0 (64.096 sec) +Saved checkpoint after 151 epoch(s) to ../data/resnet164/checkpoints/00151... +INFO:tensorflow:global_step/sec: 1.50744 +INFO:tensorflow:step = 59101, loss = 0.143865, precision = 0.992188 (66.338 sec) +INFO:tensorflow:global_step/sec: 1.5595 +INFO:tensorflow:step = 59201, loss = 0.131302, precision = 1.0 (64.123 sec) +INFO:tensorflow:global_step/sec: 1.55922 +INFO:tensorflow:step = 59301, loss = 0.132411, precision = 1.0 (64.135 sec) +INFO:tensorflow:global_step/sec: 1.5588 +INFO:tensorflow:step = 59401, loss = 0.13826, precision = 1.0 (64.152 sec) +Saved checkpoint after 152 epoch(s) to ../data/resnet164/checkpoints/00152... +INFO:tensorflow:global_step/sec: 1.51048 +INFO:tensorflow:step = 59501, loss = 0.133165, precision = 1.0 (66.204 sec) +INFO:tensorflow:global_step/sec: 1.55947 +INFO:tensorflow:step = 59601, loss = 0.134723, precision = 1.0 (64.124 sec) +INFO:tensorflow:global_step/sec: 1.56016 +INFO:tensorflow:step = 59701, loss = 0.13528, precision = 1.0 (64.096 sec) +INFO:tensorflow:global_step/sec: 1.55956 +INFO:tensorflow:step = 59801, loss = 0.13747, precision = 1.0 (64.121 sec) +Saved checkpoint after 153 epoch(s) to ../data/resnet164/checkpoints/00153... +INFO:tensorflow:global_step/sec: 1.51037 +INFO:tensorflow:step = 59901, loss = 0.129668, precision = 1.0 (66.209 sec) +INFO:tensorflow:global_step/sec: 1.5592 +INFO:tensorflow:step = 60001, loss = 0.130488, precision = 1.0 (64.135 sec) +INFO:tensorflow:global_step/sec: 1.55985 +INFO:tensorflow:step = 60101, loss = 0.133604, precision = 1.0 (64.109 sec) +INFO:tensorflow:global_step/sec: 1.56021 +INFO:tensorflow:step = 60201, loss = 0.12947, precision = 1.0 (64.094 sec) +Saved checkpoint after 154 epoch(s) to ../data/resnet164/checkpoints/00154... +INFO:tensorflow:global_step/sec: 1.50956 +INFO:tensorflow:step = 60301, loss = 0.131917, precision = 1.0 (66.245 sec) +INFO:tensorflow:global_step/sec: 1.55969 +INFO:tensorflow:step = 60401, loss = 0.131976, precision = 1.0 (64.116 sec) +INFO:tensorflow:global_step/sec: 1.55989 +INFO:tensorflow:step = 60501, loss = 0.128034, precision = 1.0 (64.107 sec) +INFO:tensorflow:global_step/sec: 1.55994 +INFO:tensorflow:step = 60601, loss = 0.128448, precision = 1.0 (64.105 sec) +Saved checkpoint after 155 epoch(s) to ../data/resnet164/checkpoints/00155... +INFO:tensorflow:global_step/sec: 1.50904 +INFO:tensorflow:step = 60701, loss = 0.129538, precision = 1.0 (66.267 sec) +INFO:tensorflow:global_step/sec: 1.56017 +INFO:tensorflow:step = 60801, loss = 0.12937, precision = 1.0 (64.095 sec) +INFO:tensorflow:global_step/sec: 1.55937 +INFO:tensorflow:step = 60901, loss = 0.133735, precision = 1.0 (64.128 sec) +Saved checkpoint after 156 epoch(s) to ../data/resnet164/checkpoints/00156... +INFO:tensorflow:global_step/sec: 1.51012 +INFO:tensorflow:step = 61001, loss = 0.128789, precision = 1.0 (66.220 sec) +INFO:tensorflow:global_step/sec: 1.55844 +INFO:tensorflow:step = 61101, loss = 0.128323, precision = 1.0 (64.167 sec) +INFO:tensorflow:global_step/sec: 1.55923 +INFO:tensorflow:step = 61201, loss = 0.127625, precision = 1.0 (64.134 sec) +INFO:tensorflow:global_step/sec: 1.55969 +INFO:tensorflow:step = 61301, loss = 0.128739, precision = 1.0 (64.115 sec) +Saved checkpoint after 157 epoch(s) to ../data/resnet164/checkpoints/00157... +INFO:tensorflow:global_step/sec: 1.51153 +INFO:tensorflow:step = 61401, loss = 0.131605, precision = 1.0 (66.158 sec) +INFO:tensorflow:global_step/sec: 1.55927 +INFO:tensorflow:step = 61501, loss = 0.129204, precision = 1.0 (64.132 sec) +INFO:tensorflow:global_step/sec: 1.55936 +INFO:tensorflow:step = 61601, loss = 0.129447, precision = 1.0 (64.129 sec) +INFO:tensorflow:global_step/sec: 1.55894 +INFO:tensorflow:step = 61701, loss = 0.126735, precision = 1.0 (64.146 sec) +Saved checkpoint after 158 epoch(s) to ../data/resnet164/checkpoints/00158... +INFO:tensorflow:global_step/sec: 1.51009 +INFO:tensorflow:step = 61801, loss = 0.126824, precision = 1.0 (66.222 sec) +INFO:tensorflow:global_step/sec: 1.5597 +INFO:tensorflow:step = 61901, loss = 0.127197, precision = 1.0 (64.114 sec) +INFO:tensorflow:global_step/sec: 1.55831 +INFO:tensorflow:step = 62001, loss = 0.138864, precision = 0.992188 (64.172 sec) +INFO:tensorflow:global_step/sec: 1.55883 +INFO:tensorflow:step = 62101, loss = 0.129096, precision = 1.0 (64.151 sec) +Saved checkpoint after 159 epoch(s) to ../data/resnet164/checkpoints/00159... +INFO:tensorflow:global_step/sec: 1.50957 +INFO:tensorflow:step = 62201, loss = 0.154458, precision = 0.992188 (66.244 sec) +INFO:tensorflow:global_step/sec: 1.55879 +INFO:tensorflow:step = 62301, loss = 0.126281, precision = 1.0 (64.152 sec) +INFO:tensorflow:global_step/sec: 1.55907 +INFO:tensorflow:step = 62401, loss = 0.130371, precision = 1.0 (64.141 sec) +INFO:tensorflow:global_step/sec: 1.55819 +INFO:tensorflow:step = 62501, loss = 0.125851, precision = 1.0 (64.177 sec) +Saved checkpoint after 160 epoch(s) to ../data/resnet164/checkpoints/00160... +INFO:tensorflow:global_step/sec: 1.50703 +INFO:tensorflow:step = 62601, loss = 0.140066, precision = 0.992188 (66.356 sec) +INFO:tensorflow:global_step/sec: 1.5588 +INFO:tensorflow:step = 62701, loss = 0.126762, precision = 1.0 (64.152 sec) +INFO:tensorflow:global_step/sec: 1.55989 +INFO:tensorflow:step = 62801, loss = 0.129405, precision = 1.0 (64.107 sec) +INFO:tensorflow:global_step/sec: 1.55896 +INFO:tensorflow:step = 62901, loss = 0.125691, precision = 1.0 (64.145 sec) +Saved checkpoint after 161 epoch(s) to ../data/resnet164/checkpoints/00161... +INFO:tensorflow:global_step/sec: 1.51068 +INFO:tensorflow:step = 63001, loss = 0.135424, precision = 0.992188 (66.195 sec) +INFO:tensorflow:global_step/sec: 1.55764 +INFO:tensorflow:step = 63101, loss = 0.126037, precision = 1.0 (64.200 sec) +INFO:tensorflow:global_step/sec: 1.55746 +INFO:tensorflow:step = 63201, loss = 0.127771, precision = 1.0 (64.207 sec) +INFO:tensorflow:global_step/sec: 1.55984 +INFO:tensorflow:step = 63301, loss = 0.125403, precision = 1.0 (64.109 sec) +Saved checkpoint after 162 epoch(s) to ../data/resnet164/checkpoints/00162... +INFO:tensorflow:global_step/sec: 1.50533 +INFO:tensorflow:step = 63401, loss = 0.125962, precision = 1.0 (66.431 sec) +INFO:tensorflow:global_step/sec: 1.55833 +INFO:tensorflow:step = 63501, loss = 0.128493, precision = 1.0 (64.171 sec) +INFO:tensorflow:global_step/sec: 1.55814 +INFO:tensorflow:step = 63601, loss = 0.125993, precision = 1.0 (64.179 sec) +INFO:tensorflow:global_step/sec: 1.55862 +INFO:tensorflow:step = 63701, loss = 0.126028, precision = 1.0 (64.159 sec) +Saved checkpoint after 163 epoch(s) to ../data/resnet164/checkpoints/00163... +INFO:tensorflow:global_step/sec: 1.51136 +INFO:tensorflow:step = 63801, loss = 0.126055, precision = 1.0 (66.166 sec) +INFO:tensorflow:global_step/sec: 1.55842 +INFO:tensorflow:step = 63901, loss = 0.124925, precision = 1.0 (64.167 sec) +INFO:tensorflow:global_step/sec: 1.55844 +INFO:tensorflow:step = 64001, loss = 0.125447, precision = 1.0 (64.167 sec) +INFO:tensorflow:global_step/sec: 1.55769 +INFO:tensorflow:step = 64101, loss = 0.124275, precision = 1.0 (64.197 sec) +Saved checkpoint after 164 epoch(s) to ../data/resnet164/checkpoints/00164... +INFO:tensorflow:global_step/sec: 1.51096 +INFO:tensorflow:step = 64201, loss = 0.124698, precision = 1.0 (66.183 sec) +INFO:tensorflow:global_step/sec: 1.55874 +INFO:tensorflow:step = 64301, loss = 0.125422, precision = 1.0 (64.154 sec) +INFO:tensorflow:global_step/sec: 1.55826 +INFO:tensorflow:step = 64401, loss = 0.125271, precision = 1.0 (64.174 sec) +INFO:tensorflow:global_step/sec: 1.55861 +INFO:tensorflow:step = 64501, loss = 0.123726, precision = 1.0 (64.160 sec) +Saved checkpoint after 165 epoch(s) to ../data/resnet164/checkpoints/00165... +INFO:tensorflow:global_step/sec: 1.51023 +INFO:tensorflow:step = 64601, loss = 0.126021, precision = 1.0 (66.216 sec) +INFO:tensorflow:global_step/sec: 1.5594 +INFO:tensorflow:step = 64701, loss = 0.123492, precision = 1.0 (64.127 sec) +INFO:tensorflow:global_step/sec: 1.55842 +INFO:tensorflow:step = 64801, loss = 0.12607, precision = 1.0 (64.168 sec) +INFO:tensorflow:global_step/sec: 1.55846 +INFO:tensorflow:step = 64901, loss = 0.128934, precision = 1.0 (64.166 sec) +Saved checkpoint after 166 epoch(s) to ../data/resnet164/checkpoints/00166... +INFO:tensorflow:global_step/sec: 1.51073 +INFO:tensorflow:step = 65001, loss = 0.123705, precision = 1.0 (66.193 sec) +INFO:tensorflow:global_step/sec: 1.55907 +INFO:tensorflow:step = 65101, loss = 0.12291, precision = 1.0 (64.141 sec) +INFO:tensorflow:global_step/sec: 1.55889 +INFO:tensorflow:step = 65201, loss = 0.123109, precision = 1.0 (64.148 sec) +Saved checkpoint after 167 epoch(s) to ../data/resnet164/checkpoints/00167... +INFO:tensorflow:global_step/sec: 1.51071 +INFO:tensorflow:step = 65301, loss = 0.127345, precision = 1.0 (66.194 sec) +INFO:tensorflow:global_step/sec: 1.55852 +INFO:tensorflow:step = 65401, loss = 0.123308, precision = 1.0 (64.163 sec) +INFO:tensorflow:global_step/sec: 1.55922 +INFO:tensorflow:step = 65501, loss = 0.1227, precision = 1.0 (64.134 sec) +INFO:tensorflow:global_step/sec: 1.55903 +INFO:tensorflow:step = 65601, loss = 0.125768, precision = 1.0 (64.143 sec) +Saved checkpoint after 168 epoch(s) to ../data/resnet164/checkpoints/00168... +INFO:tensorflow:global_step/sec: 1.50989 +INFO:tensorflow:step = 65701, loss = 0.122393, precision = 1.0 (66.230 sec) +INFO:tensorflow:global_step/sec: 1.55856 +INFO:tensorflow:step = 65801, loss = 0.121825, precision = 1.0 (64.162 sec) +INFO:tensorflow:global_step/sec: 1.55816 +INFO:tensorflow:step = 65901, loss = 0.122543, precision = 1.0 (64.178 sec) +INFO:tensorflow:global_step/sec: 1.55835 +INFO:tensorflow:step = 66001, loss = 0.122798, precision = 1.0 (64.171 sec) +Saved checkpoint after 169 epoch(s) to ../data/resnet164/checkpoints/00169... +INFO:tensorflow:global_step/sec: 1.50951 +INFO:tensorflow:step = 66101, loss = 0.122646, precision = 1.0 (66.247 sec) +INFO:tensorflow:global_step/sec: 1.55821 +INFO:tensorflow:step = 66201, loss = 0.121509, precision = 1.0 (64.176 sec) +INFO:tensorflow:global_step/sec: 1.55835 +INFO:tensorflow:step = 66301, loss = 0.124213, precision = 1.0 (64.170 sec) +INFO:tensorflow:global_step/sec: 1.55844 +INFO:tensorflow:step = 66401, loss = 0.121528, precision = 1.0 (64.167 sec) +Saved checkpoint after 170 epoch(s) to ../data/resnet164/checkpoints/00170... +INFO:tensorflow:global_step/sec: 1.5106 +INFO:tensorflow:step = 66501, loss = 0.134195, precision = 0.992188 (66.199 sec) +INFO:tensorflow:global_step/sec: 1.55814 +INFO:tensorflow:step = 66601, loss = 0.121741, precision = 1.0 (64.179 sec) +INFO:tensorflow:global_step/sec: 1.55857 +INFO:tensorflow:step = 66701, loss = 0.121711, precision = 1.0 (64.161 sec) +INFO:tensorflow:global_step/sec: 1.55789 +INFO:tensorflow:step = 66801, loss = 0.120696, precision = 1.0 (64.189 sec) +Saved checkpoint after 171 epoch(s) to ../data/resnet164/checkpoints/00171... +INFO:tensorflow:global_step/sec: 1.5089 +INFO:tensorflow:step = 66901, loss = 0.12114, precision = 1.0 (66.273 sec) +INFO:tensorflow:global_step/sec: 1.55866 +INFO:tensorflow:step = 67001, loss = 0.121215, precision = 1.0 (64.158 sec) +INFO:tensorflow:global_step/sec: 1.55794 +INFO:tensorflow:step = 67101, loss = 0.120626, precision = 1.0 (64.187 sec) +INFO:tensorflow:global_step/sec: 1.5576 +INFO:tensorflow:step = 67201, loss = 0.121051, precision = 1.0 (64.201 sec) +Saved checkpoint after 172 epoch(s) to ../data/resnet164/checkpoints/00172... +INFO:tensorflow:global_step/sec: 1.50313 +INFO:tensorflow:step = 67301, loss = 0.119957, precision = 1.0 (66.528 sec) +INFO:tensorflow:global_step/sec: 1.55722 +INFO:tensorflow:step = 67401, loss = 0.121706, precision = 1.0 (64.217 sec) +INFO:tensorflow:global_step/sec: 1.55788 +INFO:tensorflow:step = 67501, loss = 0.122747, precision = 1.0 (64.190 sec) +INFO:tensorflow:global_step/sec: 1.55743 +INFO:tensorflow:step = 67601, loss = 0.12165, precision = 1.0 (64.208 sec) +Saved checkpoint after 173 epoch(s) to ../data/resnet164/checkpoints/00173... +INFO:tensorflow:global_step/sec: 1.50688 +INFO:tensorflow:step = 67701, loss = 0.120317, precision = 1.0 (66.362 sec) +INFO:tensorflow:global_step/sec: 1.55737 +INFO:tensorflow:step = 67801, loss = 0.12329, precision = 1.0 (64.211 sec) +INFO:tensorflow:global_step/sec: 1.55848 +INFO:tensorflow:step = 67901, loss = 0.119497, precision = 1.0 (64.165 sec) +INFO:tensorflow:global_step/sec: 1.55721 +INFO:tensorflow:step = 68001, loss = 0.119797, precision = 1.0 (64.218 sec) +Saved checkpoint after 174 epoch(s) to ../data/resnet164/checkpoints/00174... +INFO:tensorflow:global_step/sec: 1.50984 +INFO:tensorflow:step = 68101, loss = 0.125819, precision = 1.0 (66.232 sec) +INFO:tensorflow:global_step/sec: 1.55875 +INFO:tensorflow:step = 68201, loss = 0.119156, precision = 1.0 (64.154 sec) +INFO:tensorflow:global_step/sec: 1.55726 +INFO:tensorflow:step = 68301, loss = 0.126852, precision = 1.0 (64.215 sec) +INFO:tensorflow:global_step/sec: 1.55845 +INFO:tensorflow:step = 68401, loss = 0.119103, precision = 1.0 (64.166 sec) +Saved checkpoint after 175 epoch(s) to ../data/resnet164/checkpoints/00175... +INFO:tensorflow:global_step/sec: 1.50857 +INFO:tensorflow:step = 68501, loss = 0.119444, precision = 1.0 (66.288 sec) +INFO:tensorflow:global_step/sec: 1.55828 +INFO:tensorflow:step = 68601, loss = 0.122888, precision = 1.0 (64.173 sec) +INFO:tensorflow:global_step/sec: 1.55839 +INFO:tensorflow:step = 68701, loss = 0.120613, precision = 1.0 (64.169 sec) +INFO:tensorflow:global_step/sec: 1.55801 +INFO:tensorflow:step = 68801, loss = 0.122729, precision = 1.0 (64.184 sec) +Saved checkpoint after 176 epoch(s) to ../data/resnet164/checkpoints/00176... +INFO:tensorflow:global_step/sec: 1.5105 +INFO:tensorflow:step = 68901, loss = 0.122097, precision = 1.0 (66.203 sec) +INFO:tensorflow:global_step/sec: 1.55829 +INFO:tensorflow:step = 69001, loss = 0.11865, precision = 1.0 (64.173 sec) +INFO:tensorflow:global_step/sec: 1.55761 +INFO:tensorflow:step = 69101, loss = 0.119951, precision = 1.0 (64.201 sec) +INFO:tensorflow:global_step/sec: 1.55817 +INFO:tensorflow:step = 69201, loss = 0.118393, precision = 1.0 (64.178 sec) +Saved checkpoint after 177 epoch(s) to ../data/resnet164/checkpoints/00177... +INFO:tensorflow:global_step/sec: 1.51117 +INFO:tensorflow:step = 69301, loss = 0.117825, precision = 1.0 (66.174 sec) +INFO:tensorflow:global_step/sec: 1.55679 +INFO:tensorflow:step = 69401, loss = 0.117592, precision = 1.0 (64.235 sec) +INFO:tensorflow:global_step/sec: 1.55686 +INFO:tensorflow:step = 69501, loss = 0.117705, precision = 1.0 (64.232 sec) +Saved checkpoint after 178 epoch(s) to ../data/resnet164/checkpoints/00178... +INFO:tensorflow:global_step/sec: 1.51042 +INFO:tensorflow:step = 69601, loss = 0.118509, precision = 1.0 (66.207 sec) +INFO:tensorflow:global_step/sec: 1.55714 +INFO:tensorflow:step = 69701, loss = 0.117037, precision = 1.0 (64.220 sec) +INFO:tensorflow:global_step/sec: 1.55786 +INFO:tensorflow:step = 69801, loss = 0.117412, precision = 1.0 (64.191 sec) +INFO:tensorflow:global_step/sec: 1.55621 +INFO:tensorflow:step = 69901, loss = 0.116905, precision = 1.0 (64.259 sec) +Saved checkpoint after 179 epoch(s) to ../data/resnet164/checkpoints/00179... +INFO:tensorflow:global_step/sec: 1.50899 +INFO:tensorflow:step = 70001, loss = 0.117325, precision = 1.0 (66.270 sec) +INFO:tensorflow:global_step/sec: 1.55914 +INFO:tensorflow:step = 70101, loss = 0.116723, precision = 1.0 (64.138 sec) +INFO:tensorflow:global_step/sec: 1.55843 +INFO:tensorflow:step = 70201, loss = 0.12166, precision = 1.0 (64.167 sec) +INFO:tensorflow:global_step/sec: 1.55801 +INFO:tensorflow:step = 70301, loss = 0.133914, precision = 0.992188 (64.184 sec) +Saved checkpoint after 180 epoch(s) to ../data/resnet164/checkpoints/00180... +INFO:tensorflow:global_step/sec: 1.50854 +INFO:tensorflow:step = 70401, loss = 0.1185, precision = 1.0 (66.289 sec) +INFO:tensorflow:global_step/sec: 1.55694 +INFO:tensorflow:step = 70501, loss = 0.11936, precision = 1.0 (64.228 sec) +INFO:tensorflow:global_step/sec: 1.55729 +INFO:tensorflow:step = 70601, loss = 0.116072, precision = 1.0 (64.214 sec) +INFO:tensorflow:global_step/sec: 1.55844 +INFO:tensorflow:step = 70701, loss = 0.116522, precision = 1.0 (64.167 sec) +Saved checkpoint after 181 epoch(s) to ../data/resnet164/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1k80_gc/resnet20_train.log b/tensorflow/CIFAR10/logs/1k80_gc/resnet20_train.log new file mode 100644 index 0000000..2a357ef --- /dev/null +++ b/tensorflow/CIFAR10/logs/1k80_gc/resnet20_train.log @@ -0,0 +1,1728 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/269.03k params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/10.38b flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + +======================End of Report========================== +2017-07-30 00:26:39.894512: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero +2017-07-30 00:26:39.895504: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla K80 +major: 3 minor: 7 memoryClockRate (GHz) 0.562 +pciBusID 0000:00:04.0 +Total memory: 11.17GiB +Free memory: 11.09GiB +2017-07-30 00:26:39.897975: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-07-30 00:26:39.898042: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-07-30 00:26:39.898069: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:00:04.0) +2017-07-30 00:26:40.194242: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 00:26:40.194306: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 8 visible devices +2017-07-30 00:26:40.196930: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x3ed1030 executing computations on platform Host. Devices: +2017-07-30 00:26:40.196960: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-07-30 00:26:40.197821: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 00:26:40.197853: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 8 visible devices +2017-07-30 00:26:40.198893: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x3ee3b80 executing computations on platform CUDA. Devices: +2017-07-30 00:26:40.198919: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla K80, Compute Capability 3.7 +INFO:tensorflow:step = 1, loss = 2.63582, precision = 0.101562 +INFO:tensorflow:global_step/sec: 8.8446 +INFO:tensorflow:step = 101, loss = 2.15141, precision = 0.289062 (11.307 sec) +INFO:tensorflow:global_step/sec: 9.00595 +INFO:tensorflow:step = 201, loss = 1.86549, precision = 0.46875 (11.104 sec) +INFO:tensorflow:global_step/sec: 8.99843 +INFO:tensorflow:step = 301, loss = 1.79615, precision = 0.46875 (11.113 sec) +total_params: 269034 +Saved checkpoint after 1 epoch(s) to ../data/resnet20/checkpoints/00001... +INFO:tensorflow:global_step/sec: 8.70876 +INFO:tensorflow:step = 401, loss = 2.01341, precision = 0.398438 (11.483 sec) +INFO:tensorflow:global_step/sec: 8.99258 +INFO:tensorflow:step = 501, loss = 1.75303, precision = 0.453125 (11.120 sec) +INFO:tensorflow:global_step/sec: 8.97388 +INFO:tensorflow:step = 601, loss = 1.53758, precision = 0.53125 (11.143 sec) +INFO:tensorflow:global_step/sec: 8.97736 +INFO:tensorflow:step = 701, loss = 1.61431, precision = 0.507812 (11.139 sec) +Saved checkpoint after 2 epoch(s) to ../data/resnet20/checkpoints/00002... +INFO:tensorflow:global_step/sec: 8.68123 +INFO:tensorflow:step = 801, loss = 1.2033, precision = 0.671875 (11.519 sec) +INFO:tensorflow:global_step/sec: 8.97709 +INFO:tensorflow:step = 901, loss = 1.2132, precision = 0.679688 (11.139 sec) +INFO:tensorflow:global_step/sec: 8.9909 +INFO:tensorflow:step = 1001, loss = 1.15159, precision = 0.6875 (11.122 sec) +INFO:tensorflow:global_step/sec: 8.99519 +INFO:tensorflow:step = 1101, loss = 1.07166, precision = 0.710938 (11.119 sec) +Saved checkpoint after 3 epoch(s) to ../data/resnet20/checkpoints/00003... +INFO:tensorflow:global_step/sec: 8.71748 +INFO:tensorflow:step = 1201, loss = 1.01629, precision = 0.734375 (11.469 sec) +INFO:tensorflow:global_step/sec: 8.97213 +INFO:tensorflow:step = 1301, loss = 0.893246, precision = 0.78125 (11.146 sec) +INFO:tensorflow:global_step/sec: 8.96901 +INFO:tensorflow:step = 1401, loss = 1.06488, precision = 0.742188 (11.150 sec) +INFO:tensorflow:global_step/sec: 8.97523 +INFO:tensorflow:step = 1501, loss = 0.938078, precision = 0.765625 (11.142 sec) +Saved checkpoint after 4 epoch(s) to ../data/resnet20/checkpoints/00004... +INFO:tensorflow:global_step/sec: 8.70954 +INFO:tensorflow:step = 1601, loss = 0.993085, precision = 0.757812 (11.481 sec) +INFO:tensorflow:global_step/sec: 8.98407 +INFO:tensorflow:step = 1701, loss = 0.9715, precision = 0.734375 (11.131 sec) +INFO:tensorflow:global_step/sec: 8.98202 +INFO:tensorflow:step = 1801, loss = 0.924286, precision = 0.734375 (11.133 sec) +INFO:tensorflow:global_step/sec: 8.97501 +INFO:tensorflow:step = 1901, loss = 0.992734, precision = 0.71875 (11.142 sec) +Saved checkpoint after 5 epoch(s) to ../data/resnet20/checkpoints/00005... +INFO:tensorflow:global_step/sec: 8.69636 +INFO:tensorflow:step = 2001, loss = 0.770762, precision = 0.835938 (11.499 sec) +INFO:tensorflow:global_step/sec: 8.98665 +INFO:tensorflow:step = 2101, loss = 0.85052, precision = 0.804688 (11.128 sec) +INFO:tensorflow:global_step/sec: 8.9931 +INFO:tensorflow:step = 2201, loss = 0.867085, precision = 0.789062 (11.120 sec) +INFO:tensorflow:global_step/sec: 8.9885 +INFO:tensorflow:step = 2301, loss = 0.891835, precision = 0.78125 (11.125 sec) +Saved checkpoint after 6 epoch(s) to ../data/resnet20/checkpoints/00006... +INFO:tensorflow:global_step/sec: 8.72795 +INFO:tensorflow:step = 2401, loss = 0.829984, precision = 0.78125 (11.458 sec) +INFO:tensorflow:global_step/sec: 8.99761 +INFO:tensorflow:step = 2501, loss = 0.700608, precision = 0.867188 (11.114 sec) +INFO:tensorflow:global_step/sec: 8.99041 +INFO:tensorflow:step = 2601, loss = 0.950402, precision = 0.71875 (11.123 sec) +INFO:tensorflow:global_step/sec: 9.00054 +INFO:tensorflow:step = 2701, loss = 0.876665, precision = 0.796875 (11.110 sec) +Saved checkpoint after 7 epoch(s) to ../data/resnet20/checkpoints/00007... +INFO:tensorflow:global_step/sec: 8.73072 +INFO:tensorflow:step = 2801, loss = 0.721362, precision = 0.828125 (11.454 sec) +INFO:tensorflow:global_step/sec: 8.98973 +INFO:tensorflow:step = 2901, loss = 0.871553, precision = 0.78125 (11.123 sec) +INFO:tensorflow:global_step/sec: 8.99944 +INFO:tensorflow:step = 3001, loss = 0.746899, precision = 0.804688 (11.112 sec) +INFO:tensorflow:global_step/sec: 8.98923 +INFO:tensorflow:step = 3101, loss = 0.833978, precision = 0.773438 (11.124 sec) +Saved checkpoint after 8 epoch(s) to ../data/resnet20/checkpoints/00008... +INFO:tensorflow:global_step/sec: 8.72348 +INFO:tensorflow:step = 3201, loss = 0.762071, precision = 0.835938 (11.463 sec) +INFO:tensorflow:global_step/sec: 8.99233 +INFO:tensorflow:step = 3301, loss = 0.730157, precision = 0.859375 (11.121 sec) +INFO:tensorflow:global_step/sec: 8.99235 +INFO:tensorflow:step = 3401, loss = 0.710143, precision = 0.84375 (11.121 sec) +INFO:tensorflow:global_step/sec: 8.99724 +INFO:tensorflow:step = 3501, loss = 0.853509, precision = 0.773438 (11.115 sec) +Saved checkpoint after 9 epoch(s) to ../data/resnet20/checkpoints/00009... +INFO:tensorflow:global_step/sec: 8.72411 +INFO:tensorflow:step = 3601, loss = 0.672151, precision = 0.84375 (11.463 sec) +INFO:tensorflow:global_step/sec: 9.00176 +INFO:tensorflow:step = 3701, loss = 0.735469, precision = 0.84375 (11.109 sec) +INFO:tensorflow:global_step/sec: 8.97793 +INFO:tensorflow:step = 3801, loss = 0.754355, precision = 0.835938 (11.138 sec) +INFO:tensorflow:global_step/sec: 8.98901 +INFO:tensorflow:step = 3901, loss = 0.654468, precision = 0.84375 (11.125 sec) +Saved checkpoint after 10 epoch(s) to ../data/resnet20/checkpoints/00010... +INFO:tensorflow:global_step/sec: 8.72032 +INFO:tensorflow:step = 4001, loss = 0.713344, precision = 0.835938 (11.467 sec) +INFO:tensorflow:global_step/sec: 8.9905 +INFO:tensorflow:step = 4101, loss = 0.848836, precision = 0.804688 (11.123 sec) +INFO:tensorflow:global_step/sec: 9.0041 +INFO:tensorflow:step = 4201, loss = 0.658022, precision = 0.859375 (11.106 sec) +Saved checkpoint after 11 epoch(s) to ../data/resnet20/checkpoints/00011... +INFO:tensorflow:global_step/sec: 8.73631 +INFO:tensorflow:step = 4301, loss = 0.799759, precision = 0.789062 (11.446 sec) +INFO:tensorflow:global_step/sec: 8.99747 +INFO:tensorflow:step = 4401, loss = 1.09402, precision = 0.710938 (11.114 sec) +INFO:tensorflow:global_step/sec: 8.99422 +INFO:tensorflow:step = 4501, loss = 0.720315, precision = 0.835938 (11.118 sec) +INFO:tensorflow:global_step/sec: 8.98034 +INFO:tensorflow:step = 4601, loss = 0.687527, precision = 0.859375 (11.135 sec) +Saved checkpoint after 12 epoch(s) to ../data/resnet20/checkpoints/00012... +INFO:tensorflow:global_step/sec: 8.74572 +INFO:tensorflow:step = 4701, loss = 0.936841, precision = 0.796875 (11.434 sec) +INFO:tensorflow:global_step/sec: 8.98445 +INFO:tensorflow:step = 4801, loss = 0.91013, precision = 0.75 (11.130 sec) +INFO:tensorflow:global_step/sec: 9.00109 +INFO:tensorflow:step = 4901, loss = 0.674012, precision = 0.8125 (11.110 sec) +INFO:tensorflow:global_step/sec: 8.99311 +INFO:tensorflow:step = 5001, loss = 0.690392, precision = 0.84375 (11.120 sec) +Saved checkpoint after 13 epoch(s) to ../data/resnet20/checkpoints/00013... +INFO:tensorflow:global_step/sec: 8.72814 +INFO:tensorflow:step = 5101, loss = 0.707073, precision = 0.835938 (11.457 sec) +INFO:tensorflow:global_step/sec: 9.00157 +INFO:tensorflow:step = 5201, loss = 0.786215, precision = 0.835938 (11.109 sec) +INFO:tensorflow:global_step/sec: 9.0004 +INFO:tensorflow:step = 5301, loss = 0.725833, precision = 0.859375 (11.111 sec) +INFO:tensorflow:global_step/sec: 8.98969 +INFO:tensorflow:step = 5401, loss = 0.718115, precision = 0.796875 (11.124 sec) +Saved checkpoint after 14 epoch(s) to ../data/resnet20/checkpoints/00014... +INFO:tensorflow:global_step/sec: 8.71502 +INFO:tensorflow:step = 5501, loss = 0.76627, precision = 0.835938 (11.475 sec) +INFO:tensorflow:global_step/sec: 9.00254 +INFO:tensorflow:step = 5601, loss = 0.586945, precision = 0.898438 (11.108 sec) +INFO:tensorflow:global_step/sec: 9.00148 +INFO:tensorflow:step = 5701, loss = 0.771857, precision = 0.828125 (11.109 sec) +INFO:tensorflow:global_step/sec: 8.98109 +INFO:tensorflow:step = 5801, loss = 0.679203, precision = 0.835938 (11.135 sec) +Saved checkpoint after 15 epoch(s) to ../data/resnet20/checkpoints/00015... +INFO:tensorflow:global_step/sec: 8.71541 +INFO:tensorflow:step = 5901, loss = 0.742206, precision = 0.820312 (11.474 sec) +INFO:tensorflow:global_step/sec: 8.99733 +INFO:tensorflow:step = 6001, loss = 0.642851, precision = 0.851562 (11.114 sec) +INFO:tensorflow:global_step/sec: 9.00909 +INFO:tensorflow:step = 6101, loss = 0.547753, precision = 0.90625 (11.100 sec) +INFO:tensorflow:global_step/sec: 8.99412 +INFO:tensorflow:step = 6201, loss = 0.798034, precision = 0.8125 (11.119 sec) +Saved checkpoint after 16 epoch(s) to ../data/resnet20/checkpoints/00016... +INFO:tensorflow:global_step/sec: 8.72166 +INFO:tensorflow:step = 6301, loss = 0.711752, precision = 0.859375 (11.466 sec) +INFO:tensorflow:global_step/sec: 8.96407 +INFO:tensorflow:step = 6401, loss = 0.664355, precision = 0.851562 (11.156 sec) +INFO:tensorflow:global_step/sec: 8.98405 +INFO:tensorflow:step = 6501, loss = 0.669178, precision = 0.835938 (11.131 sec) +INFO:tensorflow:global_step/sec: 8.99841 +INFO:tensorflow:step = 6601, loss = 0.640982, precision = 0.84375 (11.113 sec) +Saved checkpoint after 17 epoch(s) to ../data/resnet20/checkpoints/00017... +INFO:tensorflow:global_step/sec: 8.69762 +INFO:tensorflow:step = 6701, loss = 0.72218, precision = 0.84375 (11.497 sec) +INFO:tensorflow:global_step/sec: 8.99009 +INFO:tensorflow:step = 6801, loss = 0.791583, precision = 0.804688 (11.123 sec) +INFO:tensorflow:global_step/sec: 8.99648 +INFO:tensorflow:step = 6901, loss = 0.866411, precision = 0.773438 (11.115 sec) +INFO:tensorflow:global_step/sec: 8.99383 +INFO:tensorflow:step = 7001, loss = 0.671635, precision = 0.84375 (11.119 sec) +Saved checkpoint after 18 epoch(s) to ../data/resnet20/checkpoints/00018... +INFO:tensorflow:global_step/sec: 8.73892 +INFO:tensorflow:step = 7101, loss = 0.710769, precision = 0.84375 (11.443 sec) +INFO:tensorflow:global_step/sec: 8.98772 +INFO:tensorflow:step = 7201, loss = 0.772647, precision = 0.84375 (11.126 sec) +INFO:tensorflow:global_step/sec: 8.9901 +INFO:tensorflow:step = 7301, loss = 0.621804, precision = 0.898438 (11.123 sec) +INFO:tensorflow:global_step/sec: 8.99306 +INFO:tensorflow:step = 7401, loss = 0.702454, precision = 0.820312 (11.120 sec) +Saved checkpoint after 19 epoch(s) to ../data/resnet20/checkpoints/00019... +INFO:tensorflow:global_step/sec: 8.73853 +INFO:tensorflow:step = 7501, loss = 0.739654, precision = 0.828125 (11.444 sec) +INFO:tensorflow:global_step/sec: 8.99817 +INFO:tensorflow:step = 7601, loss = 0.826574, precision = 0.789062 (11.113 sec) +INFO:tensorflow:global_step/sec: 8.98969 +INFO:tensorflow:step = 7701, loss = 0.69398, precision = 0.828125 (11.124 sec) +INFO:tensorflow:global_step/sec: 8.99807 +INFO:tensorflow:step = 7801, loss = 0.733995, precision = 0.820312 (11.113 sec) +Saved checkpoint after 20 epoch(s) to ../data/resnet20/checkpoints/00020... +INFO:tensorflow:global_step/sec: 8.7233 +INFO:tensorflow:step = 7901, loss = 0.665967, precision = 0.875 (11.464 sec) +INFO:tensorflow:global_step/sec: 8.99586 +INFO:tensorflow:step = 8001, loss = 0.794531, precision = 0.8125 (11.116 sec) +INFO:tensorflow:global_step/sec: 8.98665 +INFO:tensorflow:step = 8101, loss = 0.714135, precision = 0.859375 (11.128 sec) +INFO:tensorflow:global_step/sec: 8.98377 +INFO:tensorflow:step = 8201, loss = 0.821927, precision = 0.796875 (11.131 sec) +Saved checkpoint after 21 epoch(s) to ../data/resnet20/checkpoints/00021... +INFO:tensorflow:global_step/sec: 8.71787 +INFO:tensorflow:step = 8301, loss = 0.779604, precision = 0.835938 (11.470 sec) +INFO:tensorflow:global_step/sec: 8.99581 +INFO:tensorflow:step = 8401, loss = 0.684935, precision = 0.835938 (11.116 sec) +INFO:tensorflow:global_step/sec: 9.00147 +INFO:tensorflow:step = 8501, loss = 0.811073, precision = 0.78125 (11.109 sec) +INFO:tensorflow:global_step/sec: 8.99552 +INFO:tensorflow:step = 8601, loss = 0.712185, precision = 0.828125 (11.117 sec) +Saved checkpoint after 22 epoch(s) to ../data/resnet20/checkpoints/00022... +INFO:tensorflow:global_step/sec: 8.73977 +INFO:tensorflow:step = 8701, loss = 0.668799, precision = 0.851562 (11.442 sec) +INFO:tensorflow:global_step/sec: 8.99953 +INFO:tensorflow:step = 8801, loss = 0.695721, precision = 0.859375 (11.112 sec) +INFO:tensorflow:global_step/sec: 9.00198 +INFO:tensorflow:step = 8901, loss = 0.69971, precision = 0.835938 (11.109 sec) +Saved checkpoint after 23 epoch(s) to ../data/resnet20/checkpoints/00023... +INFO:tensorflow:global_step/sec: 8.7334 +INFO:tensorflow:step = 9001, loss = 0.764861, precision = 0.84375 (11.450 sec) +INFO:tensorflow:global_step/sec: 8.99822 +INFO:tensorflow:step = 9101, loss = 0.787899, precision = 0.8125 (11.113 sec) +INFO:tensorflow:global_step/sec: 8.99699 +INFO:tensorflow:step = 9201, loss = 0.728426, precision = 0.835938 (11.115 sec) +INFO:tensorflow:global_step/sec: 9.00004 +INFO:tensorflow:step = 9301, loss = 0.767542, precision = 0.820312 (11.111 sec) +Saved checkpoint after 24 epoch(s) to ../data/resnet20/checkpoints/00024... +INFO:tensorflow:global_step/sec: 8.73854 +INFO:tensorflow:step = 9401, loss = 0.675936, precision = 0.859375 (11.444 sec) +INFO:tensorflow:global_step/sec: 9.01352 +INFO:tensorflow:step = 9501, loss = 0.841199, precision = 0.789062 (11.094 sec) +INFO:tensorflow:global_step/sec: 8.99588 +INFO:tensorflow:step = 9601, loss = 0.792053, precision = 0.820312 (11.116 sec) +INFO:tensorflow:global_step/sec: 8.99584 +INFO:tensorflow:step = 9701, loss = 0.715181, precision = 0.828125 (11.116 sec) +Saved checkpoint after 25 epoch(s) to ../data/resnet20/checkpoints/00025... +INFO:tensorflow:global_step/sec: 8.73206 +INFO:tensorflow:step = 9801, loss = 0.65681, precision = 0.867188 (11.452 sec) +INFO:tensorflow:global_step/sec: 9.00197 +INFO:tensorflow:step = 9901, loss = 0.655573, precision = 0.882812 (11.109 sec) +INFO:tensorflow:global_step/sec: 9.00157 +INFO:tensorflow:step = 10001, loss = 0.765476, precision = 0.835938 (11.109 sec) +INFO:tensorflow:global_step/sec: 8.99657 +INFO:tensorflow:step = 10101, loss = 0.665998, precision = 0.882812 (11.115 sec) +Saved checkpoint after 26 epoch(s) to ../data/resnet20/checkpoints/00026... +INFO:tensorflow:global_step/sec: 8.75023 +INFO:tensorflow:step = 10201, loss = 0.808922, precision = 0.8125 (11.428 sec) +INFO:tensorflow:global_step/sec: 8.99814 +INFO:tensorflow:step = 10301, loss = 0.667796, precision = 0.828125 (11.113 sec) +INFO:tensorflow:global_step/sec: 8.98726 +INFO:tensorflow:step = 10401, loss = 0.62003, precision = 0.875 (11.127 sec) +INFO:tensorflow:global_step/sec: 8.98845 +INFO:tensorflow:step = 10501, loss = 0.683628, precision = 0.867188 (11.125 sec) +Saved checkpoint after 27 epoch(s) to ../data/resnet20/checkpoints/00027... +INFO:tensorflow:global_step/sec: 8.74559 +INFO:tensorflow:step = 10601, loss = 0.632708, precision = 0.867188 (11.434 sec) +INFO:tensorflow:global_step/sec: 9.00611 +INFO:tensorflow:step = 10701, loss = 0.741137, precision = 0.835938 (11.104 sec) +INFO:tensorflow:global_step/sec: 8.99508 +INFO:tensorflow:step = 10801, loss = 0.79849, precision = 0.84375 (11.117 sec) +INFO:tensorflow:global_step/sec: 8.99941 +INFO:tensorflow:step = 10901, loss = 0.692262, precision = 0.867188 (11.112 sec) +Saved checkpoint after 28 epoch(s) to ../data/resnet20/checkpoints/00028... +INFO:tensorflow:global_step/sec: 8.75005 +INFO:tensorflow:step = 11001, loss = 0.766723, precision = 0.867188 (11.429 sec) +INFO:tensorflow:global_step/sec: 8.99397 +INFO:tensorflow:step = 11101, loss = 0.915596, precision = 0.765625 (11.118 sec) +INFO:tensorflow:global_step/sec: 8.9981 +INFO:tensorflow:step = 11201, loss = 0.615986, precision = 0.867188 (11.114 sec) +INFO:tensorflow:global_step/sec: 8.99391 +INFO:tensorflow:step = 11301, loss = 0.672063, precision = 0.867188 (11.119 sec) +Saved checkpoint after 29 epoch(s) to ../data/resnet20/checkpoints/00029... +INFO:tensorflow:global_step/sec: 8.76673 +INFO:tensorflow:step = 11401, loss = 0.65314, precision = 0.8125 (11.407 sec) +INFO:tensorflow:global_step/sec: 9.00527 +INFO:tensorflow:step = 11501, loss = 0.72263, precision = 0.84375 (11.104 sec) +INFO:tensorflow:global_step/sec: 8.99693 +INFO:tensorflow:step = 11601, loss = 0.716647, precision = 0.820312 (11.115 sec) +INFO:tensorflow:global_step/sec: 9.00382 +INFO:tensorflow:step = 11701, loss = 0.71635, precision = 0.851562 (11.107 sec) +Saved checkpoint after 30 epoch(s) to ../data/resnet20/checkpoints/00030... +INFO:tensorflow:global_step/sec: 8.75359 +INFO:tensorflow:step = 11801, loss = 0.637233, precision = 0.859375 (11.424 sec) +INFO:tensorflow:global_step/sec: 9.01448 +INFO:tensorflow:step = 11901, loss = 0.59484, precision = 0.90625 (11.093 sec) +INFO:tensorflow:global_step/sec: 8.9922 +INFO:tensorflow:step = 12001, loss = 0.595153, precision = 0.875 (11.121 sec) +INFO:tensorflow:global_step/sec: 9.00106 +INFO:tensorflow:step = 12101, loss = 0.661948, precision = 0.890625 (11.110 sec) +Saved checkpoint after 31 epoch(s) to ../data/resnet20/checkpoints/00031... +INFO:tensorflow:global_step/sec: 8.74422 +INFO:tensorflow:step = 12201, loss = 0.786163, precision = 0.851562 (11.436 sec) +INFO:tensorflow:global_step/sec: 9.00259 +INFO:tensorflow:step = 12301, loss = 0.684094, precision = 0.867188 (11.108 sec) +INFO:tensorflow:global_step/sec: 9.00107 +INFO:tensorflow:step = 12401, loss = 0.7956, precision = 0.828125 (11.110 sec) +INFO:tensorflow:global_step/sec: 8.99258 +INFO:tensorflow:step = 12501, loss = 0.685938, precision = 0.898438 (11.120 sec) +Saved checkpoint after 32 epoch(s) to ../data/resnet20/checkpoints/00032... +INFO:tensorflow:global_step/sec: 8.73482 +INFO:tensorflow:step = 12601, loss = 0.637323, precision = 0.890625 (11.449 sec) +INFO:tensorflow:global_step/sec: 9.02864 +INFO:tensorflow:step = 12701, loss = 0.642626, precision = 0.867188 (11.076 sec) +INFO:tensorflow:global_step/sec: 9.01118 +INFO:tensorflow:step = 12801, loss = 0.828107, precision = 0.804688 (11.097 sec) +INFO:tensorflow:global_step/sec: 8.99481 +INFO:tensorflow:step = 12901, loss = 0.530348, precision = 0.90625 (11.118 sec) +Saved checkpoint after 33 epoch(s) to ../data/resnet20/checkpoints/00033... +INFO:tensorflow:global_step/sec: 8.72601 +INFO:tensorflow:step = 13001, loss = 0.656595, precision = 0.859375 (11.460 sec) +INFO:tensorflow:global_step/sec: 8.99512 +INFO:tensorflow:step = 13101, loss = 0.641087, precision = 0.851562 (11.117 sec) +INFO:tensorflow:global_step/sec: 9.00147 +INFO:tensorflow:step = 13201, loss = 0.619355, precision = 0.882812 (11.110 sec) +Saved checkpoint after 34 epoch(s) to ../data/resnet20/checkpoints/00034... +INFO:tensorflow:global_step/sec: 8.77232 +INFO:tensorflow:step = 13301, loss = 0.718781, precision = 0.84375 (11.399 sec) +INFO:tensorflow:global_step/sec: 9.00548 +INFO:tensorflow:step = 13401, loss = 0.640606, precision = 0.851562 (11.104 sec) +INFO:tensorflow:global_step/sec: 9.00046 +INFO:tensorflow:step = 13501, loss = 0.807839, precision = 0.789062 (11.111 sec) +INFO:tensorflow:global_step/sec: 9.00515 +INFO:tensorflow:step = 13601, loss = 0.671979, precision = 0.851562 (11.105 sec) +Saved checkpoint after 35 epoch(s) to ../data/resnet20/checkpoints/00035... +INFO:tensorflow:global_step/sec: 8.71073 +INFO:tensorflow:step = 13701, loss = 0.586171, precision = 0.90625 (11.480 sec) +INFO:tensorflow:global_step/sec: 9.00816 +INFO:tensorflow:step = 13801, loss = 0.669077, precision = 0.835938 (11.101 sec) +INFO:tensorflow:global_step/sec: 9.00537 +INFO:tensorflow:step = 13901, loss = 0.806787, precision = 0.835938 (11.104 sec) +INFO:tensorflow:global_step/sec: 8.99423 +INFO:tensorflow:step = 14001, loss = 0.618961, precision = 0.890625 (11.118 sec) +Saved checkpoint after 36 epoch(s) to ../data/resnet20/checkpoints/00036... +INFO:tensorflow:global_step/sec: 8.73782 +INFO:tensorflow:step = 14101, loss = 0.671448, precision = 0.84375 (11.445 sec) +INFO:tensorflow:global_step/sec: 9.00502 +INFO:tensorflow:step = 14201, loss = 0.619879, precision = 0.890625 (11.105 sec) +INFO:tensorflow:global_step/sec: 8.98541 +INFO:tensorflow:step = 14301, loss = 0.641404, precision = 0.867188 (11.129 sec) +INFO:tensorflow:global_step/sec: 9.01076 +INFO:tensorflow:step = 14401, loss = 0.629924, precision = 0.859375 (11.098 sec) +Saved checkpoint after 37 epoch(s) to ../data/resnet20/checkpoints/00037... +INFO:tensorflow:global_step/sec: 8.72898 +INFO:tensorflow:step = 14501, loss = 0.678757, precision = 0.875 (11.456 sec) +INFO:tensorflow:global_step/sec: 8.99493 +INFO:tensorflow:step = 14601, loss = 0.711473, precision = 0.84375 (11.117 sec) +INFO:tensorflow:global_step/sec: 9.00684 +INFO:tensorflow:step = 14701, loss = 0.625429, precision = 0.882812 (11.103 sec) +INFO:tensorflow:global_step/sec: 8.99567 +INFO:tensorflow:step = 14801, loss = 0.660572, precision = 0.851562 (11.116 sec) +Saved checkpoint after 38 epoch(s) to ../data/resnet20/checkpoints/00038... +INFO:tensorflow:global_step/sec: 8.7549 +INFO:tensorflow:step = 14901, loss = 0.569198, precision = 0.90625 (11.422 sec) +INFO:tensorflow:global_step/sec: 9.00659 +INFO:tensorflow:step = 15001, loss = 0.644153, precision = 0.875 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01463 +INFO:tensorflow:step = 15101, loss = 0.737217, precision = 0.835938 (11.093 sec) +INFO:tensorflow:global_step/sec: 8.99804 +INFO:tensorflow:step = 15201, loss = 0.728753, precision = 0.84375 (11.114 sec) +Saved checkpoint after 39 epoch(s) to ../data/resnet20/checkpoints/00039... +INFO:tensorflow:global_step/sec: 8.76573 +INFO:tensorflow:step = 15301, loss = 0.681902, precision = 0.859375 (11.408 sec) +INFO:tensorflow:global_step/sec: 9.039 +INFO:tensorflow:step = 15401, loss = 0.500901, precision = 0.914062 (11.063 sec) +INFO:tensorflow:global_step/sec: 9.01579 +INFO:tensorflow:step = 15501, loss = 0.666093, precision = 0.859375 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.01199 +INFO:tensorflow:step = 15601, loss = 0.701893, precision = 0.867188 (11.096 sec) +Saved checkpoint after 40 epoch(s) to ../data/resnet20/checkpoints/00040... +INFO:tensorflow:global_step/sec: 8.72346 +INFO:tensorflow:step = 15701, loss = 0.689023, precision = 0.851562 (11.463 sec) +INFO:tensorflow:global_step/sec: 9.00843 +INFO:tensorflow:step = 15801, loss = 0.6543, precision = 0.890625 (11.101 sec) +INFO:tensorflow:global_step/sec: 9.01518 +INFO:tensorflow:step = 15901, loss = 0.78795, precision = 0.851562 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.0082 +INFO:tensorflow:step = 16001, loss = 0.867016, precision = 0.78125 (11.101 sec) +Saved checkpoint after 41 epoch(s) to ../data/resnet20/checkpoints/00041... +INFO:tensorflow:global_step/sec: 8.74538 +INFO:tensorflow:step = 16101, loss = 0.605179, precision = 0.898438 (11.435 sec) +INFO:tensorflow:global_step/sec: 9.00319 +INFO:tensorflow:step = 16201, loss = 0.618825, precision = 0.882812 (11.107 sec) +INFO:tensorflow:global_step/sec: 9.01127 +INFO:tensorflow:step = 16301, loss = 0.593303, precision = 0.898438 (11.097 sec) +INFO:tensorflow:global_step/sec: 8.99537 +INFO:tensorflow:step = 16401, loss = 0.667735, precision = 0.851562 (11.117 sec) +Saved checkpoint after 42 epoch(s) to ../data/resnet20/checkpoints/00042... +INFO:tensorflow:global_step/sec: 8.74156 +INFO:tensorflow:step = 16501, loss = 0.695862, precision = 0.851562 (11.440 sec) +INFO:tensorflow:global_step/sec: 8.99094 +INFO:tensorflow:step = 16601, loss = 0.638021, precision = 0.84375 (11.122 sec) +INFO:tensorflow:global_step/sec: 9.00501 +INFO:tensorflow:step = 16701, loss = 0.665109, precision = 0.882812 (11.105 sec) +INFO:tensorflow:global_step/sec: 9.0075 +INFO:tensorflow:step = 16801, loss = 0.650907, precision = 0.875 (11.102 sec) +Saved checkpoint after 43 epoch(s) to ../data/resnet20/checkpoints/00043... +INFO:tensorflow:global_step/sec: 8.73269 +INFO:tensorflow:step = 16901, loss = 0.641212, precision = 0.867188 (11.451 sec) +INFO:tensorflow:global_step/sec: 8.9999 +INFO:tensorflow:step = 17001, loss = 0.663768, precision = 0.890625 (11.111 sec) +INFO:tensorflow:global_step/sec: 9.00885 +INFO:tensorflow:step = 17101, loss = 0.684203, precision = 0.828125 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.01094 +INFO:tensorflow:step = 17201, loss = 0.5677, precision = 0.90625 (11.098 sec) +Saved checkpoint after 44 epoch(s) to ../data/resnet20/checkpoints/00044... +INFO:tensorflow:global_step/sec: 8.74328 +INFO:tensorflow:step = 17301, loss = 0.475229, precision = 0.945312 (11.437 sec) +INFO:tensorflow:global_step/sec: 9.00234 +INFO:tensorflow:step = 17401, loss = 0.683061, precision = 0.851562 (11.108 sec) +INFO:tensorflow:global_step/sec: 9.00536 +INFO:tensorflow:step = 17501, loss = 0.620372, precision = 0.90625 (11.104 sec) +Saved checkpoint after 45 epoch(s) to ../data/resnet20/checkpoints/00045... +INFO:tensorflow:global_step/sec: 8.74862 +INFO:tensorflow:step = 17601, loss = 0.631014, precision = 0.867188 (11.430 sec) +INFO:tensorflow:global_step/sec: 9.01432 +INFO:tensorflow:step = 17701, loss = 0.605007, precision = 0.90625 (11.093 sec) +INFO:tensorflow:global_step/sec: 8.99335 +INFO:tensorflow:step = 17801, loss = 0.6528, precision = 0.867188 (11.119 sec) +INFO:tensorflow:global_step/sec: 8.99958 +INFO:tensorflow:step = 17901, loss = 0.599646, precision = 0.890625 (11.112 sec) +Saved checkpoint after 46 epoch(s) to ../data/resnet20/checkpoints/00046... +INFO:tensorflow:global_step/sec: 8.74775 +INFO:tensorflow:step = 18001, loss = 0.689335, precision = 0.859375 (11.432 sec) +INFO:tensorflow:global_step/sec: 9.00569 +INFO:tensorflow:step = 18101, loss = 0.74802, precision = 0.804688 (11.104 sec) +INFO:tensorflow:global_step/sec: 9.01665 +INFO:tensorflow:step = 18201, loss = 0.79174, precision = 0.851562 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.00509 +INFO:tensorflow:step = 18301, loss = 0.697989, precision = 0.828125 (11.105 sec) +Saved checkpoint after 47 epoch(s) to ../data/resnet20/checkpoints/00047... +INFO:tensorflow:global_step/sec: 8.7495 +INFO:tensorflow:step = 18401, loss = 0.605176, precision = 0.921875 (11.429 sec) +INFO:tensorflow:global_step/sec: 9.00263 +INFO:tensorflow:step = 18501, loss = 0.767793, precision = 0.828125 (11.108 sec) +INFO:tensorflow:global_step/sec: 9.00364 +INFO:tensorflow:step = 18601, loss = 0.734758, precision = 0.867188 (11.107 sec) +INFO:tensorflow:global_step/sec: 9.00611 +INFO:tensorflow:step = 18701, loss = 0.59734, precision = 0.890625 (11.104 sec) +Saved checkpoint after 48 epoch(s) to ../data/resnet20/checkpoints/00048... +INFO:tensorflow:global_step/sec: 8.76041 +INFO:tensorflow:step = 18801, loss = 0.653543, precision = 0.859375 (11.415 sec) +INFO:tensorflow:global_step/sec: 9.01644 +INFO:tensorflow:step = 18901, loss = 0.711951, precision = 0.820312 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.00403 +INFO:tensorflow:step = 19001, loss = 0.63145, precision = 0.882812 (11.106 sec) +INFO:tensorflow:global_step/sec: 8.99541 +INFO:tensorflow:step = 19101, loss = 0.734876, precision = 0.859375 (11.117 sec) +Saved checkpoint after 49 epoch(s) to ../data/resnet20/checkpoints/00049... +INFO:tensorflow:global_step/sec: 8.74379 +INFO:tensorflow:step = 19201, loss = 0.701413, precision = 0.835938 (11.437 sec) +INFO:tensorflow:global_step/sec: 9.01457 +INFO:tensorflow:step = 19301, loss = 0.621667, precision = 0.882812 (11.093 sec) +INFO:tensorflow:global_step/sec: 9.01251 +INFO:tensorflow:step = 19401, loss = 0.561399, precision = 0.9375 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.00898 +INFO:tensorflow:step = 19501, loss = 0.654718, precision = 0.875 (11.100 sec) +Saved checkpoint after 50 epoch(s) to ../data/resnet20/checkpoints/00050... +INFO:tensorflow:global_step/sec: 8.74425 +INFO:tensorflow:step = 19601, loss = 0.75043, precision = 0.835938 (11.436 sec) +INFO:tensorflow:global_step/sec: 8.99906 +INFO:tensorflow:step = 19701, loss = 0.621037, precision = 0.890625 (11.112 sec) +INFO:tensorflow:global_step/sec: 9.0125 +INFO:tensorflow:step = 19801, loss = 0.599109, precision = 0.875 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.00679 +INFO:tensorflow:step = 19901, loss = 0.655139, precision = 0.875 (11.103 sec) +Saved checkpoint after 51 epoch(s) to ../data/resnet20/checkpoints/00051... +INFO:tensorflow:global_step/sec: 8.73547 +INFO:tensorflow:step = 20001, loss = 0.618254, precision = 0.882812 (11.448 sec) +INFO:tensorflow:global_step/sec: 9.00604 +INFO:tensorflow:step = 20101, loss = 0.826784, precision = 0.820312 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01481 +INFO:tensorflow:step = 20201, loss = 0.734542, precision = 0.875 (11.093 sec) +INFO:tensorflow:global_step/sec: 9.01231 +INFO:tensorflow:step = 20301, loss = 0.552979, precision = 0.921875 (11.096 sec) +Saved checkpoint after 52 epoch(s) to ../data/resnet20/checkpoints/00052... +INFO:tensorflow:global_step/sec: 8.7503 +INFO:tensorflow:step = 20401, loss = 0.660353, precision = 0.890625 (11.428 sec) +INFO:tensorflow:global_step/sec: 9.00896 +INFO:tensorflow:step = 20501, loss = 0.567329, precision = 0.882812 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.01886 +INFO:tensorflow:step = 20601, loss = 0.662227, precision = 0.859375 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.01059 +INFO:tensorflow:step = 20701, loss = 0.652509, precision = 0.882812 (11.098 sec) +Saved checkpoint after 53 epoch(s) to ../data/resnet20/checkpoints/00053... +INFO:tensorflow:global_step/sec: 8.77358 +INFO:tensorflow:step = 20801, loss = 0.592862, precision = 0.882812 (11.398 sec) +INFO:tensorflow:global_step/sec: 9.00592 +INFO:tensorflow:step = 20901, loss = 0.62968, precision = 0.859375 (11.104 sec) +INFO:tensorflow:global_step/sec: 9 +INFO:tensorflow:step = 21001, loss = 0.610757, precision = 0.875 (11.111 sec) +INFO:tensorflow:global_step/sec: 9.02326 +INFO:tensorflow:step = 21101, loss = 0.767255, precision = 0.835938 (11.082 sec) +Saved checkpoint after 54 epoch(s) to ../data/resnet20/checkpoints/00054... +INFO:tensorflow:global_step/sec: 8.74325 +INFO:tensorflow:step = 21201, loss = 0.651424, precision = 0.859375 (11.438 sec) +INFO:tensorflow:global_step/sec: 9.00761 +INFO:tensorflow:step = 21301, loss = 0.672817, precision = 0.859375 (11.102 sec) +INFO:tensorflow:global_step/sec: 9.00936 +INFO:tensorflow:step = 21401, loss = 0.621381, precision = 0.875 (11.100 sec) +INFO:tensorflow:global_step/sec: 8.99555 +INFO:tensorflow:step = 21501, loss = 0.721021, precision = 0.859375 (11.116 sec) +Saved checkpoint after 55 epoch(s) to ../data/resnet20/checkpoints/00055... +INFO:tensorflow:global_step/sec: 8.74822 +INFO:tensorflow:step = 21601, loss = 0.744732, precision = 0.835938 (11.431 sec) +INFO:tensorflow:global_step/sec: 9.00524 +INFO:tensorflow:step = 21701, loss = 0.685535, precision = 0.859375 (11.104 sec) +INFO:tensorflow:global_step/sec: 9.0041 +INFO:tensorflow:step = 21801, loss = 0.560826, precision = 0.90625 (11.106 sec) +Saved checkpoint after 56 epoch(s) to ../data/resnet20/checkpoints/00056... +INFO:tensorflow:global_step/sec: 8.7538 +INFO:tensorflow:step = 21901, loss = 0.525338, precision = 0.921875 (11.424 sec) +INFO:tensorflow:global_step/sec: 9.00704 +INFO:tensorflow:step = 22001, loss = 0.699139, precision = 0.882812 (11.102 sec) +INFO:tensorflow:global_step/sec: 9.00571 +INFO:tensorflow:step = 22101, loss = 0.577578, precision = 0.890625 (11.104 sec) +INFO:tensorflow:global_step/sec: 8.99362 +INFO:tensorflow:step = 22201, loss = 0.621527, precision = 0.890625 (11.119 sec) +Saved checkpoint after 57 epoch(s) to ../data/resnet20/checkpoints/00057... +INFO:tensorflow:global_step/sec: 8.74537 +INFO:tensorflow:step = 22301, loss = 0.662852, precision = 0.882812 (11.435 sec) +INFO:tensorflow:global_step/sec: 9.00934 +INFO:tensorflow:step = 22401, loss = 0.802737, precision = 0.8125 (11.100 sec) +INFO:tensorflow:global_step/sec: 8.99386 +INFO:tensorflow:step = 22501, loss = 0.686625, precision = 0.875 (11.118 sec) +INFO:tensorflow:global_step/sec: 9.00715 +INFO:tensorflow:step = 22601, loss = 0.632806, precision = 0.84375 (11.102 sec) +Saved checkpoint after 58 epoch(s) to ../data/resnet20/checkpoints/00058... +INFO:tensorflow:global_step/sec: 8.71344 +INFO:tensorflow:step = 22701, loss = 0.637604, precision = 0.90625 (11.477 sec) +INFO:tensorflow:global_step/sec: 9.005 +INFO:tensorflow:step = 22801, loss = 0.529126, precision = 0.9375 (11.105 sec) +INFO:tensorflow:global_step/sec: 9.00327 +INFO:tensorflow:step = 22901, loss = 0.610795, precision = 0.867188 (11.107 sec) +INFO:tensorflow:global_step/sec: 9.00254 +INFO:tensorflow:step = 23001, loss = 0.693876, precision = 0.867188 (11.108 sec) +Saved checkpoint after 59 epoch(s) to ../data/resnet20/checkpoints/00059... +INFO:tensorflow:global_step/sec: 8.75386 +INFO:tensorflow:step = 23101, loss = 0.719762, precision = 0.851562 (11.424 sec) +INFO:tensorflow:global_step/sec: 9.00169 +INFO:tensorflow:step = 23201, loss = 0.850263, precision = 0.828125 (11.109 sec) +INFO:tensorflow:global_step/sec: 9.01184 +INFO:tensorflow:step = 23301, loss = 0.765196, precision = 0.796875 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.0018 +INFO:tensorflow:step = 23401, loss = 0.572434, precision = 0.875 (11.109 sec) +Saved checkpoint after 60 epoch(s) to ../data/resnet20/checkpoints/00060... +INFO:tensorflow:global_step/sec: 8.76729 +INFO:tensorflow:step = 23501, loss = 0.616719, precision = 0.875 (11.406 sec) +INFO:tensorflow:global_step/sec: 9.00958 +INFO:tensorflow:step = 23601, loss = 0.625643, precision = 0.898438 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.00417 +INFO:tensorflow:step = 23701, loss = 0.70243, precision = 0.890625 (11.106 sec) +INFO:tensorflow:global_step/sec: 8.9954 +INFO:tensorflow:step = 23801, loss = 0.807366, precision = 0.828125 (11.117 sec) +Saved checkpoint after 61 epoch(s) to ../data/resnet20/checkpoints/00061... +INFO:tensorflow:global_step/sec: 8.73632 +INFO:tensorflow:step = 23901, loss = 0.568691, precision = 0.890625 (11.447 sec) +INFO:tensorflow:global_step/sec: 9.01475 +INFO:tensorflow:step = 24001, loss = 0.553465, precision = 0.90625 (11.093 sec) +INFO:tensorflow:global_step/sec: 9.01654 +INFO:tensorflow:step = 24101, loss = 0.622372, precision = 0.90625 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.00425 +INFO:tensorflow:step = 24201, loss = 0.730691, precision = 0.859375 (11.106 sec) +Saved checkpoint after 62 epoch(s) to ../data/resnet20/checkpoints/00062... +INFO:tensorflow:global_step/sec: 8.75488 +INFO:tensorflow:step = 24301, loss = 0.719186, precision = 0.851562 (11.422 sec) +INFO:tensorflow:global_step/sec: 8.99936 +INFO:tensorflow:step = 24401, loss = 0.648375, precision = 0.875 (11.112 sec) +INFO:tensorflow:global_step/sec: 9.021 +INFO:tensorflow:step = 24501, loss = 0.652865, precision = 0.875 (11.085 sec) +INFO:tensorflow:global_step/sec: 9.00876 +INFO:tensorflow:step = 24601, loss = 0.651849, precision = 0.851562 (11.100 sec) +Saved checkpoint after 63 epoch(s) to ../data/resnet20/checkpoints/00063... +INFO:tensorflow:global_step/sec: 8.7414 +INFO:tensorflow:step = 24701, loss = 0.740944, precision = 0.875 (11.440 sec) +INFO:tensorflow:global_step/sec: 8.9998 +INFO:tensorflow:step = 24801, loss = 0.718432, precision = 0.851562 (11.111 sec) +INFO:tensorflow:global_step/sec: 9.00668 +INFO:tensorflow:step = 24901, loss = 0.694204, precision = 0.820312 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.00694 +INFO:tensorflow:step = 25001, loss = 0.622659, precision = 0.890625 (11.103 sec) +Saved checkpoint after 64 epoch(s) to ../data/resnet20/checkpoints/00064... +INFO:tensorflow:global_step/sec: 8.7463 +INFO:tensorflow:step = 25101, loss = 0.633964, precision = 0.882812 (11.433 sec) +INFO:tensorflow:global_step/sec: 9.01242 +INFO:tensorflow:step = 25201, loss = 0.769593, precision = 0.78125 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.00273 +INFO:tensorflow:step = 25301, loss = 0.581373, precision = 0.90625 (11.108 sec) +INFO:tensorflow:global_step/sec: 9.01037 +INFO:tensorflow:step = 25401, loss = 0.73258, precision = 0.875 (11.098 sec) +Saved checkpoint after 65 epoch(s) to ../data/resnet20/checkpoints/00065... +INFO:tensorflow:global_step/sec: 8.745 +INFO:tensorflow:step = 25501, loss = 0.679221, precision = 0.867188 (11.435 sec) +INFO:tensorflow:global_step/sec: 9.00851 +INFO:tensorflow:step = 25601, loss = 0.61362, precision = 0.90625 (11.101 sec) +INFO:tensorflow:global_step/sec: 9.01862 +INFO:tensorflow:step = 25701, loss = 0.647252, precision = 0.867188 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.00972 +INFO:tensorflow:step = 25801, loss = 0.741438, precision = 0.851562 (11.099 sec) +Saved checkpoint after 66 epoch(s) to ../data/resnet20/checkpoints/00066... +INFO:tensorflow:global_step/sec: 8.75153 +INFO:tensorflow:step = 25901, loss = 0.782403, precision = 0.835938 (11.427 sec) +INFO:tensorflow:global_step/sec: 9.00198 +INFO:tensorflow:step = 26001, loss = 0.768763, precision = 0.84375 (11.109 sec) +INFO:tensorflow:global_step/sec: 9.00697 +INFO:tensorflow:step = 26101, loss = 0.681015, precision = 0.867188 (11.103 sec) +Saved checkpoint after 67 epoch(s) to ../data/resnet20/checkpoints/00067... +INFO:tensorflow:global_step/sec: 8.75302 +INFO:tensorflow:step = 26201, loss = 0.742703, precision = 0.867188 (11.425 sec) +INFO:tensorflow:global_step/sec: 8.99485 +INFO:tensorflow:step = 26301, loss = 0.686986, precision = 0.859375 (11.117 sec) +INFO:tensorflow:global_step/sec: 9.02394 +INFO:tensorflow:step = 26401, loss = 0.682185, precision = 0.84375 (11.082 sec) +INFO:tensorflow:global_step/sec: 9.01243 +INFO:tensorflow:step = 26501, loss = 0.655739, precision = 0.851562 (11.095 sec) +Saved checkpoint after 68 epoch(s) to ../data/resnet20/checkpoints/00068... +INFO:tensorflow:global_step/sec: 8.77312 +INFO:tensorflow:step = 26601, loss = 0.592039, precision = 0.890625 (11.399 sec) +INFO:tensorflow:global_step/sec: 9.00391 +INFO:tensorflow:step = 26701, loss = 0.6023, precision = 0.875 (11.106 sec) +INFO:tensorflow:global_step/sec: 9.0085 +INFO:tensorflow:step = 26801, loss = 0.591694, precision = 0.890625 (11.101 sec) +INFO:tensorflow:global_step/sec: 8.99811 +INFO:tensorflow:step = 26901, loss = 0.69348, precision = 0.875 (11.113 sec) +Saved checkpoint after 69 epoch(s) to ../data/resnet20/checkpoints/00069... +INFO:tensorflow:global_step/sec: 8.75294 +INFO:tensorflow:step = 27001, loss = 0.64141, precision = 0.875 (11.425 sec) +INFO:tensorflow:global_step/sec: 9.02206 +INFO:tensorflow:step = 27101, loss = 0.579568, precision = 0.875 (11.084 sec) +INFO:tensorflow:global_step/sec: 9.01733 +INFO:tensorflow:step = 27201, loss = 0.506963, precision = 0.914062 (11.090 sec) +INFO:tensorflow:global_step/sec: 9.01184 +INFO:tensorflow:step = 27301, loss = 0.822248, precision = 0.851562 (11.096 sec) +Saved checkpoint after 70 epoch(s) to ../data/resnet20/checkpoints/00070... +INFO:tensorflow:global_step/sec: 8.77391 +INFO:tensorflow:step = 27401, loss = 0.639129, precision = 0.875 (11.397 sec) +INFO:tensorflow:global_step/sec: 9.01003 +INFO:tensorflow:step = 27501, loss = 0.681452, precision = 0.875 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.00966 +INFO:tensorflow:step = 27601, loss = 0.668974, precision = 0.867188 (11.099 sec) +INFO:tensorflow:global_step/sec: 8.98657 +INFO:tensorflow:step = 27701, loss = 0.609514, precision = 0.851562 (11.128 sec) +Saved checkpoint after 71 epoch(s) to ../data/resnet20/checkpoints/00071... +INFO:tensorflow:global_step/sec: 8.76443 +INFO:tensorflow:step = 27801, loss = 0.61849, precision = 0.867188 (11.410 sec) +INFO:tensorflow:global_step/sec: 9.01253 +INFO:tensorflow:step = 27901, loss = 0.656372, precision = 0.859375 (11.096 sec) +INFO:tensorflow:global_step/sec: 8.99695 +INFO:tensorflow:step = 28001, loss = 0.794817, precision = 0.835938 (11.115 sec) +INFO:tensorflow:global_step/sec: 9.01469 +INFO:tensorflow:step = 28101, loss = 0.773907, precision = 0.835938 (11.093 sec) +Saved checkpoint after 72 epoch(s) to ../data/resnet20/checkpoints/00072... +INFO:tensorflow:global_step/sec: 8.76268 +INFO:tensorflow:step = 28201, loss = 0.589017, precision = 0.890625 (11.412 sec) +INFO:tensorflow:global_step/sec: 9.01768 +INFO:tensorflow:step = 28301, loss = 0.64568, precision = 0.859375 (11.089 sec) +INFO:tensorflow:global_step/sec: 9.01578 +INFO:tensorflow:step = 28401, loss = 0.663255, precision = 0.882812 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.02018 +INFO:tensorflow:step = 28501, loss = 0.630632, precision = 0.875 (11.086 sec) +Saved checkpoint after 73 epoch(s) to ../data/resnet20/checkpoints/00073... +INFO:tensorflow:global_step/sec: 8.76119 +INFO:tensorflow:step = 28601, loss = 0.599779, precision = 0.875 (11.414 sec) +INFO:tensorflow:global_step/sec: 9.0011 +INFO:tensorflow:step = 28701, loss = 0.629374, precision = 0.867188 (11.109 sec) +INFO:tensorflow:global_step/sec: 9.01572 +INFO:tensorflow:step = 28801, loss = 1.01234, precision = 0.78125 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.01728 +INFO:tensorflow:step = 28901, loss = 0.569924, precision = 0.90625 (11.090 sec) +Saved checkpoint after 74 epoch(s) to ../data/resnet20/checkpoints/00074... +INFO:tensorflow:global_step/sec: 8.74934 +INFO:tensorflow:step = 29001, loss = 0.714978, precision = 0.851562 (11.430 sec) +INFO:tensorflow:global_step/sec: 9.00358 +INFO:tensorflow:step = 29101, loss = 0.695946, precision = 0.875 (11.107 sec) +INFO:tensorflow:global_step/sec: 9.00258 +INFO:tensorflow:step = 29201, loss = 0.577124, precision = 0.898438 (11.108 sec) +INFO:tensorflow:global_step/sec: 9.00121 +INFO:tensorflow:step = 29301, loss = 0.696459, precision = 0.820312 (11.110 sec) +Saved checkpoint after 75 epoch(s) to ../data/resnet20/checkpoints/00075... +INFO:tensorflow:global_step/sec: 8.75396 +INFO:tensorflow:step = 29401, loss = 0.608321, precision = 0.875 (11.424 sec) +INFO:tensorflow:global_step/sec: 9.01241 +INFO:tensorflow:step = 29501, loss = 0.663575, precision = 0.851562 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.00673 +INFO:tensorflow:step = 29601, loss = 0.673633, precision = 0.851562 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01056 +INFO:tensorflow:step = 29701, loss = 0.723981, precision = 0.84375 (11.098 sec) +Saved checkpoint after 76 epoch(s) to ../data/resnet20/checkpoints/00076... +INFO:tensorflow:global_step/sec: 8.73869 +INFO:tensorflow:step = 29801, loss = 0.700332, precision = 0.851562 (11.443 sec) +INFO:tensorflow:global_step/sec: 9.0071 +INFO:tensorflow:step = 29901, loss = 0.713434, precision = 0.859375 (11.102 sec) +INFO:tensorflow:global_step/sec: 9.0188 +INFO:tensorflow:step = 30001, loss = 0.779215, precision = 0.820312 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.01406 +INFO:tensorflow:step = 30101, loss = 0.600551, precision = 0.882812 (11.094 sec) +Saved checkpoint after 77 epoch(s) to ../data/resnet20/checkpoints/00077... +INFO:tensorflow:global_step/sec: 8.72576 +INFO:tensorflow:step = 30201, loss = 0.537466, precision = 0.914062 (11.460 sec) +INFO:tensorflow:global_step/sec: 9.01026 +INFO:tensorflow:step = 30301, loss = 0.592956, precision = 0.90625 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.01684 +INFO:tensorflow:step = 30401, loss = 0.656627, precision = 0.875 (11.090 sec) +Saved checkpoint after 78 epoch(s) to ../data/resnet20/checkpoints/00078... +INFO:tensorflow:global_step/sec: 8.75763 +INFO:tensorflow:step = 30501, loss = 0.640646, precision = 0.835938 (11.419 sec) +INFO:tensorflow:global_step/sec: 9.01238 +INFO:tensorflow:step = 30601, loss = 0.688703, precision = 0.859375 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.00694 +INFO:tensorflow:step = 30701, loss = 0.607189, precision = 0.875 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01922 +INFO:tensorflow:step = 30801, loss = 0.610779, precision = 0.890625 (11.087 sec) +Saved checkpoint after 79 epoch(s) to ../data/resnet20/checkpoints/00079... +INFO:tensorflow:global_step/sec: 8.74963 +INFO:tensorflow:step = 30901, loss = 0.797176, precision = 0.820312 (11.429 sec) +INFO:tensorflow:global_step/sec: 9.02498 +INFO:tensorflow:step = 31001, loss = 0.563304, precision = 0.890625 (11.080 sec) +INFO:tensorflow:global_step/sec: 9.02467 +INFO:tensorflow:step = 31101, loss = 0.636114, precision = 0.859375 (11.081 sec) +INFO:tensorflow:global_step/sec: 9.01681 +INFO:tensorflow:step = 31201, loss = 0.638349, precision = 0.898438 (11.090 sec) +Saved checkpoint after 80 epoch(s) to ../data/resnet20/checkpoints/00080... +INFO:tensorflow:global_step/sec: 8.74304 +INFO:tensorflow:step = 31301, loss = 0.629247, precision = 0.90625 (11.438 sec) +INFO:tensorflow:global_step/sec: 9.0154 +INFO:tensorflow:step = 31401, loss = 0.718365, precision = 0.859375 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.02069 +INFO:tensorflow:step = 31501, loss = 0.604595, precision = 0.875 (11.086 sec) +INFO:tensorflow:global_step/sec: 9.0035 +INFO:tensorflow:step = 31601, loss = 0.716713, precision = 0.859375 (11.107 sec) +Saved checkpoint after 81 epoch(s) to ../data/resnet20/checkpoints/00081... +INFO:tensorflow:global_step/sec: 8.75111 +INFO:tensorflow:step = 31701, loss = 0.670176, precision = 0.890625 (11.427 sec) +INFO:tensorflow:global_step/sec: 9.02044 +INFO:tensorflow:step = 31801, loss = 0.566608, precision = 0.859375 (11.086 sec) +INFO:tensorflow:global_step/sec: 9.00781 +INFO:tensorflow:step = 31901, loss = 0.560194, precision = 0.914062 (11.102 sec) +INFO:tensorflow:global_step/sec: 9.01339 +INFO:tensorflow:step = 32001, loss = 0.780303, precision = 0.828125 (11.094 sec) +Saved checkpoint after 82 epoch(s) to ../data/resnet20/checkpoints/00082... +INFO:tensorflow:global_step/sec: 8.76617 +INFO:tensorflow:step = 32101, loss = 0.616743, precision = 0.859375 (11.408 sec) +INFO:tensorflow:global_step/sec: 9.02356 +INFO:tensorflow:step = 32201, loss = 0.58518, precision = 0.890625 (11.082 sec) +INFO:tensorflow:global_step/sec: 9.00954 +INFO:tensorflow:step = 32301, loss = 0.552516, precision = 0.914062 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.00689 +INFO:tensorflow:step = 32401, loss = 0.695361, precision = 0.851562 (11.103 sec) +Saved checkpoint after 83 epoch(s) to ../data/resnet20/checkpoints/00083... +INFO:tensorflow:global_step/sec: 8.77534 +INFO:tensorflow:step = 32501, loss = 0.600469, precision = 0.898438 (11.396 sec) +INFO:tensorflow:global_step/sec: 9.01331 +INFO:tensorflow:step = 32601, loss = 0.761679, precision = 0.820312 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.00289 +INFO:tensorflow:step = 32701, loss = 0.549471, precision = 0.90625 (11.107 sec) +INFO:tensorflow:global_step/sec: 9.01002 +INFO:tensorflow:step = 32801, loss = 0.502498, precision = 0.898438 (11.099 sec) +Saved checkpoint after 84 epoch(s) to ../data/resnet20/checkpoints/00084... +INFO:tensorflow:global_step/sec: 8.75687 +INFO:tensorflow:step = 32901, loss = 0.506382, precision = 0.921875 (11.419 sec) +INFO:tensorflow:global_step/sec: 9.00823 +INFO:tensorflow:step = 33001, loss = 0.716862, precision = 0.84375 (11.101 sec) +INFO:tensorflow:global_step/sec: 9.00683 +INFO:tensorflow:step = 33101, loss = 0.587327, precision = 0.898438 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.00953 +INFO:tensorflow:step = 33201, loss = 0.598701, precision = 0.898438 (11.099 sec) +Saved checkpoint after 85 epoch(s) to ../data/resnet20/checkpoints/00085... +INFO:tensorflow:global_step/sec: 8.73145 +INFO:tensorflow:step = 33301, loss = 0.545781, precision = 0.882812 (11.453 sec) +INFO:tensorflow:global_step/sec: 9.01882 +INFO:tensorflow:step = 33401, loss = 0.749463, precision = 0.84375 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.01913 +INFO:tensorflow:step = 33501, loss = 0.587927, precision = 0.898438 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.01492 +INFO:tensorflow:step = 33601, loss = 0.602201, precision = 0.875 (11.093 sec) +Saved checkpoint after 86 epoch(s) to ../data/resnet20/checkpoints/00086... +INFO:tensorflow:global_step/sec: 8.74322 +INFO:tensorflow:step = 33701, loss = 0.634178, precision = 0.882812 (11.438 sec) +INFO:tensorflow:global_step/sec: 9.01664 +INFO:tensorflow:step = 33801, loss = 0.668187, precision = 0.851562 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.00792 +INFO:tensorflow:step = 33901, loss = 0.703278, precision = 0.851562 (11.101 sec) +INFO:tensorflow:global_step/sec: 9.01285 +INFO:tensorflow:step = 34001, loss = 0.548128, precision = 0.875 (11.095 sec) +Saved checkpoint after 87 epoch(s) to ../data/resnet20/checkpoints/00087... +INFO:tensorflow:global_step/sec: 8.75229 +INFO:tensorflow:step = 34101, loss = 0.706518, precision = 0.84375 (11.426 sec) +INFO:tensorflow:global_step/sec: 9.00555 +INFO:tensorflow:step = 34201, loss = 0.618722, precision = 0.890625 (11.104 sec) +INFO:tensorflow:global_step/sec: 9.01789 +INFO:tensorflow:step = 34301, loss = 0.502085, precision = 0.9375 (11.089 sec) +INFO:tensorflow:global_step/sec: 9.01362 +INFO:tensorflow:step = 34401, loss = 0.746867, precision = 0.835938 (11.094 sec) +Saved checkpoint after 88 epoch(s) to ../data/resnet20/checkpoints/00088... +INFO:tensorflow:global_step/sec: 8.75323 +INFO:tensorflow:step = 34501, loss = 0.635168, precision = 0.882812 (11.424 sec) +INFO:tensorflow:global_step/sec: 9.01108 +INFO:tensorflow:step = 34601, loss = 0.636282, precision = 0.890625 (11.098 sec) +INFO:tensorflow:global_step/sec: 9.00676 +INFO:tensorflow:step = 34701, loss = 0.721462, precision = 0.851562 (11.103 sec) +Saved checkpoint after 89 epoch(s) to ../data/resnet20/checkpoints/00089... +INFO:tensorflow:global_step/sec: 8.766 +INFO:tensorflow:step = 34801, loss = 0.584088, precision = 0.882812 (11.408 sec) +INFO:tensorflow:global_step/sec: 9.01204 +INFO:tensorflow:step = 34901, loss = 0.549122, precision = 0.929688 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.01128 +INFO:tensorflow:step = 35001, loss = 0.62851, precision = 0.867188 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.01496 +INFO:tensorflow:step = 35101, loss = 0.6026, precision = 0.882812 (11.093 sec) +Saved checkpoint after 90 epoch(s) to ../data/resnet20/checkpoints/00090... +INFO:tensorflow:global_step/sec: 8.74682 +INFO:tensorflow:step = 35201, loss = 0.676822, precision = 0.898438 (11.433 sec) +INFO:tensorflow:global_step/sec: 9.01596 +INFO:tensorflow:step = 35301, loss = 0.640556, precision = 0.859375 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.00717 +INFO:tensorflow:step = 35401, loss = 0.554639, precision = 0.914062 (11.102 sec) +INFO:tensorflow:global_step/sec: 9.00338 +INFO:tensorflow:step = 35501, loss = 0.840462, precision = 0.796875 (11.107 sec) +Saved checkpoint after 91 epoch(s) to ../data/resnet20/checkpoints/00091... +INFO:tensorflow:global_step/sec: 8.73899 +INFO:tensorflow:step = 35601, loss = 0.591701, precision = 0.90625 (11.443 sec) +INFO:tensorflow:global_step/sec: 9.00632 +INFO:tensorflow:step = 35701, loss = 0.417147, precision = 0.945312 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01275 +INFO:tensorflow:step = 35801, loss = 0.46049, precision = 0.945312 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.0075 +INFO:tensorflow:step = 35901, loss = 0.556271, precision = 0.882812 (11.102 sec) +Saved checkpoint after 92 epoch(s) to ../data/resnet20/checkpoints/00092... +INFO:tensorflow:global_step/sec: 8.75927 +INFO:tensorflow:step = 36001, loss = 0.501671, precision = 0.929688 (11.417 sec) +INFO:tensorflow:global_step/sec: 9.01395 +INFO:tensorflow:step = 36101, loss = 0.492486, precision = 0.9375 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.01148 +INFO:tensorflow:step = 36201, loss = 0.476714, precision = 0.914062 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.01639 +INFO:tensorflow:step = 36301, loss = 0.425025, precision = 0.953125 (11.091 sec) +Saved checkpoint after 93 epoch(s) to ../data/resnet20/checkpoints/00093... +INFO:tensorflow:global_step/sec: 8.75944 +INFO:tensorflow:step = 36401, loss = 0.500866, precision = 0.929688 (11.416 sec) +INFO:tensorflow:global_step/sec: 9.008 +INFO:tensorflow:step = 36501, loss = 0.410685, precision = 0.929688 (11.101 sec) +INFO:tensorflow:global_step/sec: 9.00091 +INFO:tensorflow:step = 36601, loss = 0.424024, precision = 0.9375 (11.110 sec) +INFO:tensorflow:global_step/sec: 9.02649 +INFO:tensorflow:step = 36701, loss = 0.435975, precision = 0.921875 (11.079 sec) +Saved checkpoint after 94 epoch(s) to ../data/resnet20/checkpoints/00094... +INFO:tensorflow:global_step/sec: 8.75294 +INFO:tensorflow:step = 36801, loss = 0.435267, precision = 0.9375 (11.425 sec) +INFO:tensorflow:global_step/sec: 9.02532 +INFO:tensorflow:step = 36901, loss = 0.390481, precision = 0.953125 (11.080 sec) +INFO:tensorflow:global_step/sec: 9.00486 +INFO:tensorflow:step = 37001, loss = 0.532244, precision = 0.890625 (11.105 sec) +INFO:tensorflow:global_step/sec: 9.00621 +INFO:tensorflow:step = 37101, loss = 0.437021, precision = 0.914062 (11.103 sec) +Saved checkpoint after 95 epoch(s) to ../data/resnet20/checkpoints/00095... +INFO:tensorflow:global_step/sec: 8.74534 +INFO:tensorflow:step = 37201, loss = 0.43896, precision = 0.945312 (11.435 sec) +INFO:tensorflow:global_step/sec: 9.01357 +INFO:tensorflow:step = 37301, loss = 0.390559, precision = 0.953125 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.00969 +INFO:tensorflow:step = 37401, loss = 0.394162, precision = 0.9375 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.01531 +INFO:tensorflow:step = 37501, loss = 0.407274, precision = 0.921875 (11.092 sec) +Saved checkpoint after 96 epoch(s) to ../data/resnet20/checkpoints/00096... +INFO:tensorflow:global_step/sec: 8.72154 +INFO:tensorflow:step = 37601, loss = 0.342958, precision = 0.984375 (11.466 sec) +INFO:tensorflow:global_step/sec: 9.01418 +INFO:tensorflow:step = 37701, loss = 0.400225, precision = 0.953125 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.00655 +INFO:tensorflow:step = 37801, loss = 0.419037, precision = 0.9375 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.02113 +INFO:tensorflow:step = 37901, loss = 0.370381, precision = 0.929688 (11.085 sec) +Saved checkpoint after 97 epoch(s) to ../data/resnet20/checkpoints/00097... +INFO:tensorflow:global_step/sec: 8.74933 +INFO:tensorflow:step = 38001, loss = 0.432328, precision = 0.9375 (11.429 sec) +INFO:tensorflow:global_step/sec: 9.00919 +INFO:tensorflow:step = 38101, loss = 0.36883, precision = 0.953125 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.0222 +INFO:tensorflow:step = 38201, loss = 0.43204, precision = 0.929688 (11.084 sec) +INFO:tensorflow:global_step/sec: 9.01537 +INFO:tensorflow:step = 38301, loss = 0.30894, precision = 0.96875 (11.092 sec) +Saved checkpoint after 98 epoch(s) to ../data/resnet20/checkpoints/00098... +INFO:tensorflow:global_step/sec: 8.74633 +INFO:tensorflow:step = 38401, loss = 0.38563, precision = 0.945312 (11.434 sec) +INFO:tensorflow:global_step/sec: 9.01837 +INFO:tensorflow:step = 38501, loss = 0.350069, precision = 0.945312 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.01235 +INFO:tensorflow:step = 38601, loss = 0.426081, precision = 0.9375 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.02132 +INFO:tensorflow:step = 38701, loss = 0.393595, precision = 0.945312 (11.085 sec) +Saved checkpoint after 99 epoch(s) to ../data/resnet20/checkpoints/00099... +INFO:tensorflow:global_step/sec: 8.7447 +INFO:tensorflow:step = 38801, loss = 0.379145, precision = 0.96875 (11.436 sec) +INFO:tensorflow:global_step/sec: 9.01328 +INFO:tensorflow:step = 38901, loss = 0.308018, precision = 0.976562 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.01894 +INFO:tensorflow:step = 39001, loss = 0.265771, precision = 0.992188 (11.088 sec) +Saved checkpoint after 100 epoch(s) to ../data/resnet20/checkpoints/00100... +INFO:tensorflow:global_step/sec: 8.76626 +INFO:tensorflow:step = 39101, loss = 0.315631, precision = 0.96875 (11.408 sec) +INFO:tensorflow:global_step/sec: 9.00687 +INFO:tensorflow:step = 39201, loss = 0.320705, precision = 0.96875 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01207 +INFO:tensorflow:step = 39301, loss = 0.315002, precision = 0.960938 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.01836 +INFO:tensorflow:step = 39401, loss = 0.378302, precision = 0.96875 (11.088 sec) +Saved checkpoint after 101 epoch(s) to ../data/resnet20/checkpoints/00101... +INFO:tensorflow:global_step/sec: 8.74822 +INFO:tensorflow:step = 39501, loss = 0.312256, precision = 0.960938 (11.431 sec) +INFO:tensorflow:global_step/sec: 9.00591 +INFO:tensorflow:step = 39601, loss = 0.321204, precision = 0.96875 (11.104 sec) +INFO:tensorflow:global_step/sec: 9.02661 +INFO:tensorflow:step = 39701, loss = 0.330682, precision = 0.960938 (11.078 sec) +INFO:tensorflow:global_step/sec: 9.00273 +INFO:tensorflow:step = 39801, loss = 0.331432, precision = 0.945312 (11.108 sec) +Saved checkpoint after 102 epoch(s) to ../data/resnet20/checkpoints/00102... +INFO:tensorflow:global_step/sec: 8.75585 +INFO:tensorflow:step = 39901, loss = 0.319653, precision = 0.953125 (11.421 sec) +INFO:tensorflow:global_step/sec: 9.00603 +INFO:tensorflow:step = 40001, loss = 0.329484, precision = 0.960938 (11.104 sec) +INFO:tensorflow:global_step/sec: 9.00874 +INFO:tensorflow:step = 40101, loss = 0.27351, precision = 0.984375 (11.100 sec) +INFO:tensorflow:global_step/sec: 8.99893 +INFO:tensorflow:step = 40201, loss = 0.352933, precision = 0.9375 (11.113 sec) +Saved checkpoint after 103 epoch(s) to ../data/resnet20/checkpoints/00103... +INFO:tensorflow:global_step/sec: 8.75286 +INFO:tensorflow:step = 40301, loss = 0.334791, precision = 0.976562 (11.425 sec) +INFO:tensorflow:global_step/sec: 8.99593 +INFO:tensorflow:step = 40401, loss = 0.293281, precision = 0.96875 (11.116 sec) +INFO:tensorflow:global_step/sec: 9.01913 +INFO:tensorflow:step = 40501, loss = 0.251796, precision = 0.976562 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.00447 +INFO:tensorflow:step = 40601, loss = 0.374165, precision = 0.921875 (11.106 sec) +Saved checkpoint after 104 epoch(s) to ../data/resnet20/checkpoints/00104... +INFO:tensorflow:global_step/sec: 8.75719 +INFO:tensorflow:step = 40701, loss = 0.351609, precision = 0.929688 (11.419 sec) +INFO:tensorflow:global_step/sec: 9.01712 +INFO:tensorflow:step = 40801, loss = 0.359336, precision = 0.953125 (11.090 sec) +INFO:tensorflow:global_step/sec: 9.00677 +INFO:tensorflow:step = 40901, loss = 0.313301, precision = 0.953125 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01567 +INFO:tensorflow:step = 41001, loss = 0.321315, precision = 0.9375 (11.092 sec) +Saved checkpoint after 105 epoch(s) to ../data/resnet20/checkpoints/00105... +INFO:tensorflow:global_step/sec: 8.75717 +INFO:tensorflow:step = 41101, loss = 0.311392, precision = 0.960938 (11.419 sec) +INFO:tensorflow:global_step/sec: 8.98529 +INFO:tensorflow:step = 41201, loss = 0.310681, precision = 0.960938 (11.129 sec) +INFO:tensorflow:global_step/sec: 9.01757 +INFO:tensorflow:step = 41301, loss = 0.287605, precision = 0.984375 (11.090 sec) +INFO:tensorflow:global_step/sec: 9.00978 +INFO:tensorflow:step = 41401, loss = 0.270993, precision = 0.960938 (11.099 sec) +Saved checkpoint after 106 epoch(s) to ../data/resnet20/checkpoints/00106... +INFO:tensorflow:global_step/sec: 8.74766 +INFO:tensorflow:step = 41501, loss = 0.362662, precision = 0.953125 (11.432 sec) +INFO:tensorflow:global_step/sec: 9.01823 +INFO:tensorflow:step = 41601, loss = 0.296828, precision = 0.96875 (11.089 sec) +INFO:tensorflow:global_step/sec: 9.00634 +INFO:tensorflow:step = 41701, loss = 0.361416, precision = 0.921875 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01621 +INFO:tensorflow:step = 41801, loss = 0.246006, precision = 0.976562 (11.091 sec) +Saved checkpoint after 107 epoch(s) to ../data/resnet20/checkpoints/00107... +INFO:tensorflow:global_step/sec: 8.74913 +INFO:tensorflow:step = 41901, loss = 0.294432, precision = 0.960938 (11.429 sec) +INFO:tensorflow:global_step/sec: 9.00882 +INFO:tensorflow:step = 42001, loss = 0.285762, precision = 0.953125 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.00669 +INFO:tensorflow:step = 42101, loss = 0.309329, precision = 0.960938 (11.103 sec) +INFO:tensorflow:global_step/sec: 8.99974 +INFO:tensorflow:step = 42201, loss = 0.364849, precision = 0.9375 (11.111 sec) +Saved checkpoint after 108 epoch(s) to ../data/resnet20/checkpoints/00108... +INFO:tensorflow:global_step/sec: 8.75587 +INFO:tensorflow:step = 42301, loss = 0.299221, precision = 0.945312 (11.421 sec) +INFO:tensorflow:global_step/sec: 9.00103 +INFO:tensorflow:step = 42401, loss = 0.234287, precision = 0.984375 (11.110 sec) +INFO:tensorflow:global_step/sec: 9.00947 +INFO:tensorflow:step = 42501, loss = 0.275395, precision = 0.96875 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.01362 +INFO:tensorflow:step = 42601, loss = 0.274846, precision = 0.96875 (11.094 sec) +Saved checkpoint after 109 epoch(s) to ../data/resnet20/checkpoints/00109... +INFO:tensorflow:global_step/sec: 8.75114 +INFO:tensorflow:step = 42701, loss = 0.291756, precision = 0.953125 (11.427 sec) +INFO:tensorflow:global_step/sec: 9.01954 +INFO:tensorflow:step = 42801, loss = 0.241716, precision = 0.96875 (11.087 sec) +INFO:tensorflow:global_step/sec: 9.01033 +INFO:tensorflow:step = 42901, loss = 0.307645, precision = 0.953125 (11.098 sec) +INFO:tensorflow:global_step/sec: 9.01208 +INFO:tensorflow:step = 43001, loss = 0.289848, precision = 0.960938 (11.096 sec) +Saved checkpoint after 110 epoch(s) to ../data/resnet20/checkpoints/00110... +INFO:tensorflow:global_step/sec: 8.73591 +INFO:tensorflow:step = 43101, loss = 0.236719, precision = 0.960938 (11.447 sec) +INFO:tensorflow:global_step/sec: 9.00593 +INFO:tensorflow:step = 43201, loss = 0.289968, precision = 0.960938 (11.104 sec) +INFO:tensorflow:global_step/sec: 9.00922 +INFO:tensorflow:step = 43301, loss = 0.303104, precision = 0.945312 (11.100 sec) +Saved checkpoint after 111 epoch(s) to ../data/resnet20/checkpoints/00111... +INFO:tensorflow:global_step/sec: 8.74654 +INFO:tensorflow:step = 43401, loss = 0.267911, precision = 0.984375 (11.433 sec) +INFO:tensorflow:global_step/sec: 9.01127 +INFO:tensorflow:step = 43501, loss = 0.296184, precision = 0.953125 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.01585 +INFO:tensorflow:step = 43601, loss = 0.262742, precision = 0.976562 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.01246 +INFO:tensorflow:step = 43701, loss = 0.299303, precision = 0.953125 (11.096 sec) +Saved checkpoint after 112 epoch(s) to ../data/resnet20/checkpoints/00112... +INFO:tensorflow:global_step/sec: 8.74081 +INFO:tensorflow:step = 43801, loss = 0.314915, precision = 0.945312 (11.441 sec) +INFO:tensorflow:global_step/sec: 9.01451 +INFO:tensorflow:step = 43901, loss = 0.242575, precision = 0.96875 (11.093 sec) +INFO:tensorflow:global_step/sec: 9.0102 +INFO:tensorflow:step = 44001, loss = 0.306546, precision = 0.945312 (11.098 sec) +INFO:tensorflow:global_step/sec: 9.0084 +INFO:tensorflow:step = 44101, loss = 0.314002, precision = 0.953125 (11.101 sec) +Saved checkpoint after 113 epoch(s) to ../data/resnet20/checkpoints/00113... +INFO:tensorflow:global_step/sec: 8.75442 +INFO:tensorflow:step = 44201, loss = 0.259139, precision = 0.96875 (11.423 sec) +INFO:tensorflow:global_step/sec: 9.01066 +INFO:tensorflow:step = 44301, loss = 0.280753, precision = 0.945312 (11.098 sec) +INFO:tensorflow:global_step/sec: 9.01946 +INFO:tensorflow:step = 44401, loss = 0.321928, precision = 0.945312 (11.087 sec) +INFO:tensorflow:global_step/sec: 9.0148 +INFO:tensorflow:step = 44501, loss = 0.303085, precision = 0.945312 (11.093 sec) +Saved checkpoint after 114 epoch(s) to ../data/resnet20/checkpoints/00114... +INFO:tensorflow:global_step/sec: 8.71824 +INFO:tensorflow:step = 44601, loss = 0.280336, precision = 0.953125 (11.470 sec) +INFO:tensorflow:global_step/sec: 8.99948 +INFO:tensorflow:step = 44701, loss = 0.270849, precision = 0.945312 (11.112 sec) +INFO:tensorflow:global_step/sec: 9.01508 +INFO:tensorflow:step = 44801, loss = 0.244212, precision = 0.960938 (11.093 sec) +INFO:tensorflow:global_step/sec: 9.01429 +INFO:tensorflow:step = 44901, loss = 0.222509, precision = 0.96875 (11.094 sec) +Saved checkpoint after 115 epoch(s) to ../data/resnet20/checkpoints/00115... +INFO:tensorflow:global_step/sec: 8.75848 +INFO:tensorflow:step = 45001, loss = 0.239266, precision = 0.960938 (11.418 sec) +INFO:tensorflow:global_step/sec: 9.00988 +INFO:tensorflow:step = 45101, loss = 0.298769, precision = 0.9375 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.01227 +INFO:tensorflow:step = 45201, loss = 0.248224, precision = 0.953125 (11.096 sec) +INFO:tensorflow:global_step/sec: 8.99597 +INFO:tensorflow:step = 45301, loss = 0.232787, precision = 0.976562 (11.116 sec) +Saved checkpoint after 116 epoch(s) to ../data/resnet20/checkpoints/00116... +INFO:tensorflow:global_step/sec: 8.74535 +INFO:tensorflow:step = 45401, loss = 0.234362, precision = 0.96875 (11.435 sec) +INFO:tensorflow:global_step/sec: 9.01406 +INFO:tensorflow:step = 45501, loss = 0.319444, precision = 0.9375 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.01421 +INFO:tensorflow:step = 45601, loss = 0.315393, precision = 0.945312 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.00823 +INFO:tensorflow:step = 45701, loss = 0.318256, precision = 0.921875 (11.101 sec) +Saved checkpoint after 117 epoch(s) to ../data/resnet20/checkpoints/00117... +INFO:tensorflow:global_step/sec: 8.73983 +INFO:tensorflow:step = 45801, loss = 0.331186, precision = 0.9375 (11.442 sec) +INFO:tensorflow:global_step/sec: 9.01029 +INFO:tensorflow:step = 45901, loss = 0.278036, precision = 0.945312 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.00867 +INFO:tensorflow:step = 46001, loss = 0.271486, precision = 0.960938 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.01792 +INFO:tensorflow:step = 46101, loss = 0.231086, precision = 0.984375 (11.089 sec) +Saved checkpoint after 118 epoch(s) to ../data/resnet20/checkpoints/00118... +INFO:tensorflow:global_step/sec: 8.76144 +INFO:tensorflow:step = 46201, loss = 0.216466, precision = 0.960938 (11.414 sec) +INFO:tensorflow:global_step/sec: 9.02201 +INFO:tensorflow:step = 46301, loss = 0.192067, precision = 1.0 (11.084 sec) +INFO:tensorflow:global_step/sec: 9.00729 +INFO:tensorflow:step = 46401, loss = 0.294274, precision = 0.96875 (11.102 sec) +INFO:tensorflow:global_step/sec: 8.99714 +INFO:tensorflow:step = 46501, loss = 0.214813, precision = 0.976562 (11.115 sec) +Saved checkpoint after 119 epoch(s) to ../data/resnet20/checkpoints/00119... +INFO:tensorflow:global_step/sec: 8.75373 +INFO:tensorflow:step = 46601, loss = 0.260278, precision = 0.953125 (11.424 sec) +INFO:tensorflow:global_step/sec: 9.00076 +INFO:tensorflow:step = 46701, loss = 0.353015, precision = 0.921875 (11.110 sec) +INFO:tensorflow:global_step/sec: 9.00737 +INFO:tensorflow:step = 46801, loss = 0.306823, precision = 0.953125 (11.102 sec) +INFO:tensorflow:global_step/sec: 9.01207 +INFO:tensorflow:step = 46901, loss = 0.276564, precision = 0.914062 (11.096 sec) +Saved checkpoint after 120 epoch(s) to ../data/resnet20/checkpoints/00120... +INFO:tensorflow:global_step/sec: 8.74718 +INFO:tensorflow:step = 47001, loss = 0.222523, precision = 0.976562 (11.432 sec) +INFO:tensorflow:global_step/sec: 9.01393 +INFO:tensorflow:step = 47101, loss = 0.293465, precision = 0.953125 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.01797 +INFO:tensorflow:step = 47201, loss = 0.205114, precision = 0.984375 (11.089 sec) +INFO:tensorflow:global_step/sec: 9.01256 +INFO:tensorflow:step = 47301, loss = 0.237512, precision = 0.960938 (11.096 sec) +Saved checkpoint after 121 epoch(s) to ../data/resnet20/checkpoints/00121... +INFO:tensorflow:global_step/sec: 8.74244 +INFO:tensorflow:step = 47401, loss = 0.225021, precision = 0.96875 (11.438 sec) +INFO:tensorflow:global_step/sec: 8.99935 +INFO:tensorflow:step = 47501, loss = 0.328246, precision = 0.9375 (11.112 sec) +INFO:tensorflow:global_step/sec: 9.02895 +INFO:tensorflow:step = 47601, loss = 0.270316, precision = 0.945312 (11.075 sec) +INFO:tensorflow:global_step/sec: 9.0062 +INFO:tensorflow:step = 47701, loss = 0.222482, precision = 0.960938 (11.103 sec) +Saved checkpoint after 122 epoch(s) to ../data/resnet20/checkpoints/00122... +INFO:tensorflow:global_step/sec: 8.75525 +INFO:tensorflow:step = 47801, loss = 0.325652, precision = 0.914062 (11.422 sec) +INFO:tensorflow:global_step/sec: 8.99818 +INFO:tensorflow:step = 47901, loss = 0.291383, precision = 0.945312 (11.113 sec) +INFO:tensorflow:global_step/sec: 9.0195 +INFO:tensorflow:step = 48001, loss = 0.267326, precision = 0.945312 (11.087 sec) +Saved checkpoint after 123 epoch(s) to ../data/resnet20/checkpoints/00123... +INFO:tensorflow:global_step/sec: 8.75692 +INFO:tensorflow:step = 48101, loss = 0.259803, precision = 0.96875 (11.419 sec) +INFO:tensorflow:global_step/sec: 9.01956 +INFO:tensorflow:step = 48201, loss = 0.251688, precision = 0.945312 (11.087 sec) +INFO:tensorflow:global_step/sec: 9.0021 +INFO:tensorflow:step = 48301, loss = 0.28553, precision = 0.953125 (11.108 sec) +INFO:tensorflow:global_step/sec: 9.00536 +INFO:tensorflow:step = 48401, loss = 0.214308, precision = 0.976562 (11.105 sec) +Saved checkpoint after 124 epoch(s) to ../data/resnet20/checkpoints/00124... +INFO:tensorflow:global_step/sec: 8.74482 +INFO:tensorflow:step = 48501, loss = 0.291169, precision = 0.929688 (11.435 sec) +INFO:tensorflow:global_step/sec: 9.01065 +INFO:tensorflow:step = 48601, loss = 0.220252, precision = 0.96875 (11.098 sec) +INFO:tensorflow:global_step/sec: 9.01773 +INFO:tensorflow:step = 48701, loss = 0.314935, precision = 0.945312 (11.089 sec) +INFO:tensorflow:global_step/sec: 9.01526 +INFO:tensorflow:step = 48801, loss = 0.236005, precision = 0.96875 (11.092 sec) +Saved checkpoint after 125 epoch(s) to ../data/resnet20/checkpoints/00125... +INFO:tensorflow:global_step/sec: 8.74783 +INFO:tensorflow:step = 48901, loss = 0.332051, precision = 0.921875 (11.431 sec) +INFO:tensorflow:global_step/sec: 9.01342 +INFO:tensorflow:step = 49001, loss = 0.21209, precision = 0.96875 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.01143 +INFO:tensorflow:step = 49101, loss = 0.230055, precision = 0.96875 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.00567 +INFO:tensorflow:step = 49201, loss = 0.276684, precision = 0.9375 (11.104 sec) +Saved checkpoint after 126 epoch(s) to ../data/resnet20/checkpoints/00126... +INFO:tensorflow:global_step/sec: 8.72893 +INFO:tensorflow:step = 49301, loss = 0.195963, precision = 0.984375 (11.456 sec) +INFO:tensorflow:global_step/sec: 9.00488 +INFO:tensorflow:step = 49401, loss = 0.300385, precision = 0.945312 (11.105 sec) +INFO:tensorflow:global_step/sec: 9.02021 +INFO:tensorflow:step = 49501, loss = 0.250037, precision = 0.953125 (11.086 sec) +INFO:tensorflow:global_step/sec: 9.00775 +INFO:tensorflow:step = 49601, loss = 0.242279, precision = 0.960938 (11.101 sec) +Saved checkpoint after 127 epoch(s) to ../data/resnet20/checkpoints/00127... +INFO:tensorflow:global_step/sec: 8.74969 +INFO:tensorflow:step = 49701, loss = 0.24471, precision = 0.960938 (11.429 sec) +INFO:tensorflow:global_step/sec: 9.00647 +INFO:tensorflow:step = 49801, loss = 0.220247, precision = 0.960938 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01689 +INFO:tensorflow:step = 49901, loss = 0.311207, precision = 0.9375 (11.090 sec) +INFO:tensorflow:global_step/sec: 9.00148 +INFO:tensorflow:step = 50001, loss = 0.304568, precision = 0.921875 (11.109 sec) +Saved checkpoint after 128 epoch(s) to ../data/resnet20/checkpoints/00128... +INFO:tensorflow:global_step/sec: 8.74128 +INFO:tensorflow:step = 50101, loss = 0.323316, precision = 0.945312 (11.440 sec) +INFO:tensorflow:global_step/sec: 9.01436 +INFO:tensorflow:step = 50201, loss = 0.340589, precision = 0.921875 (11.093 sec) +INFO:tensorflow:global_step/sec: 8.99009 +INFO:tensorflow:step = 50301, loss = 0.241719, precision = 0.960938 (11.123 sec) +INFO:tensorflow:global_step/sec: 9.01939 +INFO:tensorflow:step = 50401, loss = 0.230264, precision = 0.984375 (11.087 sec) +Saved checkpoint after 129 epoch(s) to ../data/resnet20/checkpoints/00129... +INFO:tensorflow:global_step/sec: 8.75746 +INFO:tensorflow:step = 50501, loss = 0.279854, precision = 0.9375 (11.419 sec) +INFO:tensorflow:global_step/sec: 9.01249 +INFO:tensorflow:step = 50601, loss = 0.184431, precision = 0.976562 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.02411 +INFO:tensorflow:step = 50701, loss = 0.339309, precision = 0.9375 (11.081 sec) +INFO:tensorflow:global_step/sec: 9.01791 +INFO:tensorflow:step = 50801, loss = 0.229369, precision = 0.953125 (11.089 sec) +Saved checkpoint after 130 epoch(s) to ../data/resnet20/checkpoints/00130... +INFO:tensorflow:global_step/sec: 8.74114 +INFO:tensorflow:step = 50901, loss = 0.192114, precision = 0.984375 (11.440 sec) +INFO:tensorflow:global_step/sec: 9.01683 +INFO:tensorflow:step = 51001, loss = 0.241979, precision = 0.96875 (11.090 sec) +INFO:tensorflow:global_step/sec: 9.00697 +INFO:tensorflow:step = 51101, loss = 0.255652, precision = 0.953125 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01628 +INFO:tensorflow:step = 51201, loss = 0.290382, precision = 0.945312 (11.091 sec) +Saved checkpoint after 131 epoch(s) to ../data/resnet20/checkpoints/00131... +INFO:tensorflow:global_step/sec: 8.76498 +INFO:tensorflow:step = 51301, loss = 0.22099, precision = 0.960938 (11.409 sec) +INFO:tensorflow:global_step/sec: 9.01639 +INFO:tensorflow:step = 51401, loss = 0.288048, precision = 0.953125 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.01614 +INFO:tensorflow:step = 51501, loss = 0.247262, precision = 0.96875 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.01667 +INFO:tensorflow:step = 51601, loss = 0.246674, precision = 0.960938 (11.091 sec) +Saved checkpoint after 132 epoch(s) to ../data/resnet20/checkpoints/00132... +INFO:tensorflow:global_step/sec: 8.74163 +INFO:tensorflow:step = 51701, loss = 0.182372, precision = 0.976562 (11.439 sec) +INFO:tensorflow:global_step/sec: 9.01691 +INFO:tensorflow:step = 51801, loss = 0.247798, precision = 0.96875 (11.090 sec) +INFO:tensorflow:global_step/sec: 9.01564 +INFO:tensorflow:step = 51901, loss = 0.36362, precision = 0.9375 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.0139 +INFO:tensorflow:step = 52001, loss = 0.265308, precision = 0.953125 (11.094 sec) +Saved checkpoint after 133 epoch(s) to ../data/resnet20/checkpoints/00133... +INFO:tensorflow:global_step/sec: 8.75084 +INFO:tensorflow:step = 52101, loss = 0.234101, precision = 0.976562 (11.427 sec) +INFO:tensorflow:global_step/sec: 9.02352 +INFO:tensorflow:step = 52201, loss = 0.209635, precision = 0.976562 (11.082 sec) +INFO:tensorflow:global_step/sec: 9.00647 +INFO:tensorflow:step = 52301, loss = 0.237694, precision = 0.976562 (11.103 sec) +Saved checkpoint after 134 epoch(s) to ../data/resnet20/checkpoints/00134... +INFO:tensorflow:global_step/sec: 8.70434 +INFO:tensorflow:step = 52401, loss = 0.216798, precision = 0.96875 (11.489 sec) +INFO:tensorflow:global_step/sec: 9.01054 +INFO:tensorflow:step = 52501, loss = 0.25417, precision = 0.953125 (11.098 sec) +INFO:tensorflow:global_step/sec: 9.00975 +INFO:tensorflow:step = 52601, loss = 0.198886, precision = 0.992188 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.01961 +INFO:tensorflow:step = 52701, loss = 0.227357, precision = 0.960938 (11.087 sec) +Saved checkpoint after 135 epoch(s) to ../data/resnet20/checkpoints/00135... +INFO:tensorflow:global_step/sec: 8.75508 +INFO:tensorflow:step = 52801, loss = 0.246265, precision = 0.960938 (11.422 sec) +INFO:tensorflow:global_step/sec: 9.01529 +INFO:tensorflow:step = 52901, loss = 0.260796, precision = 0.953125 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.02282 +INFO:tensorflow:step = 53001, loss = 0.210368, precision = 0.984375 (11.083 sec) +INFO:tensorflow:global_step/sec: 9.01412 +INFO:tensorflow:step = 53101, loss = 0.196672, precision = 0.976562 (11.094 sec) +Saved checkpoint after 136 epoch(s) to ../data/resnet20/checkpoints/00136... +INFO:tensorflow:global_step/sec: 8.75804 +INFO:tensorflow:step = 53201, loss = 0.215664, precision = 0.976562 (11.418 sec) +INFO:tensorflow:global_step/sec: 9.00858 +INFO:tensorflow:step = 53301, loss = 0.159766, precision = 1.0 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.01526 +INFO:tensorflow:step = 53401, loss = 0.237688, precision = 0.960938 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.01835 +INFO:tensorflow:step = 53501, loss = 0.260839, precision = 0.929688 (11.088 sec) +Saved checkpoint after 137 epoch(s) to ../data/resnet20/checkpoints/00137... +INFO:tensorflow:global_step/sec: 8.74868 +INFO:tensorflow:step = 53601, loss = 0.175075, precision = 0.976562 (11.430 sec) +INFO:tensorflow:global_step/sec: 9.00502 +INFO:tensorflow:step = 53701, loss = 0.187011, precision = 0.976562 (11.105 sec) +INFO:tensorflow:global_step/sec: 9.01402 +INFO:tensorflow:step = 53801, loss = 0.175857, precision = 0.992188 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.01655 +INFO:tensorflow:step = 53901, loss = 0.215817, precision = 0.96875 (11.091 sec) +Saved checkpoint after 138 epoch(s) to ../data/resnet20/checkpoints/00138... +INFO:tensorflow:global_step/sec: 8.74293 +INFO:tensorflow:step = 54001, loss = 0.226281, precision = 0.960938 (11.438 sec) +INFO:tensorflow:global_step/sec: 9.01863 +INFO:tensorflow:step = 54101, loss = 0.180381, precision = 0.984375 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.01919 +INFO:tensorflow:step = 54201, loss = 0.225013, precision = 0.960938 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.01081 +INFO:tensorflow:step = 54301, loss = 0.164974, precision = 0.984375 (11.098 sec) +Saved checkpoint after 139 epoch(s) to ../data/resnet20/checkpoints/00139... +INFO:tensorflow:global_step/sec: 8.74045 +INFO:tensorflow:step = 54401, loss = 0.153513, precision = 1.0 (11.441 sec) +INFO:tensorflow:global_step/sec: 9.01804 +INFO:tensorflow:step = 54501, loss = 0.200838, precision = 0.96875 (11.089 sec) +INFO:tensorflow:global_step/sec: 9.00944 +INFO:tensorflow:step = 54601, loss = 0.184138, precision = 0.984375 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.0178 +INFO:tensorflow:step = 54701, loss = 0.163899, precision = 0.992188 (11.089 sec) +Saved checkpoint after 140 epoch(s) to ../data/resnet20/checkpoints/00140... +INFO:tensorflow:global_step/sec: 8.73878 +INFO:tensorflow:step = 54801, loss = 0.18048, precision = 0.984375 (11.443 sec) +INFO:tensorflow:global_step/sec: 9.02691 +INFO:tensorflow:step = 54901, loss = 0.184233, precision = 0.984375 (11.078 sec) +INFO:tensorflow:global_step/sec: 9.02325 +INFO:tensorflow:step = 55001, loss = 0.178442, precision = 0.984375 (11.083 sec) +INFO:tensorflow:global_step/sec: 9.00126 +INFO:tensorflow:step = 55101, loss = 0.175062, precision = 0.976562 (11.109 sec) +Saved checkpoint after 141 epoch(s) to ../data/resnet20/checkpoints/00141... +INFO:tensorflow:global_step/sec: 8.71698 +INFO:tensorflow:step = 55201, loss = 0.171435, precision = 0.984375 (11.472 sec) +INFO:tensorflow:global_step/sec: 9.01244 +INFO:tensorflow:step = 55301, loss = 0.180263, precision = 0.984375 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.03287 +INFO:tensorflow:step = 55401, loss = 0.195016, precision = 0.976562 (11.070 sec) +INFO:tensorflow:global_step/sec: 9.02135 +INFO:tensorflow:step = 55501, loss = 0.162751, precision = 1.0 (11.085 sec) +Saved checkpoint after 142 epoch(s) to ../data/resnet20/checkpoints/00142... +INFO:tensorflow:global_step/sec: 8.79371 +INFO:tensorflow:step = 55601, loss = 0.18352, precision = 0.984375 (11.372 sec) +INFO:tensorflow:global_step/sec: 9.02465 +INFO:tensorflow:step = 55701, loss = 0.164027, precision = 0.992188 (11.081 sec) +INFO:tensorflow:global_step/sec: 9.016 +INFO:tensorflow:step = 55801, loss = 0.162047, precision = 1.0 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.01462 +INFO:tensorflow:step = 55901, loss = 0.179946, precision = 0.984375 (11.093 sec) +Saved checkpoint after 143 epoch(s) to ../data/resnet20/checkpoints/00143... +INFO:tensorflow:global_step/sec: 8.74256 +INFO:tensorflow:step = 56001, loss = 0.182114, precision = 0.984375 (11.438 sec) +INFO:tensorflow:global_step/sec: 9.00576 +INFO:tensorflow:step = 56101, loss = 0.170763, precision = 0.992188 (11.104 sec) +INFO:tensorflow:global_step/sec: 9.01305 +INFO:tensorflow:step = 56201, loss = 0.167513, precision = 0.992188 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.0212 +INFO:tensorflow:step = 56301, loss = 0.159383, precision = 1.0 (11.085 sec) +Saved checkpoint after 144 epoch(s) to ../data/resnet20/checkpoints/00144... +INFO:tensorflow:global_step/sec: 8.75429 +INFO:tensorflow:step = 56401, loss = 0.171626, precision = 0.992188 (11.423 sec) +INFO:tensorflow:global_step/sec: 9.01435 +INFO:tensorflow:step = 56501, loss = 0.166502, precision = 1.0 (11.093 sec) +INFO:tensorflow:global_step/sec: 9.012 +INFO:tensorflow:step = 56601, loss = 0.143635, precision = 1.0 (11.096 sec) +Saved checkpoint after 145 epoch(s) to ../data/resnet20/checkpoints/00145... +INFO:tensorflow:global_step/sec: 8.74278 +INFO:tensorflow:step = 56701, loss = 0.145622, precision = 1.0 (11.438 sec) +INFO:tensorflow:global_step/sec: 9.00042 +INFO:tensorflow:step = 56801, loss = 0.183842, precision = 0.992188 (11.110 sec) +INFO:tensorflow:global_step/sec: 9.00463 +INFO:tensorflow:step = 56901, loss = 0.145274, precision = 1.0 (11.105 sec) +INFO:tensorflow:global_step/sec: 9.0112 +INFO:tensorflow:step = 57001, loss = 0.16861, precision = 0.992188 (11.097 sec) +Saved checkpoint after 146 epoch(s) to ../data/resnet20/checkpoints/00146... +INFO:tensorflow:global_step/sec: 8.75143 +INFO:tensorflow:step = 57101, loss = 0.21456, precision = 0.960938 (11.427 sec) +INFO:tensorflow:global_step/sec: 9.00002 +INFO:tensorflow:step = 57201, loss = 0.150277, precision = 1.0 (11.111 sec) +INFO:tensorflow:global_step/sec: 9.01173 +INFO:tensorflow:step = 57301, loss = 0.184113, precision = 0.992188 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.00788 +INFO:tensorflow:step = 57401, loss = 0.172192, precision = 0.984375 (11.101 sec) +Saved checkpoint after 147 epoch(s) to ../data/resnet20/checkpoints/00147... +INFO:tensorflow:global_step/sec: 8.74821 +INFO:tensorflow:step = 57501, loss = 0.17664, precision = 0.992188 (11.431 sec) +INFO:tensorflow:global_step/sec: 9.00485 +INFO:tensorflow:step = 57601, loss = 0.156077, precision = 1.0 (11.105 sec) +INFO:tensorflow:global_step/sec: 9.00401 +INFO:tensorflow:step = 57701, loss = 0.147273, precision = 1.0 (11.106 sec) +INFO:tensorflow:global_step/sec: 9.00034 +INFO:tensorflow:step = 57801, loss = 0.169329, precision = 1.0 (11.111 sec) +Saved checkpoint after 148 epoch(s) to ../data/resnet20/checkpoints/00148... +INFO:tensorflow:global_step/sec: 8.74495 +INFO:tensorflow:step = 57901, loss = 0.174517, precision = 0.992188 (11.435 sec) +INFO:tensorflow:global_step/sec: 9.01644 +INFO:tensorflow:step = 58001, loss = 0.166003, precision = 0.992188 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.01726 +INFO:tensorflow:step = 58101, loss = 0.142055, precision = 1.0 (11.090 sec) +INFO:tensorflow:global_step/sec: 9.01592 +INFO:tensorflow:step = 58201, loss = 0.151138, precision = 1.0 (11.092 sec) +Saved checkpoint after 149 epoch(s) to ../data/resnet20/checkpoints/00149... +INFO:tensorflow:global_step/sec: 8.75115 +INFO:tensorflow:step = 58301, loss = 0.164182, precision = 0.992188 (11.427 sec) +INFO:tensorflow:global_step/sec: 9.01716 +INFO:tensorflow:step = 58401, loss = 0.149472, precision = 1.0 (11.090 sec) +INFO:tensorflow:global_step/sec: 9.0045 +INFO:tensorflow:step = 58501, loss = 0.164995, precision = 0.984375 (11.106 sec) +INFO:tensorflow:global_step/sec: 9.01174 +INFO:tensorflow:step = 58601, loss = 0.197323, precision = 0.96875 (11.097 sec) +Saved checkpoint after 150 epoch(s) to ../data/resnet20/checkpoints/00150... +INFO:tensorflow:global_step/sec: 8.74899 +INFO:tensorflow:step = 58701, loss = 0.139994, precision = 1.0 (11.430 sec) +INFO:tensorflow:global_step/sec: 9.01025 +INFO:tensorflow:step = 58801, loss = 0.145577, precision = 1.0 (11.098 sec) +INFO:tensorflow:global_step/sec: 9.00014 +INFO:tensorflow:step = 58901, loss = 0.187565, precision = 0.984375 (11.111 sec) +INFO:tensorflow:global_step/sec: 9.00781 +INFO:tensorflow:step = 59001, loss = 0.17021, precision = 0.984375 (11.102 sec) +Saved checkpoint after 151 epoch(s) to ../data/resnet20/checkpoints/00151... +INFO:tensorflow:global_step/sec: 8.75533 +INFO:tensorflow:step = 59101, loss = 0.180391, precision = 0.984375 (11.422 sec) +INFO:tensorflow:global_step/sec: 9.01002 +INFO:tensorflow:step = 59201, loss = 0.157191, precision = 0.992188 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.01485 +INFO:tensorflow:step = 59301, loss = 0.169206, precision = 0.984375 (11.093 sec) +INFO:tensorflow:global_step/sec: 9.01767 +INFO:tensorflow:step = 59401, loss = 0.187292, precision = 0.984375 (11.089 sec) +Saved checkpoint after 152 epoch(s) to ../data/resnet20/checkpoints/00152... +INFO:tensorflow:global_step/sec: 8.70853 +INFO:tensorflow:step = 59501, loss = 0.151279, precision = 0.992188 (11.483 sec) +INFO:tensorflow:global_step/sec: 9.00849 +INFO:tensorflow:step = 59601, loss = 0.172783, precision = 0.984375 (11.101 sec) +INFO:tensorflow:global_step/sec: 9.00193 +INFO:tensorflow:step = 59701, loss = 0.151812, precision = 1.0 (11.109 sec) +INFO:tensorflow:global_step/sec: 9.01903 +INFO:tensorflow:step = 59801, loss = 0.155974, precision = 0.992188 (11.088 sec) +Saved checkpoint after 153 epoch(s) to ../data/resnet20/checkpoints/00153... +INFO:tensorflow:global_step/sec: 8.75185 +INFO:tensorflow:step = 59901, loss = 0.157641, precision = 0.992188 (11.426 sec) +INFO:tensorflow:global_step/sec: 9.01651 +INFO:tensorflow:step = 60001, loss = 0.165674, precision = 0.992188 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.01169 +INFO:tensorflow:step = 60101, loss = 0.193482, precision = 0.984375 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.0135 +INFO:tensorflow:step = 60201, loss = 0.15571, precision = 0.992188 (11.094 sec) +Saved checkpoint after 154 epoch(s) to ../data/resnet20/checkpoints/00154... +INFO:tensorflow:global_step/sec: 8.73681 +INFO:tensorflow:step = 60301, loss = 0.165418, precision = 0.992188 (11.446 sec) +INFO:tensorflow:global_step/sec: 9.00942 +INFO:tensorflow:step = 60401, loss = 0.145432, precision = 0.992188 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.01338 +INFO:tensorflow:step = 60501, loss = 0.168494, precision = 0.976562 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.01399 +INFO:tensorflow:step = 60601, loss = 0.159389, precision = 1.0 (11.094 sec) +Saved checkpoint after 155 epoch(s) to ../data/resnet20/checkpoints/00155... +INFO:tensorflow:global_step/sec: 8.74791 +INFO:tensorflow:step = 60701, loss = 0.164596, precision = 1.0 (11.431 sec) +INFO:tensorflow:global_step/sec: 9.01219 +INFO:tensorflow:step = 60801, loss = 0.14627, precision = 0.992188 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.01204 +INFO:tensorflow:step = 60901, loss = 0.196722, precision = 0.984375 (11.096 sec) +Saved checkpoint after 156 epoch(s) to ../data/resnet20/checkpoints/00156... +INFO:tensorflow:global_step/sec: 8.75347 +INFO:tensorflow:step = 61001, loss = 0.160638, precision = 0.984375 (11.424 sec) +INFO:tensorflow:global_step/sec: 9.01371 +INFO:tensorflow:step = 61101, loss = 0.144823, precision = 1.0 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.01201 +INFO:tensorflow:step = 61201, loss = 0.148998, precision = 0.992188 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.02068 +INFO:tensorflow:step = 61301, loss = 0.138945, precision = 1.0 (11.086 sec) +Saved checkpoint after 157 epoch(s) to ../data/resnet20/checkpoints/00157... +INFO:tensorflow:global_step/sec: 8.74052 +INFO:tensorflow:step = 61401, loss = 0.153152, precision = 1.0 (11.441 sec) +INFO:tensorflow:global_step/sec: 9.00513 +INFO:tensorflow:step = 61501, loss = 0.14752, precision = 1.0 (11.105 sec) +INFO:tensorflow:global_step/sec: 9.0121 +INFO:tensorflow:step = 61601, loss = 0.170335, precision = 0.984375 (11.096 sec) +INFO:tensorflow:global_step/sec: 8.99864 +INFO:tensorflow:step = 61701, loss = 0.162012, precision = 0.984375 (11.113 sec) +Saved checkpoint after 158 epoch(s) to ../data/resnet20/checkpoints/00158... +INFO:tensorflow:global_step/sec: 8.7512 +INFO:tensorflow:step = 61801, loss = 0.155955, precision = 1.0 (11.427 sec) +INFO:tensorflow:global_step/sec: 9.01989 +INFO:tensorflow:step = 61901, loss = 0.169432, precision = 0.992188 (11.086 sec) +INFO:tensorflow:global_step/sec: 9.01306 +INFO:tensorflow:step = 62001, loss = 0.192641, precision = 0.976562 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.02499 +INFO:tensorflow:step = 62101, loss = 0.140117, precision = 1.0 (11.080 sec) +Saved checkpoint after 159 epoch(s) to ../data/resnet20/checkpoints/00159... +INFO:tensorflow:global_step/sec: 8.756 +INFO:tensorflow:step = 62201, loss = 0.137998, precision = 1.0 (11.421 sec) +INFO:tensorflow:global_step/sec: 9.00968 +INFO:tensorflow:step = 62301, loss = 0.180306, precision = 0.976562 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.01436 +INFO:tensorflow:step = 62401, loss = 0.152066, precision = 1.0 (11.093 sec) +INFO:tensorflow:global_step/sec: 9.00459 +INFO:tensorflow:step = 62501, loss = 0.142707, precision = 1.0 (11.105 sec) +Saved checkpoint after 160 epoch(s) to ../data/resnet20/checkpoints/00160... +INFO:tensorflow:global_step/sec: 8.74904 +INFO:tensorflow:step = 62601, loss = 0.148146, precision = 0.992188 (11.430 sec) +INFO:tensorflow:global_step/sec: 9.01295 +INFO:tensorflow:step = 62701, loss = 0.158771, precision = 0.984375 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.01739 +INFO:tensorflow:step = 62801, loss = 0.152611, precision = 1.0 (11.090 sec) +INFO:tensorflow:global_step/sec: 9.0241 +INFO:tensorflow:step = 62901, loss = 0.164142, precision = 0.984375 (11.082 sec) +Saved checkpoint after 161 epoch(s) to ../data/resnet20/checkpoints/00161... +INFO:tensorflow:global_step/sec: 8.73648 +INFO:tensorflow:step = 63001, loss = 0.137607, precision = 1.0 (11.446 sec) +INFO:tensorflow:global_step/sec: 9.01186 +INFO:tensorflow:step = 63101, loss = 0.141982, precision = 1.0 (11.096 sec) +INFO:tensorflow:global_step/sec: 9.01117 +INFO:tensorflow:step = 63201, loss = 0.177779, precision = 0.976562 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.01377 +INFO:tensorflow:step = 63301, loss = 0.170612, precision = 0.984375 (11.094 sec) +Saved checkpoint after 162 epoch(s) to ../data/resnet20/checkpoints/00162... +INFO:tensorflow:global_step/sec: 8.74951 +INFO:tensorflow:step = 63401, loss = 0.147997, precision = 1.0 (11.429 sec) +INFO:tensorflow:global_step/sec: 9.00408 +INFO:tensorflow:step = 63501, loss = 0.14821, precision = 1.0 (11.106 sec) +INFO:tensorflow:global_step/sec: 9.01532 +INFO:tensorflow:step = 63601, loss = 0.167541, precision = 0.984375 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.00859 +INFO:tensorflow:step = 63701, loss = 0.138984, precision = 1.0 (11.100 sec) +Saved checkpoint after 163 epoch(s) to ../data/resnet20/checkpoints/00163... +INFO:tensorflow:global_step/sec: 8.74924 +INFO:tensorflow:step = 63801, loss = 0.148629, precision = 0.992188 (11.430 sec) +INFO:tensorflow:global_step/sec: 9.00719 +INFO:tensorflow:step = 63901, loss = 0.132807, precision = 1.0 (11.102 sec) +INFO:tensorflow:global_step/sec: 9.01757 +INFO:tensorflow:step = 64001, loss = 0.178735, precision = 0.976562 (11.089 sec) +INFO:tensorflow:global_step/sec: 9.01987 +INFO:tensorflow:step = 64101, loss = 0.168593, precision = 0.992188 (11.086 sec) +Saved checkpoint after 164 epoch(s) to ../data/resnet20/checkpoints/00164... +INFO:tensorflow:global_step/sec: 8.73638 +INFO:tensorflow:step = 64201, loss = 0.154234, precision = 0.992188 (11.447 sec) +INFO:tensorflow:global_step/sec: 9.02075 +INFO:tensorflow:step = 64301, loss = 0.150639, precision = 0.992188 (11.085 sec) +INFO:tensorflow:global_step/sec: 9.01177 +INFO:tensorflow:step = 64401, loss = 0.144318, precision = 0.992188 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.01842 +INFO:tensorflow:step = 64501, loss = 0.156943, precision = 0.992188 (11.089 sec) +Saved checkpoint after 165 epoch(s) to ../data/resnet20/checkpoints/00165... +INFO:tensorflow:global_step/sec: 8.75286 +INFO:tensorflow:step = 64601, loss = 0.152557, precision = 1.0 (11.425 sec) +INFO:tensorflow:global_step/sec: 9.01562 +INFO:tensorflow:step = 64701, loss = 0.169628, precision = 0.992188 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.01312 +INFO:tensorflow:step = 64801, loss = 0.156913, precision = 0.992188 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.01258 +INFO:tensorflow:step = 64901, loss = 0.14049, precision = 1.0 (11.096 sec) +Saved checkpoint after 166 epoch(s) to ../data/resnet20/checkpoints/00166... +INFO:tensorflow:global_step/sec: 8.73672 +INFO:tensorflow:step = 65001, loss = 0.13466, precision = 1.0 (11.446 sec) +INFO:tensorflow:global_step/sec: 9.01408 +INFO:tensorflow:step = 65101, loss = 0.165865, precision = 0.992188 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.01657 +INFO:tensorflow:step = 65201, loss = 0.157933, precision = 0.992188 (11.091 sec) +Saved checkpoint after 167 epoch(s) to ../data/resnet20/checkpoints/00167... +INFO:tensorflow:global_step/sec: 8.75517 +INFO:tensorflow:step = 65301, loss = 0.146613, precision = 0.992188 (11.422 sec) +INFO:tensorflow:global_step/sec: 9.02466 +INFO:tensorflow:step = 65401, loss = 0.134488, precision = 1.0 (11.081 sec) +INFO:tensorflow:global_step/sec: 9.01166 +INFO:tensorflow:step = 65501, loss = 0.143888, precision = 0.992188 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.00935 +INFO:tensorflow:step = 65601, loss = 0.167562, precision = 0.984375 (11.099 sec) +Saved checkpoint after 168 epoch(s) to ../data/resnet20/checkpoints/00168... +INFO:tensorflow:global_step/sec: 8.74757 +INFO:tensorflow:step = 65701, loss = 0.144126, precision = 1.0 (11.432 sec) +INFO:tensorflow:global_step/sec: 9.00525 +INFO:tensorflow:step = 65801, loss = 0.176463, precision = 0.976562 (11.105 sec) +INFO:tensorflow:global_step/sec: 8.99879 +INFO:tensorflow:step = 65901, loss = 0.154245, precision = 0.992188 (11.113 sec) +INFO:tensorflow:global_step/sec: 9.01522 +INFO:tensorflow:step = 66001, loss = 0.134072, precision = 1.0 (11.092 sec) +Saved checkpoint after 169 epoch(s) to ../data/resnet20/checkpoints/00169... +INFO:tensorflow:global_step/sec: 8.73667 +INFO:tensorflow:step = 66101, loss = 0.136621, precision = 1.0 (11.446 sec) +INFO:tensorflow:global_step/sec: 9.00837 +INFO:tensorflow:step = 66201, loss = 0.138597, precision = 1.0 (11.101 sec) +INFO:tensorflow:global_step/sec: 9.01389 +INFO:tensorflow:step = 66301, loss = 0.148408, precision = 0.992188 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.00756 +INFO:tensorflow:step = 66401, loss = 0.142596, precision = 1.0 (11.102 sec) +Saved checkpoint after 170 epoch(s) to ../data/resnet20/checkpoints/00170... +INFO:tensorflow:global_step/sec: 8.70881 +INFO:tensorflow:step = 66501, loss = 0.126363, precision = 1.0 (11.483 sec) +INFO:tensorflow:global_step/sec: 9.00059 +INFO:tensorflow:step = 66601, loss = 0.167514, precision = 0.984375 (11.110 sec) +INFO:tensorflow:global_step/sec: 9.01867 +INFO:tensorflow:step = 66701, loss = 0.172163, precision = 0.992188 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.01357 +INFO:tensorflow:step = 66801, loss = 0.133803, precision = 1.0 (11.094 sec) +Saved checkpoint after 171 epoch(s) to ../data/resnet20/checkpoints/00171... +INFO:tensorflow:global_step/sec: 8.75405 +INFO:tensorflow:step = 66901, loss = 0.133399, precision = 1.0 (11.423 sec) +INFO:tensorflow:global_step/sec: 9.01308 +INFO:tensorflow:step = 67001, loss = 0.153717, precision = 0.992188 (11.095 sec) +INFO:tensorflow:global_step/sec: 9.00779 +INFO:tensorflow:step = 67101, loss = 0.132468, precision = 1.0 (11.102 sec) +INFO:tensorflow:global_step/sec: 9.01331 +INFO:tensorflow:step = 67201, loss = 0.153338, precision = 1.0 (11.095 sec) +Saved checkpoint after 172 epoch(s) to ../data/resnet20/checkpoints/00172... +INFO:tensorflow:global_step/sec: 8.74175 +INFO:tensorflow:step = 67301, loss = 0.139698, precision = 1.0 (11.439 sec) +INFO:tensorflow:global_step/sec: 9.02 +INFO:tensorflow:step = 67401, loss = 0.160688, precision = 0.976562 (11.086 sec) +INFO:tensorflow:global_step/sec: 9.00696 +INFO:tensorflow:step = 67501, loss = 0.130813, precision = 1.0 (11.103 sec) +INFO:tensorflow:global_step/sec: 9.01744 +INFO:tensorflow:step = 67601, loss = 0.14963, precision = 0.992188 (11.090 sec) +Saved checkpoint after 173 epoch(s) to ../data/resnet20/checkpoints/00173... +INFO:tensorflow:global_step/sec: 8.75123 +INFO:tensorflow:step = 67701, loss = 0.136603, precision = 1.0 (11.427 sec) +INFO:tensorflow:global_step/sec: 9.01868 +INFO:tensorflow:step = 67801, loss = 0.146449, precision = 0.992188 (11.088 sec) +INFO:tensorflow:global_step/sec: 9.01131 +INFO:tensorflow:step = 67901, loss = 0.165587, precision = 0.984375 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.01705 +INFO:tensorflow:step = 68001, loss = 0.141451, precision = 0.992188 (11.090 sec) +Saved checkpoint after 174 epoch(s) to ../data/resnet20/checkpoints/00174... +INFO:tensorflow:global_step/sec: 8.74982 +INFO:tensorflow:step = 68101, loss = 0.176943, precision = 0.984375 (11.429 sec) +INFO:tensorflow:global_step/sec: 9.00867 +INFO:tensorflow:step = 68201, loss = 0.138215, precision = 0.992188 (11.100 sec) +INFO:tensorflow:global_step/sec: 9.01341 +INFO:tensorflow:step = 68301, loss = 0.139693, precision = 1.0 (11.094 sec) +INFO:tensorflow:global_step/sec: 9.00072 +INFO:tensorflow:step = 68401, loss = 0.132229, precision = 1.0 (11.110 sec) +Saved checkpoint after 175 epoch(s) to ../data/resnet20/checkpoints/00175... +INFO:tensorflow:global_step/sec: 8.75364 +INFO:tensorflow:step = 68501, loss = 0.159547, precision = 0.976562 (11.424 sec) +INFO:tensorflow:global_step/sec: 9.01584 +INFO:tensorflow:step = 68601, loss = 0.144231, precision = 0.984375 (11.092 sec) +INFO:tensorflow:global_step/sec: 9.00967 +INFO:tensorflow:step = 68701, loss = 0.145858, precision = 0.992188 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.0036 +INFO:tensorflow:step = 68801, loss = 0.152583, precision = 0.984375 (11.106 sec) +Saved checkpoint after 176 epoch(s) to ../data/resnet20/checkpoints/00176... +INFO:tensorflow:global_step/sec: 8.74745 +INFO:tensorflow:step = 68901, loss = 0.150114, precision = 0.992188 (11.432 sec) +INFO:tensorflow:global_step/sec: 9.02654 +INFO:tensorflow:step = 69001, loss = 0.175582, precision = 0.984375 (11.078 sec) +INFO:tensorflow:global_step/sec: 9.02482 +INFO:tensorflow:step = 69101, loss = 0.137297, precision = 1.0 (11.080 sec) +INFO:tensorflow:global_step/sec: 9.0268 +INFO:tensorflow:step = 69201, loss = 0.138794, precision = 0.992188 (11.078 sec) +Saved checkpoint after 177 epoch(s) to ../data/resnet20/checkpoints/00177... +INFO:tensorflow:global_step/sec: 8.73673 +INFO:tensorflow:step = 69301, loss = 0.141637, precision = 0.992188 (11.446 sec) +INFO:tensorflow:global_step/sec: 9.01154 +INFO:tensorflow:step = 69401, loss = 0.147266, precision = 0.984375 (11.097 sec) +INFO:tensorflow:global_step/sec: 9.01131 +INFO:tensorflow:step = 69501, loss = 0.142301, precision = 0.992188 (11.097 sec) +Saved checkpoint after 178 epoch(s) to ../data/resnet20/checkpoints/00178... +INFO:tensorflow:global_step/sec: 8.73453 +INFO:tensorflow:step = 69601, loss = 0.14074, precision = 1.0 (11.449 sec) +INFO:tensorflow:global_step/sec: 9.00214 +INFO:tensorflow:step = 69701, loss = 0.154856, precision = 0.984375 (11.108 sec) +INFO:tensorflow:global_step/sec: 9.01629 +INFO:tensorflow:step = 69801, loss = 0.146389, precision = 0.992188 (11.091 sec) +INFO:tensorflow:global_step/sec: 9.00597 +INFO:tensorflow:step = 69901, loss = 0.1417, precision = 0.992188 (11.104 sec) +Saved checkpoint after 179 epoch(s) to ../data/resnet20/checkpoints/00179... +INFO:tensorflow:global_step/sec: 8.74997 +INFO:tensorflow:step = 70001, loss = 0.150674, precision = 0.992188 (11.429 sec) +INFO:tensorflow:global_step/sec: 9.01481 +INFO:tensorflow:step = 70101, loss = 0.134261, precision = 1.0 (11.093 sec) +INFO:tensorflow:global_step/sec: 9.00235 +INFO:tensorflow:step = 70201, loss = 0.128187, precision = 1.0 (11.108 sec) +INFO:tensorflow:global_step/sec: 9.0152 +INFO:tensorflow:step = 70301, loss = 0.125868, precision = 1.0 (11.093 sec) +Saved checkpoint after 180 epoch(s) to ../data/resnet20/checkpoints/00180... +INFO:tensorflow:global_step/sec: 8.75492 +INFO:tensorflow:step = 70401, loss = 0.138747, precision = 0.992188 (11.422 sec) +INFO:tensorflow:global_step/sec: 9.00992 +INFO:tensorflow:step = 70501, loss = 0.154714, precision = 0.984375 (11.099 sec) +INFO:tensorflow:global_step/sec: 9.01063 +INFO:tensorflow:step = 70601, loss = 0.129542, precision = 1.0 (11.098 sec) +INFO:tensorflow:global_step/sec: 9.02483 +INFO:tensorflow:step = 70701, loss = 0.140172, precision = 1.0 (11.081 sec) +Saved checkpoint after 181 epoch(s) to ../data/resnet20/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1k80_gc/resnet56_train.log b/tensorflow/CIFAR10/logs/1k80_gc/resnet56_train.log new file mode 100644 index 0000000..26bf580 --- /dev/null +++ b/tensorflow/CIFAR10/logs/1k80_gc/resnet56_train.log @@ -0,0 +1,1836 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/850.99k params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_3/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_4/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_5/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_6/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_7/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_8/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_3/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_4/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_5/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_6/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_7/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_8/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_3/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_4/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_5/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_6/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_7/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_8/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/32.12b flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + +======================End of Report========================== +2017-07-30 00:29:40.434704: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero +2017-07-30 00:29:40.435514: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla K80 +major: 3 minor: 7 memoryClockRate (GHz) 0.562 +pciBusID 0000:00:04.0 +Total memory: 11.17GiB +Free memory: 11.09GiB +2017-07-30 00:29:40.435546: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-07-30 00:29:40.435558: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-07-30 00:29:40.435571: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:00:04.0) +2017-07-30 00:29:41.016506: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 00:29:41.016567: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 8 visible devices +2017-07-30 00:29:41.021213: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x63994e0 executing computations on platform Host. Devices: +2017-07-30 00:29:41.021253: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-07-30 00:29:41.022122: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-07-30 00:29:41.022156: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 8 visible devices +2017-07-30 00:29:41.023223: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x6290f10 executing computations on platform CUDA. Devices: +2017-07-30 00:29:41.023250: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla K80, Compute Capability 3.7 +INFO:tensorflow:step = 1, loss = 3.35626, precision = 0.0859375 +INFO:tensorflow:global_step/sec: 3.18669 +INFO:tensorflow:step = 101, loss = 2.6409, precision = 0.320312 (31.381 sec) +INFO:tensorflow:global_step/sec: 3.23395 +INFO:tensorflow:step = 201, loss = 2.58921, precision = 0.375 (30.922 sec) +INFO:tensorflow:global_step/sec: 3.23671 +INFO:tensorflow:step = 301, loss = 2.41665, precision = 0.4375 (30.896 sec) +total_params: 850986 +Saved checkpoint after 1 epoch(s) to ../data/resnet56/checkpoints/00001... +INFO:tensorflow:global_step/sec: 3.14668 +INFO:tensorflow:step = 401, loss = 2.67574, precision = 0.335938 (31.780 sec) +INFO:tensorflow:global_step/sec: 3.23981 +INFO:tensorflow:step = 501, loss = 2.25297, precision = 0.53125 (30.866 sec) +INFO:tensorflow:global_step/sec: 3.23217 +INFO:tensorflow:step = 601, loss = 1.97639, precision = 0.625 (30.939 sec) +INFO:tensorflow:global_step/sec: 3.23485 +INFO:tensorflow:step = 701, loss = 1.93673, precision = 0.546875 (30.913 sec) +Saved checkpoint after 2 epoch(s) to ../data/resnet56/checkpoints/00002... +INFO:tensorflow:global_step/sec: 3.15791 +INFO:tensorflow:step = 801, loss = 1.89409, precision = 0.59375 (31.667 sec) +INFO:tensorflow:global_step/sec: 3.23516 +INFO:tensorflow:step = 901, loss = 1.53702, precision = 0.671875 (30.910 sec) +INFO:tensorflow:global_step/sec: 3.23004 +INFO:tensorflow:step = 1001, loss = 1.54945, precision = 0.648438 (30.960 sec) +INFO:tensorflow:global_step/sec: 3.2343 +INFO:tensorflow:step = 1101, loss = 1.50841, precision = 0.726562 (30.919 sec) +Saved checkpoint after 3 epoch(s) to ../data/resnet56/checkpoints/00003... +INFO:tensorflow:global_step/sec: 3.15175 +INFO:tensorflow:step = 1201, loss = 1.38049, precision = 0.710938 (31.728 sec) +INFO:tensorflow:global_step/sec: 3.2296 +INFO:tensorflow:step = 1301, loss = 1.42572, precision = 0.695312 (30.963 sec) +INFO:tensorflow:global_step/sec: 3.23494 +INFO:tensorflow:step = 1401, loss = 1.19575, precision = 0.773438 (30.913 sec) +INFO:tensorflow:global_step/sec: 3.23198 +INFO:tensorflow:step = 1501, loss = 1.19997, precision = 0.71875 (30.941 sec) +Saved checkpoint after 4 epoch(s) to ../data/resnet56/checkpoints/00004... +INFO:tensorflow:global_step/sec: 3.1548 +INFO:tensorflow:step = 1601, loss = 1.23073, precision = 0.734375 (31.698 sec) +INFO:tensorflow:global_step/sec: 3.23107 +INFO:tensorflow:step = 1701, loss = 1.15108, precision = 0.757812 (30.949 sec) +INFO:tensorflow:global_step/sec: 3.23011 +INFO:tensorflow:step = 1801, loss = 1.13718, precision = 0.757812 (30.959 sec) +INFO:tensorflow:global_step/sec: 3.23483 +INFO:tensorflow:step = 1901, loss = 1.15922, precision = 0.726562 (30.914 sec) +Saved checkpoint after 5 epoch(s) to ../data/resnet56/checkpoints/00005... +INFO:tensorflow:global_step/sec: 3.15687 +INFO:tensorflow:step = 2001, loss = 0.949058, precision = 0.84375 (31.677 sec) +INFO:tensorflow:global_step/sec: 3.23072 +INFO:tensorflow:step = 2101, loss = 1.07618, precision = 0.726562 (30.953 sec) +INFO:tensorflow:global_step/sec: 3.23016 +INFO:tensorflow:step = 2201, loss = 0.889877, precision = 0.796875 (30.958 sec) +INFO:tensorflow:global_step/sec: 3.23278 +INFO:tensorflow:step = 2301, loss = 0.892277, precision = 0.828125 (30.933 sec) +Saved checkpoint after 6 epoch(s) to ../data/resnet56/checkpoints/00006... +INFO:tensorflow:global_step/sec: 3.14318 +INFO:tensorflow:step = 2401, loss = 0.825054, precision = 0.84375 (31.815 sec) +INFO:tensorflow:global_step/sec: 3.22971 +INFO:tensorflow:step = 2501, loss = 0.950418, precision = 0.765625 (30.962 sec) +INFO:tensorflow:global_step/sec: 3.22907 +INFO:tensorflow:step = 2601, loss = 0.910684, precision = 0.773438 (30.969 sec) +INFO:tensorflow:global_step/sec: 3.23491 +INFO:tensorflow:step = 2701, loss = 0.947331, precision = 0.773438 (30.913 sec) +Saved checkpoint after 7 epoch(s) to ../data/resnet56/checkpoints/00007... +INFO:tensorflow:global_step/sec: 3.1527 +INFO:tensorflow:step = 2801, loss = 0.944863, precision = 0.765625 (31.720 sec) +INFO:tensorflow:global_step/sec: 3.2315 +INFO:tensorflow:step = 2901, loss = 0.846113, precision = 0.789062 (30.945 sec) +INFO:tensorflow:global_step/sec: 3.23074 +INFO:tensorflow:step = 3001, loss = 0.809746, precision = 0.828125 (30.953 sec) +INFO:tensorflow:global_step/sec: 3.23154 +INFO:tensorflow:step = 3101, loss = 1.01018, precision = 0.765625 (30.945 sec) +Saved checkpoint after 8 epoch(s) to ../data/resnet56/checkpoints/00008... +INFO:tensorflow:global_step/sec: 3.1482 +INFO:tensorflow:step = 3201, loss = 0.914783, precision = 0.773438 (31.764 sec) +INFO:tensorflow:global_step/sec: 3.23247 +INFO:tensorflow:step = 3301, loss = 0.812369, precision = 0.804688 (30.936 sec) +INFO:tensorflow:global_step/sec: 3.23346 +INFO:tensorflow:step = 3401, loss = 0.752785, precision = 0.890625 (30.926 sec) +INFO:tensorflow:global_step/sec: 3.23257 +INFO:tensorflow:step = 3501, loss = 0.883095, precision = 0.8125 (30.935 sec) +Saved checkpoint after 9 epoch(s) to ../data/resnet56/checkpoints/00009... +INFO:tensorflow:global_step/sec: 3.15099 +INFO:tensorflow:step = 3601, loss = 0.855998, precision = 0.757812 (31.737 sec) +INFO:tensorflow:global_step/sec: 3.23172 +INFO:tensorflow:step = 3701, loss = 0.90322, precision = 0.742188 (30.943 sec) +INFO:tensorflow:global_step/sec: 3.23442 +INFO:tensorflow:step = 3801, loss = 0.804968, precision = 0.835938 (30.917 sec) +INFO:tensorflow:global_step/sec: 3.23813 +INFO:tensorflow:step = 3901, loss = 0.980718, precision = 0.757812 (30.882 sec) +Saved checkpoint after 10 epoch(s) to ../data/resnet56/checkpoints/00010... +INFO:tensorflow:global_step/sec: 3.15718 +INFO:tensorflow:step = 4001, loss = 0.683121, precision = 0.867188 (31.674 sec) +INFO:tensorflow:global_step/sec: 3.23007 +INFO:tensorflow:step = 4101, loss = 0.710009, precision = 0.84375 (30.959 sec) +INFO:tensorflow:global_step/sec: 3.23134 +INFO:tensorflow:step = 4201, loss = 0.637894, precision = 0.875 (30.947 sec) +Saved checkpoint after 11 epoch(s) to ../data/resnet56/checkpoints/00011... +INFO:tensorflow:global_step/sec: 3.15436 +INFO:tensorflow:step = 4301, loss = 0.661044, precision = 0.851562 (31.702 sec) +INFO:tensorflow:global_step/sec: 3.23035 +INFO:tensorflow:step = 4401, loss = 0.743781, precision = 0.820312 (30.956 sec) +INFO:tensorflow:global_step/sec: 3.23139 +INFO:tensorflow:step = 4501, loss = 0.719936, precision = 0.851562 (30.946 sec) +INFO:tensorflow:global_step/sec: 3.23174 +INFO:tensorflow:step = 4601, loss = 0.858645, precision = 0.804688 (30.943 sec) +Saved checkpoint after 12 epoch(s) to ../data/resnet56/checkpoints/00012... +INFO:tensorflow:global_step/sec: 3.15305 +INFO:tensorflow:step = 4701, loss = 0.693362, precision = 0.875 (31.716 sec) +INFO:tensorflow:global_step/sec: 3.23021 +INFO:tensorflow:step = 4801, loss = 0.806446, precision = 0.804688 (30.957 sec) +INFO:tensorflow:global_step/sec: 3.23137 +INFO:tensorflow:step = 4901, loss = 0.748623, precision = 0.804688 (30.947 sec) +INFO:tensorflow:global_step/sec: 3.23623 +INFO:tensorflow:step = 5001, loss = 0.808611, precision = 0.820312 (30.900 sec) +Saved checkpoint after 13 epoch(s) to ../data/resnet56/checkpoints/00013... +INFO:tensorflow:global_step/sec: 3.15619 +INFO:tensorflow:step = 5101, loss = 0.850106, precision = 0.789062 (31.684 sec) +INFO:tensorflow:global_step/sec: 3.23409 +INFO:tensorflow:step = 5201, loss = 0.816003, precision = 0.828125 (30.920 sec) +INFO:tensorflow:global_step/sec: 3.23288 +INFO:tensorflow:step = 5301, loss = 0.697633, precision = 0.84375 (30.932 sec) +INFO:tensorflow:global_step/sec: 3.23344 +INFO:tensorflow:step = 5401, loss = 0.800693, precision = 0.820312 (30.927 sec) +Saved checkpoint after 14 epoch(s) to ../data/resnet56/checkpoints/00014... +INFO:tensorflow:global_step/sec: 3.15606 +INFO:tensorflow:step = 5501, loss = 0.747658, precision = 0.820312 (31.685 sec) +INFO:tensorflow:global_step/sec: 3.23243 +INFO:tensorflow:step = 5601, loss = 0.817077, precision = 0.820312 (30.936 sec) +INFO:tensorflow:global_step/sec: 3.23332 +INFO:tensorflow:step = 5701, loss = 0.62075, precision = 0.867188 (30.928 sec) +INFO:tensorflow:global_step/sec: 3.23508 +INFO:tensorflow:step = 5801, loss = 0.892585, precision = 0.757812 (30.911 sec) +Saved checkpoint after 15 epoch(s) to ../data/resnet56/checkpoints/00015... +INFO:tensorflow:global_step/sec: 3.15414 +INFO:tensorflow:step = 5901, loss = 0.682428, precision = 0.828125 (31.705 sec) +INFO:tensorflow:global_step/sec: 3.23456 +INFO:tensorflow:step = 6001, loss = 0.726659, precision = 0.851562 (30.916 sec) +INFO:tensorflow:global_step/sec: 3.23272 +INFO:tensorflow:step = 6101, loss = 0.629095, precision = 0.867188 (30.933 sec) +INFO:tensorflow:global_step/sec: 3.23671 +INFO:tensorflow:step = 6201, loss = 0.717856, precision = 0.851562 (30.896 sec) +Saved checkpoint after 16 epoch(s) to ../data/resnet56/checkpoints/00016... +INFO:tensorflow:global_step/sec: 3.15758 +INFO:tensorflow:step = 6301, loss = 0.636011, precision = 0.867188 (31.670 sec) +INFO:tensorflow:global_step/sec: 3.23078 +INFO:tensorflow:step = 6401, loss = 0.732979, precision = 0.820312 (30.952 sec) +INFO:tensorflow:global_step/sec: 3.23477 +INFO:tensorflow:step = 6501, loss = 0.69541, precision = 0.828125 (30.914 sec) +INFO:tensorflow:global_step/sec: 3.23737 +INFO:tensorflow:step = 6601, loss = 0.671405, precision = 0.851562 (30.889 sec) +Saved checkpoint after 17 epoch(s) to ../data/resnet56/checkpoints/00017... +INFO:tensorflow:global_step/sec: 3.1585 +INFO:tensorflow:step = 6701, loss = 0.661803, precision = 0.890625 (31.661 sec) +INFO:tensorflow:global_step/sec: 3.23142 +INFO:tensorflow:step = 6801, loss = 0.702098, precision = 0.84375 (30.946 sec) +INFO:tensorflow:global_step/sec: 3.23212 +INFO:tensorflow:step = 6901, loss = 0.906199, precision = 0.8125 (30.939 sec) +INFO:tensorflow:global_step/sec: 3.23316 +INFO:tensorflow:step = 7001, loss = 0.889709, precision = 0.773438 (30.929 sec) +Saved checkpoint after 18 epoch(s) to ../data/resnet56/checkpoints/00018... +INFO:tensorflow:global_step/sec: 3.15645 +INFO:tensorflow:step = 7101, loss = 0.667089, precision = 0.851562 (31.681 sec) +INFO:tensorflow:global_step/sec: 3.23539 +INFO:tensorflow:step = 7201, loss = 0.739311, precision = 0.804688 (30.909 sec) +INFO:tensorflow:global_step/sec: 3.23304 +INFO:tensorflow:step = 7301, loss = 0.876782, precision = 0.804688 (30.930 sec) +INFO:tensorflow:global_step/sec: 3.23249 +INFO:tensorflow:step = 7401, loss = 0.73286, precision = 0.828125 (30.936 sec) +Saved checkpoint after 19 epoch(s) to ../data/resnet56/checkpoints/00019... +INFO:tensorflow:global_step/sec: 3.14923 +INFO:tensorflow:step = 7501, loss = 0.673659, precision = 0.835938 (31.754 sec) +INFO:tensorflow:global_step/sec: 3.23242 +INFO:tensorflow:step = 7601, loss = 0.667104, precision = 0.875 (30.937 sec) +INFO:tensorflow:global_step/sec: 3.23571 +INFO:tensorflow:step = 7701, loss = 0.754114, precision = 0.820312 (30.905 sec) +INFO:tensorflow:global_step/sec: 3.23426 +INFO:tensorflow:step = 7801, loss = 0.618319, precision = 0.890625 (30.919 sec) +Saved checkpoint after 20 epoch(s) to ../data/resnet56/checkpoints/00020... +INFO:tensorflow:global_step/sec: 3.14914 +INFO:tensorflow:step = 7901, loss = 0.737545, precision = 0.84375 (31.755 sec) +INFO:tensorflow:global_step/sec: 3.23356 +INFO:tensorflow:step = 8001, loss = 0.629587, precision = 0.898438 (30.926 sec) +INFO:tensorflow:global_step/sec: 3.23586 +INFO:tensorflow:step = 8101, loss = 0.688766, precision = 0.875 (30.904 sec) +INFO:tensorflow:global_step/sec: 3.2349 +INFO:tensorflow:step = 8201, loss = 0.768978, precision = 0.851562 (30.913 sec) +Saved checkpoint after 21 epoch(s) to ../data/resnet56/checkpoints/00021... +INFO:tensorflow:global_step/sec: 3.14244 +INFO:tensorflow:step = 8301, loss = 0.77372, precision = 0.820312 (31.823 sec) +INFO:tensorflow:global_step/sec: 3.23093 +INFO:tensorflow:step = 8401, loss = 0.691194, precision = 0.914062 (30.951 sec) +INFO:tensorflow:global_step/sec: 3.2336 +INFO:tensorflow:step = 8501, loss = 0.662998, precision = 0.867188 (30.925 sec) +INFO:tensorflow:global_step/sec: 3.23718 +INFO:tensorflow:step = 8601, loss = 0.753373, precision = 0.828125 (30.891 sec) +Saved checkpoint after 22 epoch(s) to ../data/resnet56/checkpoints/00022... +INFO:tensorflow:global_step/sec: 3.14946 +INFO:tensorflow:step = 8701, loss = 0.814094, precision = 0.796875 (31.752 sec) +INFO:tensorflow:global_step/sec: 3.23207 +INFO:tensorflow:step = 8801, loss = 0.705831, precision = 0.835938 (30.940 sec) +INFO:tensorflow:global_step/sec: 3.23551 +INFO:tensorflow:step = 8901, loss = 0.593091, precision = 0.890625 (30.907 sec) +Saved checkpoint after 23 epoch(s) to ../data/resnet56/checkpoints/00023... +INFO:tensorflow:global_step/sec: 3.15242 +INFO:tensorflow:step = 9001, loss = 0.588878, precision = 0.882812 (31.722 sec) +INFO:tensorflow:global_step/sec: 3.23755 +INFO:tensorflow:step = 9101, loss = 0.588432, precision = 0.898438 (30.887 sec) +INFO:tensorflow:global_step/sec: 3.2339 +INFO:tensorflow:step = 9201, loss = 0.780507, precision = 0.851562 (30.922 sec) +INFO:tensorflow:global_step/sec: 3.23268 +INFO:tensorflow:step = 9301, loss = 0.863732, precision = 0.835938 (30.934 sec) +Saved checkpoint after 24 epoch(s) to ../data/resnet56/checkpoints/00024... +INFO:tensorflow:global_step/sec: 3.15394 +INFO:tensorflow:step = 9401, loss = 0.591801, precision = 0.914062 (31.707 sec) +INFO:tensorflow:global_step/sec: 3.23182 +INFO:tensorflow:step = 9501, loss = 0.60676, precision = 0.898438 (30.942 sec) +INFO:tensorflow:global_step/sec: 3.23559 +INFO:tensorflow:step = 9601, loss = 0.76303, precision = 0.804688 (30.906 sec) +INFO:tensorflow:global_step/sec: 3.23643 +INFO:tensorflow:step = 9701, loss = 0.743796, precision = 0.851562 (30.898 sec) +Saved checkpoint after 25 epoch(s) to ../data/resnet56/checkpoints/00025... +INFO:tensorflow:global_step/sec: 3.15008 +INFO:tensorflow:step = 9801, loss = 0.704286, precision = 0.828125 (31.745 sec) +INFO:tensorflow:global_step/sec: 3.2307 +INFO:tensorflow:step = 9901, loss = 0.708626, precision = 0.875 (30.953 sec) +INFO:tensorflow:global_step/sec: 3.23541 +INFO:tensorflow:step = 10001, loss = 0.788165, precision = 0.859375 (30.908 sec) +INFO:tensorflow:global_step/sec: 3.23811 +INFO:tensorflow:step = 10101, loss = 0.67289, precision = 0.882812 (30.882 sec) +Saved checkpoint after 26 epoch(s) to ../data/resnet56/checkpoints/00026... +INFO:tensorflow:global_step/sec: 3.15218 +INFO:tensorflow:step = 10201, loss = 0.606063, precision = 0.882812 (31.724 sec) +INFO:tensorflow:global_step/sec: 3.2346 +INFO:tensorflow:step = 10301, loss = 0.696597, precision = 0.859375 (30.916 sec) +INFO:tensorflow:global_step/sec: 3.23698 +INFO:tensorflow:step = 10401, loss = 0.686375, precision = 0.851562 (30.893 sec) +INFO:tensorflow:global_step/sec: 3.23879 +INFO:tensorflow:step = 10501, loss = 0.630678, precision = 0.882812 (30.876 sec) +Saved checkpoint after 27 epoch(s) to ../data/resnet56/checkpoints/00027... +INFO:tensorflow:global_step/sec: 3.15216 +INFO:tensorflow:step = 10601, loss = 0.636456, precision = 0.875 (31.724 sec) +INFO:tensorflow:global_step/sec: 3.23685 +INFO:tensorflow:step = 10701, loss = 0.71209, precision = 0.867188 (30.894 sec) +INFO:tensorflow:global_step/sec: 3.23633 +INFO:tensorflow:step = 10801, loss = 0.618316, precision = 0.882812 (30.899 sec) +INFO:tensorflow:global_step/sec: 3.23981 +INFO:tensorflow:step = 10901, loss = 0.729042, precision = 0.867188 (30.866 sec) +Saved checkpoint after 28 epoch(s) to ../data/resnet56/checkpoints/00028... +INFO:tensorflow:global_step/sec: 3.15604 +INFO:tensorflow:step = 11001, loss = 0.687836, precision = 0.859375 (31.685 sec) +INFO:tensorflow:global_step/sec: 3.23486 +INFO:tensorflow:step = 11101, loss = 0.733657, precision = 0.820312 (30.913 sec) +INFO:tensorflow:global_step/sec: 3.23905 +INFO:tensorflow:step = 11201, loss = 0.574234, precision = 0.914062 (30.873 sec) +INFO:tensorflow:global_step/sec: 3.2335 +INFO:tensorflow:step = 11301, loss = 0.727152, precision = 0.867188 (30.926 sec) +Saved checkpoint after 29 epoch(s) to ../data/resnet56/checkpoints/00029... +INFO:tensorflow:global_step/sec: 3.14901 +INFO:tensorflow:step = 11401, loss = 0.700464, precision = 0.882812 (31.756 sec) +INFO:tensorflow:global_step/sec: 3.2354 +INFO:tensorflow:step = 11501, loss = 0.658418, precision = 0.867188 (30.908 sec) +INFO:tensorflow:global_step/sec: 3.23705 +INFO:tensorflow:step = 11601, loss = 0.628699, precision = 0.875 (30.893 sec) +INFO:tensorflow:global_step/sec: 3.23459 +INFO:tensorflow:step = 11701, loss = 0.714451, precision = 0.859375 (30.916 sec) +Saved checkpoint after 30 epoch(s) to ../data/resnet56/checkpoints/00030... +INFO:tensorflow:global_step/sec: 3.15304 +INFO:tensorflow:step = 11801, loss = 0.714667, precision = 0.828125 (31.715 sec) +INFO:tensorflow:global_step/sec: 3.23598 +INFO:tensorflow:step = 11901, loss = 0.55919, precision = 0.90625 (30.903 sec) +INFO:tensorflow:global_step/sec: 3.23915 +INFO:tensorflow:step = 12001, loss = 0.584597, precision = 0.90625 (30.873 sec) +INFO:tensorflow:global_step/sec: 3.23529 +INFO:tensorflow:step = 12101, loss = 0.635457, precision = 0.859375 (30.909 sec) +Saved checkpoint after 31 epoch(s) to ../data/resnet56/checkpoints/00031... +INFO:tensorflow:global_step/sec: 3.15269 +INFO:tensorflow:step = 12201, loss = 0.717016, precision = 0.867188 (31.719 sec) +INFO:tensorflow:global_step/sec: 3.23663 +INFO:tensorflow:step = 12301, loss = 0.559691, precision = 0.914062 (30.896 sec) +INFO:tensorflow:global_step/sec: 3.23666 +INFO:tensorflow:step = 12401, loss = 0.670371, precision = 0.851562 (30.896 sec) +INFO:tensorflow:global_step/sec: 3.23603 +INFO:tensorflow:step = 12501, loss = 0.672076, precision = 0.875 (30.902 sec) +Saved checkpoint after 32 epoch(s) to ../data/resnet56/checkpoints/00032... +INFO:tensorflow:global_step/sec: 3.15547 +INFO:tensorflow:step = 12601, loss = 0.583399, precision = 0.890625 (31.691 sec) +INFO:tensorflow:global_step/sec: 3.24198 +INFO:tensorflow:step = 12701, loss = 0.716198, precision = 0.8125 (30.845 sec) +INFO:tensorflow:global_step/sec: 3.24302 +INFO:tensorflow:step = 12801, loss = 0.629108, precision = 0.882812 (30.836 sec) +INFO:tensorflow:global_step/sec: 3.2379 +INFO:tensorflow:step = 12901, loss = 0.540336, precision = 0.914062 (30.884 sec) +Saved checkpoint after 33 epoch(s) to ../data/resnet56/checkpoints/00033... +INFO:tensorflow:global_step/sec: 3.15439 +INFO:tensorflow:step = 13001, loss = 0.574518, precision = 0.882812 (31.702 sec) +INFO:tensorflow:global_step/sec: 3.23715 +INFO:tensorflow:step = 13101, loss = 0.756781, precision = 0.828125 (30.892 sec) +INFO:tensorflow:global_step/sec: 3.23731 +INFO:tensorflow:step = 13201, loss = 0.701908, precision = 0.851562 (30.890 sec) +Saved checkpoint after 34 epoch(s) to ../data/resnet56/checkpoints/00034... +INFO:tensorflow:global_step/sec: 3.14231 +INFO:tensorflow:step = 13301, loss = 0.616909, precision = 0.890625 (31.824 sec) +INFO:tensorflow:global_step/sec: 3.23234 +INFO:tensorflow:step = 13401, loss = 0.663032, precision = 0.851562 (30.937 sec) +INFO:tensorflow:global_step/sec: 3.23685 +INFO:tensorflow:step = 13501, loss = 0.726003, precision = 0.835938 (30.894 sec) +INFO:tensorflow:global_step/sec: 3.23364 +INFO:tensorflow:step = 13601, loss = 0.640403, precision = 0.882812 (30.925 sec) +Saved checkpoint after 35 epoch(s) to ../data/resnet56/checkpoints/00035... +INFO:tensorflow:global_step/sec: 3.15337 +INFO:tensorflow:step = 13701, loss = 0.636706, precision = 0.882812 (31.712 sec) +INFO:tensorflow:global_step/sec: 3.23792 +INFO:tensorflow:step = 13801, loss = 0.731824, precision = 0.820312 (30.884 sec) +INFO:tensorflow:global_step/sec: 3.23832 +INFO:tensorflow:step = 13901, loss = 0.691535, precision = 0.851562 (30.880 sec) +INFO:tensorflow:global_step/sec: 3.2387 +INFO:tensorflow:step = 14001, loss = 0.557708, precision = 0.921875 (30.877 sec) +Saved checkpoint after 36 epoch(s) to ../data/resnet56/checkpoints/00036... +INFO:tensorflow:global_step/sec: 3.15606 +INFO:tensorflow:step = 14101, loss = 0.596632, precision = 0.882812 (31.685 sec) +INFO:tensorflow:global_step/sec: 3.2393 +INFO:tensorflow:step = 14201, loss = 0.702643, precision = 0.859375 (30.871 sec) +INFO:tensorflow:global_step/sec: 3.23796 +INFO:tensorflow:step = 14301, loss = 0.674132, precision = 0.835938 (30.884 sec) +INFO:tensorflow:global_step/sec: 3.23791 +INFO:tensorflow:step = 14401, loss = 0.914108, precision = 0.765625 (30.884 sec) +Saved checkpoint after 37 epoch(s) to ../data/resnet56/checkpoints/00037... +INFO:tensorflow:global_step/sec: 3.15517 +INFO:tensorflow:step = 14501, loss = 0.505436, precision = 0.9375 (31.694 sec) +INFO:tensorflow:global_step/sec: 3.23772 +INFO:tensorflow:step = 14601, loss = 0.66477, precision = 0.90625 (30.886 sec) +INFO:tensorflow:global_step/sec: 3.2377 +INFO:tensorflow:step = 14701, loss = 0.641865, precision = 0.898438 (30.886 sec) +INFO:tensorflow:global_step/sec: 3.23652 +INFO:tensorflow:step = 14801, loss = 0.652834, precision = 0.890625 (30.898 sec) +Saved checkpoint after 38 epoch(s) to ../data/resnet56/checkpoints/00038... +INFO:tensorflow:global_step/sec: 3.15761 +INFO:tensorflow:step = 14901, loss = 0.533776, precision = 0.929688 (31.669 sec) +INFO:tensorflow:global_step/sec: 3.2363 +INFO:tensorflow:step = 15001, loss = 0.638798, precision = 0.890625 (30.900 sec) +INFO:tensorflow:global_step/sec: 3.23673 +INFO:tensorflow:step = 15101, loss = 0.822696, precision = 0.835938 (30.895 sec) +INFO:tensorflow:global_step/sec: 3.23501 +INFO:tensorflow:step = 15201, loss = 0.617717, precision = 0.882812 (30.912 sec) +Saved checkpoint after 39 epoch(s) to ../data/resnet56/checkpoints/00039... +INFO:tensorflow:global_step/sec: 3.15373 +INFO:tensorflow:step = 15301, loss = 0.603889, precision = 0.914062 (31.708 sec) +INFO:tensorflow:global_step/sec: 3.23644 +INFO:tensorflow:step = 15401, loss = 0.7833, precision = 0.867188 (30.898 sec) +INFO:tensorflow:global_step/sec: 3.23741 +INFO:tensorflow:step = 15501, loss = 0.579127, precision = 0.867188 (30.889 sec) +INFO:tensorflow:global_step/sec: 3.23722 +INFO:tensorflow:step = 15601, loss = 0.559307, precision = 0.914062 (30.890 sec) +Saved checkpoint after 40 epoch(s) to ../data/resnet56/checkpoints/00040... +INFO:tensorflow:global_step/sec: 3.15312 +INFO:tensorflow:step = 15701, loss = 0.704686, precision = 0.882812 (31.715 sec) +INFO:tensorflow:global_step/sec: 3.23983 +INFO:tensorflow:step = 15801, loss = 0.500363, precision = 0.921875 (30.866 sec) +INFO:tensorflow:global_step/sec: 3.23922 +INFO:tensorflow:step = 15901, loss = 0.698547, precision = 0.84375 (30.872 sec) +INFO:tensorflow:global_step/sec: 3.23703 +INFO:tensorflow:step = 16001, loss = 0.653773, precision = 0.890625 (30.893 sec) +Saved checkpoint after 41 epoch(s) to ../data/resnet56/checkpoints/00041... +INFO:tensorflow:global_step/sec: 3.15599 +INFO:tensorflow:step = 16101, loss = 0.501936, precision = 0.921875 (31.686 sec) +INFO:tensorflow:global_step/sec: 3.23616 +INFO:tensorflow:step = 16201, loss = 0.68073, precision = 0.867188 (30.901 sec) +INFO:tensorflow:global_step/sec: 3.23665 +INFO:tensorflow:step = 16301, loss = 0.701701, precision = 0.859375 (30.896 sec) +INFO:tensorflow:global_step/sec: 3.23627 +INFO:tensorflow:step = 16401, loss = 0.621525, precision = 0.867188 (30.900 sec) +Saved checkpoint after 42 epoch(s) to ../data/resnet56/checkpoints/00042... +INFO:tensorflow:global_step/sec: 3.15942 +INFO:tensorflow:step = 16501, loss = 0.75846, precision = 0.851562 (31.651 sec) +INFO:tensorflow:global_step/sec: 3.23609 +INFO:tensorflow:step = 16601, loss = 0.672655, precision = 0.875 (30.902 sec) +INFO:tensorflow:global_step/sec: 3.23682 +INFO:tensorflow:step = 16701, loss = 0.67987, precision = 0.867188 (30.895 sec) +INFO:tensorflow:global_step/sec: 3.23805 +INFO:tensorflow:step = 16801, loss = 0.717172, precision = 0.867188 (30.883 sec) +Saved checkpoint after 43 epoch(s) to ../data/resnet56/checkpoints/00043... +INFO:tensorflow:global_step/sec: 3.15391 +INFO:tensorflow:step = 16901, loss = 0.547641, precision = 0.914062 (31.707 sec) +INFO:tensorflow:global_step/sec: 3.24037 +INFO:tensorflow:step = 17001, loss = 0.639162, precision = 0.882812 (30.860 sec) +INFO:tensorflow:global_step/sec: 3.23712 +INFO:tensorflow:step = 17101, loss = 0.587917, precision = 0.882812 (30.892 sec) +INFO:tensorflow:global_step/sec: 3.23589 +INFO:tensorflow:step = 17201, loss = 0.656253, precision = 0.867188 (30.904 sec) +Saved checkpoint after 44 epoch(s) to ../data/resnet56/checkpoints/00044... +INFO:tensorflow:global_step/sec: 3.15616 +INFO:tensorflow:step = 17301, loss = 0.639221, precision = 0.890625 (31.684 sec) +INFO:tensorflow:global_step/sec: 3.23821 +INFO:tensorflow:step = 17401, loss = 0.688384, precision = 0.875 (30.881 sec) +INFO:tensorflow:global_step/sec: 3.2362 +INFO:tensorflow:step = 17501, loss = 0.551193, precision = 0.921875 (30.901 sec) +Saved checkpoint after 45 epoch(s) to ../data/resnet56/checkpoints/00045... +INFO:tensorflow:global_step/sec: 3.15937 +INFO:tensorflow:step = 17601, loss = 0.588563, precision = 0.90625 (31.652 sec) +INFO:tensorflow:global_step/sec: 3.23485 +INFO:tensorflow:step = 17701, loss = 0.702937, precision = 0.84375 (30.913 sec) +INFO:tensorflow:global_step/sec: 3.24054 +INFO:tensorflow:step = 17801, loss = 0.583214, precision = 0.90625 (30.859 sec) +INFO:tensorflow:global_step/sec: 3.23576 +INFO:tensorflow:step = 17901, loss = 0.607691, precision = 0.875 (30.904 sec) +Saved checkpoint after 46 epoch(s) to ../data/resnet56/checkpoints/00046... +INFO:tensorflow:global_step/sec: 3.14438 +INFO:tensorflow:step = 18001, loss = 0.634985, precision = 0.914062 (31.818 sec) +INFO:tensorflow:global_step/sec: 3.23668 +INFO:tensorflow:step = 18101, loss = 0.743974, precision = 0.84375 (30.880 sec) +INFO:tensorflow:global_step/sec: 3.23986 +INFO:tensorflow:step = 18201, loss = 0.501111, precision = 0.929688 (30.866 sec) +INFO:tensorflow:global_step/sec: 3.23692 +INFO:tensorflow:step = 18301, loss = 0.827658, precision = 0.820312 (30.894 sec) +Saved checkpoint after 47 epoch(s) to ../data/resnet56/checkpoints/00047... +INFO:tensorflow:global_step/sec: 3.15648 +INFO:tensorflow:step = 18401, loss = 0.719198, precision = 0.851562 (31.681 sec) +INFO:tensorflow:global_step/sec: 3.24025 +INFO:tensorflow:step = 18501, loss = 0.671451, precision = 0.859375 (30.862 sec) +INFO:tensorflow:global_step/sec: 3.24171 +INFO:tensorflow:step = 18601, loss = 0.708731, precision = 0.851562 (30.848 sec) +INFO:tensorflow:global_step/sec: 3.23903 +INFO:tensorflow:step = 18701, loss = 0.677597, precision = 0.875 (30.873 sec) +Saved checkpoint after 48 epoch(s) to ../data/resnet56/checkpoints/00048... +INFO:tensorflow:global_step/sec: 3.15563 +INFO:tensorflow:step = 18801, loss = 0.608053, precision = 0.882812 (31.689 sec) +INFO:tensorflow:global_step/sec: 3.23882 +INFO:tensorflow:step = 18901, loss = 0.579284, precision = 0.875 (30.875 sec) +INFO:tensorflow:global_step/sec: 3.24 +INFO:tensorflow:step = 19001, loss = 0.650907, precision = 0.890625 (30.864 sec) +INFO:tensorflow:global_step/sec: 3.23666 +INFO:tensorflow:step = 19101, loss = 0.588773, precision = 0.914062 (30.896 sec) +Saved checkpoint after 49 epoch(s) to ../data/resnet56/checkpoints/00049... +INFO:tensorflow:global_step/sec: 3.15762 +INFO:tensorflow:step = 19201, loss = 0.613147, precision = 0.898438 (31.669 sec) +INFO:tensorflow:global_step/sec: 3.23941 +INFO:tensorflow:step = 19301, loss = 0.636423, precision = 0.90625 (30.870 sec) +INFO:tensorflow:global_step/sec: 3.24068 +INFO:tensorflow:step = 19401, loss = 0.608439, precision = 0.875 (30.858 sec) +INFO:tensorflow:global_step/sec: 3.23512 +INFO:tensorflow:step = 19501, loss = 0.607607, precision = 0.882812 (30.911 sec) +Saved checkpoint after 50 epoch(s) to ../data/resnet56/checkpoints/00050... +INFO:tensorflow:global_step/sec: 3.15602 +INFO:tensorflow:step = 19601, loss = 0.622962, precision = 0.890625 (31.686 sec) +INFO:tensorflow:global_step/sec: 3.24445 +INFO:tensorflow:step = 19701, loss = 0.605348, precision = 0.921875 (30.822 sec) +INFO:tensorflow:global_step/sec: 3.2406 +INFO:tensorflow:step = 19801, loss = 0.591986, precision = 0.867188 (30.859 sec) +INFO:tensorflow:global_step/sec: 3.2365 +INFO:tensorflow:step = 19901, loss = 0.721384, precision = 0.867188 (30.897 sec) +Saved checkpoint after 51 epoch(s) to ../data/resnet56/checkpoints/00051... +INFO:tensorflow:global_step/sec: 3.16276 +INFO:tensorflow:step = 20001, loss = 0.582411, precision = 0.898438 (31.618 sec) +INFO:tensorflow:global_step/sec: 3.24005 +INFO:tensorflow:step = 20101, loss = 0.620796, precision = 0.890625 (30.864 sec) +INFO:tensorflow:global_step/sec: 3.23747 +INFO:tensorflow:step = 20201, loss = 0.601823, precision = 0.914062 (30.888 sec) +INFO:tensorflow:global_step/sec: 3.23534 +INFO:tensorflow:step = 20301, loss = 0.599424, precision = 0.890625 (30.908 sec) +Saved checkpoint after 52 epoch(s) to ../data/resnet56/checkpoints/00052... +INFO:tensorflow:global_step/sec: 3.15916 +INFO:tensorflow:step = 20401, loss = 0.553379, precision = 0.914062 (31.654 sec) +INFO:tensorflow:global_step/sec: 3.23831 +INFO:tensorflow:step = 20501, loss = 0.652253, precision = 0.875 (30.880 sec) +INFO:tensorflow:global_step/sec: 3.23861 +INFO:tensorflow:step = 20601, loss = 0.703549, precision = 0.851562 (30.877 sec) +INFO:tensorflow:global_step/sec: 3.24074 +INFO:tensorflow:step = 20701, loss = 0.752355, precision = 0.84375 (30.857 sec) +Saved checkpoint after 53 epoch(s) to ../data/resnet56/checkpoints/00053... +INFO:tensorflow:global_step/sec: 3.15781 +INFO:tensorflow:step = 20801, loss = 0.74283, precision = 0.859375 (31.667 sec) +INFO:tensorflow:global_step/sec: 3.24106 +INFO:tensorflow:step = 20901, loss = 0.599904, precision = 0.890625 (30.854 sec) +INFO:tensorflow:global_step/sec: 3.24125 +INFO:tensorflow:step = 21001, loss = 0.647433, precision = 0.882812 (30.852 sec) +INFO:tensorflow:global_step/sec: 3.2381 +INFO:tensorflow:step = 21101, loss = 0.595209, precision = 0.898438 (30.883 sec) +Saved checkpoint after 54 epoch(s) to ../data/resnet56/checkpoints/00054... +INFO:tensorflow:global_step/sec: 3.15756 +INFO:tensorflow:step = 21201, loss = 0.710204, precision = 0.851562 (31.670 sec) +INFO:tensorflow:global_step/sec: 3.24055 +INFO:tensorflow:step = 21301, loss = 0.721238, precision = 0.867188 (30.859 sec) +INFO:tensorflow:global_step/sec: 3.23659 +INFO:tensorflow:step = 21401, loss = 0.612211, precision = 0.882812 (30.897 sec) +INFO:tensorflow:global_step/sec: 3.23499 +INFO:tensorflow:step = 21501, loss = 0.689592, precision = 0.867188 (30.912 sec) +Saved checkpoint after 55 epoch(s) to ../data/resnet56/checkpoints/00055... +INFO:tensorflow:global_step/sec: 3.1638 +INFO:tensorflow:step = 21601, loss = 0.5745, precision = 0.90625 (31.608 sec) +INFO:tensorflow:global_step/sec: 3.24092 +INFO:tensorflow:step = 21701, loss = 0.568409, precision = 0.90625 (30.855 sec) +INFO:tensorflow:global_step/sec: 3.23603 +INFO:tensorflow:step = 21801, loss = 0.688212, precision = 0.882812 (30.902 sec) +Saved checkpoint after 56 epoch(s) to ../data/resnet56/checkpoints/00056... +INFO:tensorflow:global_step/sec: 3.16014 +INFO:tensorflow:step = 21901, loss = 0.673793, precision = 0.921875 (31.644 sec) +INFO:tensorflow:global_step/sec: 3.24099 +INFO:tensorflow:step = 22001, loss = 0.705811, precision = 0.851562 (30.855 sec) +INFO:tensorflow:global_step/sec: 3.24209 +INFO:tensorflow:step = 22101, loss = 0.700635, precision = 0.867188 (30.845 sec) +INFO:tensorflow:global_step/sec: 3.23696 +INFO:tensorflow:step = 22201, loss = 0.659931, precision = 0.882812 (30.893 sec) +Saved checkpoint after 57 epoch(s) to ../data/resnet56/checkpoints/00057... +INFO:tensorflow:global_step/sec: 3.15546 +INFO:tensorflow:step = 22301, loss = 0.552536, precision = 0.9375 (31.691 sec) +INFO:tensorflow:global_step/sec: 3.23258 +INFO:tensorflow:step = 22401, loss = 0.696495, precision = 0.84375 (30.935 sec) +INFO:tensorflow:global_step/sec: 3.23933 +INFO:tensorflow:step = 22501, loss = 0.7164, precision = 0.835938 (30.871 sec) +INFO:tensorflow:global_step/sec: 3.23969 +INFO:tensorflow:step = 22601, loss = 0.608863, precision = 0.859375 (30.867 sec) +Saved checkpoint after 58 epoch(s) to ../data/resnet56/checkpoints/00058... +INFO:tensorflow:global_step/sec: 3.13714 +INFO:tensorflow:step = 22701, loss = 0.664865, precision = 0.875 (31.876 sec) +INFO:tensorflow:global_step/sec: 3.24017 +INFO:tensorflow:step = 22801, loss = 0.604152, precision = 0.890625 (30.862 sec) +INFO:tensorflow:global_step/sec: 3.23653 +INFO:tensorflow:step = 22901, loss = 0.62488, precision = 0.882812 (30.897 sec) +INFO:tensorflow:global_step/sec: 3.2371 +INFO:tensorflow:step = 23001, loss = 0.676499, precision = 0.882812 (30.892 sec) +Saved checkpoint after 59 epoch(s) to ../data/resnet56/checkpoints/00059... +INFO:tensorflow:global_step/sec: 3.15986 +INFO:tensorflow:step = 23101, loss = 0.605079, precision = 0.882812 (31.647 sec) +INFO:tensorflow:global_step/sec: 3.23498 +INFO:tensorflow:step = 23201, loss = 0.642589, precision = 0.882812 (30.912 sec) +INFO:tensorflow:global_step/sec: 3.23646 +INFO:tensorflow:step = 23301, loss = 0.698477, precision = 0.90625 (30.898 sec) +INFO:tensorflow:global_step/sec: 3.23597 +INFO:tensorflow:step = 23401, loss = 0.708709, precision = 0.851562 (30.903 sec) +Saved checkpoint after 60 epoch(s) to ../data/resnet56/checkpoints/00060... +INFO:tensorflow:global_step/sec: 3.15825 +INFO:tensorflow:step = 23501, loss = 0.630185, precision = 0.882812 (31.663 sec) +INFO:tensorflow:global_step/sec: 3.24267 +INFO:tensorflow:step = 23601, loss = 0.728331, precision = 0.882812 (30.839 sec) +INFO:tensorflow:global_step/sec: 3.23641 +INFO:tensorflow:step = 23701, loss = 0.51455, precision = 0.9375 (30.898 sec) +INFO:tensorflow:global_step/sec: 3.23811 +INFO:tensorflow:step = 23801, loss = 0.677047, precision = 0.882812 (30.882 sec) +Saved checkpoint after 61 epoch(s) to ../data/resnet56/checkpoints/00061... +INFO:tensorflow:global_step/sec: 3.16256 +INFO:tensorflow:step = 23901, loss = 0.546025, precision = 0.921875 (31.620 sec) +INFO:tensorflow:global_step/sec: 3.24007 +INFO:tensorflow:step = 24001, loss = 0.732307, precision = 0.867188 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.23719 +INFO:tensorflow:step = 24101, loss = 0.663219, precision = 0.882812 (30.891 sec) +INFO:tensorflow:global_step/sec: 3.23608 +INFO:tensorflow:step = 24201, loss = 0.614973, precision = 0.914062 (30.902 sec) +Saved checkpoint after 62 epoch(s) to ../data/resnet56/checkpoints/00062... +INFO:tensorflow:global_step/sec: 3.15917 +INFO:tensorflow:step = 24301, loss = 0.585191, precision = 0.914062 (31.654 sec) +INFO:tensorflow:global_step/sec: 3.2404 +INFO:tensorflow:step = 24401, loss = 0.681217, precision = 0.867188 (30.860 sec) +INFO:tensorflow:global_step/sec: 3.2406 +INFO:tensorflow:step = 24501, loss = 0.889586, precision = 0.820312 (30.859 sec) +INFO:tensorflow:global_step/sec: 3.23665 +INFO:tensorflow:step = 24601, loss = 0.553324, precision = 0.914062 (30.896 sec) +Saved checkpoint after 63 epoch(s) to ../data/resnet56/checkpoints/00063... +INFO:tensorflow:global_step/sec: 3.16253 +INFO:tensorflow:step = 24701, loss = 0.656803, precision = 0.851562 (31.620 sec) +INFO:tensorflow:global_step/sec: 3.23973 +INFO:tensorflow:step = 24801, loss = 0.705029, precision = 0.859375 (30.867 sec) +INFO:tensorflow:global_step/sec: 3.24189 +INFO:tensorflow:step = 24901, loss = 0.705455, precision = 0.867188 (30.846 sec) +INFO:tensorflow:global_step/sec: 3.23803 +INFO:tensorflow:step = 25001, loss = 0.620605, precision = 0.90625 (30.883 sec) +Saved checkpoint after 64 epoch(s) to ../data/resnet56/checkpoints/00064... +INFO:tensorflow:global_step/sec: 3.16307 +INFO:tensorflow:step = 25101, loss = 0.730816, precision = 0.835938 (31.615 sec) +INFO:tensorflow:global_step/sec: 3.23787 +INFO:tensorflow:step = 25201, loss = 0.586629, precision = 0.890625 (30.885 sec) +INFO:tensorflow:global_step/sec: 3.23846 +INFO:tensorflow:step = 25301, loss = 0.614401, precision = 0.914062 (30.878 sec) +INFO:tensorflow:global_step/sec: 3.23756 +INFO:tensorflow:step = 25401, loss = 0.586756, precision = 0.914062 (30.888 sec) +Saved checkpoint after 65 epoch(s) to ../data/resnet56/checkpoints/00065... +INFO:tensorflow:global_step/sec: 3.15577 +INFO:tensorflow:step = 25501, loss = 0.650317, precision = 0.882812 (31.688 sec) +INFO:tensorflow:global_step/sec: 3.24078 +INFO:tensorflow:step = 25601, loss = 0.568821, precision = 0.914062 (30.857 sec) +INFO:tensorflow:global_step/sec: 3.2367 +INFO:tensorflow:step = 25701, loss = 0.844624, precision = 0.8125 (30.895 sec) +INFO:tensorflow:global_step/sec: 3.23848 +INFO:tensorflow:step = 25801, loss = 0.646571, precision = 0.890625 (30.879 sec) +Saved checkpoint after 66 epoch(s) to ../data/resnet56/checkpoints/00066... +INFO:tensorflow:global_step/sec: 3.15783 +INFO:tensorflow:step = 25901, loss = 0.610248, precision = 0.890625 (31.667 sec) +INFO:tensorflow:global_step/sec: 3.2402 +INFO:tensorflow:step = 26001, loss = 0.714345, precision = 0.867188 (30.862 sec) +INFO:tensorflow:global_step/sec: 3.23719 +INFO:tensorflow:step = 26101, loss = 0.766845, precision = 0.828125 (30.891 sec) +Saved checkpoint after 67 epoch(s) to ../data/resnet56/checkpoints/00067... +INFO:tensorflow:global_step/sec: 3.16213 +INFO:tensorflow:step = 26201, loss = 0.570452, precision = 0.921875 (31.624 sec) +INFO:tensorflow:global_step/sec: 3.241 +INFO:tensorflow:step = 26301, loss = 0.638689, precision = 0.882812 (30.855 sec) +INFO:tensorflow:global_step/sec: 3.23994 +INFO:tensorflow:step = 26401, loss = 0.575185, precision = 0.898438 (30.865 sec) +INFO:tensorflow:global_step/sec: 3.24252 +INFO:tensorflow:step = 26501, loss = 0.539279, precision = 0.9375 (30.840 sec) +Saved checkpoint after 68 epoch(s) to ../data/resnet56/checkpoints/00068... +INFO:tensorflow:global_step/sec: 3.16102 +INFO:tensorflow:step = 26601, loss = 0.592931, precision = 0.890625 (31.635 sec) +INFO:tensorflow:global_step/sec: 3.24316 +INFO:tensorflow:step = 26701, loss = 0.773641, precision = 0.828125 (30.834 sec) +INFO:tensorflow:global_step/sec: 3.24141 +INFO:tensorflow:step = 26801, loss = 0.660425, precision = 0.898438 (30.851 sec) +INFO:tensorflow:global_step/sec: 3.24019 +INFO:tensorflow:step = 26901, loss = 0.72001, precision = 0.875 (30.862 sec) +Saved checkpoint after 69 epoch(s) to ../data/resnet56/checkpoints/00069... +INFO:tensorflow:global_step/sec: 3.1656 +INFO:tensorflow:step = 27001, loss = 0.609474, precision = 0.898438 (31.590 sec) +INFO:tensorflow:global_step/sec: 3.24385 +INFO:tensorflow:step = 27101, loss = 0.630652, precision = 0.90625 (30.827 sec) +INFO:tensorflow:global_step/sec: 3.23893 +INFO:tensorflow:step = 27201, loss = 0.532151, precision = 0.929688 (30.875 sec) +INFO:tensorflow:global_step/sec: 3.23744 +INFO:tensorflow:step = 27301, loss = 0.667772, precision = 0.867188 (30.889 sec) +Saved checkpoint after 70 epoch(s) to ../data/resnet56/checkpoints/00070... +INFO:tensorflow:global_step/sec: 3.16008 +INFO:tensorflow:step = 27401, loss = 0.64001, precision = 0.882812 (31.644 sec) +INFO:tensorflow:global_step/sec: 3.23925 +INFO:tensorflow:step = 27501, loss = 0.657556, precision = 0.890625 (30.872 sec) +INFO:tensorflow:global_step/sec: 3.23719 +INFO:tensorflow:step = 27601, loss = 0.627059, precision = 0.90625 (30.891 sec) +INFO:tensorflow:global_step/sec: 3.23918 +INFO:tensorflow:step = 27701, loss = 0.516353, precision = 0.9375 (30.872 sec) +Saved checkpoint after 71 epoch(s) to ../data/resnet56/checkpoints/00071... +INFO:tensorflow:global_step/sec: 3.15567 +INFO:tensorflow:step = 27801, loss = 0.635233, precision = 0.914062 (31.689 sec) +INFO:tensorflow:global_step/sec: 3.24019 +INFO:tensorflow:step = 27901, loss = 0.611332, precision = 0.898438 (30.862 sec) +INFO:tensorflow:global_step/sec: 3.23816 +INFO:tensorflow:step = 28001, loss = 0.747859, precision = 0.835938 (30.882 sec) +INFO:tensorflow:global_step/sec: 3.23997 +INFO:tensorflow:step = 28101, loss = 0.588478, precision = 0.914062 (30.864 sec) +Saved checkpoint after 72 epoch(s) to ../data/resnet56/checkpoints/00072... +INFO:tensorflow:global_step/sec: 3.1587 +INFO:tensorflow:step = 28201, loss = 0.687534, precision = 0.875 (31.659 sec) +INFO:tensorflow:global_step/sec: 3.23931 +INFO:tensorflow:step = 28301, loss = 0.545856, precision = 0.929688 (30.871 sec) +INFO:tensorflow:global_step/sec: 3.24021 +INFO:tensorflow:step = 28401, loss = 0.645181, precision = 0.890625 (30.862 sec) +INFO:tensorflow:global_step/sec: 3.24094 +INFO:tensorflow:step = 28501, loss = 0.808302, precision = 0.820312 (30.855 sec) +Saved checkpoint after 73 epoch(s) to ../data/resnet56/checkpoints/00073... +INFO:tensorflow:global_step/sec: 3.16326 +INFO:tensorflow:step = 28601, loss = 0.551508, precision = 0.882812 (31.613 sec) +INFO:tensorflow:global_step/sec: 3.23964 +INFO:tensorflow:step = 28701, loss = 0.645944, precision = 0.90625 (30.868 sec) +INFO:tensorflow:global_step/sec: 3.23633 +INFO:tensorflow:step = 28801, loss = 0.550621, precision = 0.945312 (30.899 sec) +INFO:tensorflow:global_step/sec: 3.23563 +INFO:tensorflow:step = 28901, loss = 0.563335, precision = 0.945312 (30.906 sec) +Saved checkpoint after 74 epoch(s) to ../data/resnet56/checkpoints/00074... +INFO:tensorflow:global_step/sec: 3.16655 +INFO:tensorflow:step = 29001, loss = 0.758797, precision = 0.835938 (31.580 sec) +INFO:tensorflow:global_step/sec: 3.23986 +INFO:tensorflow:step = 29101, loss = 0.616957, precision = 0.90625 (30.866 sec) +INFO:tensorflow:global_step/sec: 3.2364 +INFO:tensorflow:step = 29201, loss = 0.690081, precision = 0.875 (30.899 sec) +INFO:tensorflow:global_step/sec: 3.24136 +INFO:tensorflow:step = 29301, loss = 0.639352, precision = 0.882812 (30.851 sec) +Saved checkpoint after 75 epoch(s) to ../data/resnet56/checkpoints/00075... +INFO:tensorflow:global_step/sec: 3.16625 +INFO:tensorflow:step = 29401, loss = 0.66787, precision = 0.921875 (31.583 sec) +INFO:tensorflow:global_step/sec: 3.23935 +INFO:tensorflow:step = 29501, loss = 0.614219, precision = 0.898438 (30.871 sec) +INFO:tensorflow:global_step/sec: 3.23934 +INFO:tensorflow:step = 29601, loss = 0.74642, precision = 0.8125 (30.871 sec) +INFO:tensorflow:global_step/sec: 3.24076 +INFO:tensorflow:step = 29701, loss = 0.643869, precision = 0.890625 (30.857 sec) +Saved checkpoint after 76 epoch(s) to ../data/resnet56/checkpoints/00076... +INFO:tensorflow:global_step/sec: 3.16875 +INFO:tensorflow:step = 29801, loss = 0.534243, precision = 0.953125 (31.558 sec) +INFO:tensorflow:global_step/sec: 3.24016 +INFO:tensorflow:step = 29901, loss = 0.810334, precision = 0.851562 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.23821 +INFO:tensorflow:step = 30001, loss = 0.711269, precision = 0.875 (30.881 sec) +INFO:tensorflow:global_step/sec: 3.2436 +INFO:tensorflow:step = 30101, loss = 0.564615, precision = 0.898438 (30.830 sec) +Saved checkpoint after 77 epoch(s) to ../data/resnet56/checkpoints/00077... +INFO:tensorflow:global_step/sec: 3.16576 +INFO:tensorflow:step = 30201, loss = 0.728617, precision = 0.835938 (31.588 sec) +INFO:tensorflow:global_step/sec: 3.24281 +INFO:tensorflow:step = 30301, loss = 0.814668, precision = 0.851562 (30.837 sec) +INFO:tensorflow:global_step/sec: 3.23847 +INFO:tensorflow:step = 30401, loss = 0.577226, precision = 0.921875 (30.879 sec) +Saved checkpoint after 78 epoch(s) to ../data/resnet56/checkpoints/00078... +INFO:tensorflow:global_step/sec: 3.16215 +INFO:tensorflow:step = 30501, loss = 0.568029, precision = 0.914062 (31.624 sec) +INFO:tensorflow:global_step/sec: 3.24015 +INFO:tensorflow:step = 30601, loss = 0.687378, precision = 0.851562 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.23923 +INFO:tensorflow:step = 30701, loss = 0.785272, precision = 0.835938 (30.871 sec) +INFO:tensorflow:global_step/sec: 3.24279 +INFO:tensorflow:step = 30801, loss = 0.508065, precision = 0.921875 (30.838 sec) +Saved checkpoint after 79 epoch(s) to ../data/resnet56/checkpoints/00079... +INFO:tensorflow:global_step/sec: 3.16661 +INFO:tensorflow:step = 30901, loss = 0.69954, precision = 0.835938 (31.580 sec) +INFO:tensorflow:global_step/sec: 3.24198 +INFO:tensorflow:step = 31001, loss = 0.595411, precision = 0.90625 (30.845 sec) +INFO:tensorflow:global_step/sec: 3.23955 +INFO:tensorflow:step = 31101, loss = 0.493298, precision = 0.953125 (30.868 sec) +INFO:tensorflow:global_step/sec: 3.24054 +INFO:tensorflow:step = 31201, loss = 0.670995, precision = 0.875 (30.860 sec) +Saved checkpoint after 80 epoch(s) to ../data/resnet56/checkpoints/00080... +INFO:tensorflow:global_step/sec: 3.16596 +INFO:tensorflow:step = 31301, loss = 0.542607, precision = 0.929688 (31.586 sec) +INFO:tensorflow:global_step/sec: 3.23952 +INFO:tensorflow:step = 31401, loss = 0.553784, precision = 0.921875 (30.869 sec) +INFO:tensorflow:global_step/sec: 3.23855 +INFO:tensorflow:step = 31501, loss = 0.56379, precision = 0.921875 (30.878 sec) +INFO:tensorflow:global_step/sec: 3.24064 +INFO:tensorflow:step = 31601, loss = 0.67815, precision = 0.875 (30.858 sec) +Saved checkpoint after 81 epoch(s) to ../data/resnet56/checkpoints/00081... +INFO:tensorflow:global_step/sec: 3.16032 +INFO:tensorflow:step = 31701, loss = 0.70078, precision = 0.84375 (31.643 sec) +INFO:tensorflow:global_step/sec: 3.23694 +INFO:tensorflow:step = 31801, loss = 0.73544, precision = 0.851562 (30.893 sec) +INFO:tensorflow:global_step/sec: 3.23941 +INFO:tensorflow:step = 31901, loss = 0.637603, precision = 0.875 (30.870 sec) +INFO:tensorflow:global_step/sec: 3.24183 +INFO:tensorflow:step = 32001, loss = 0.638791, precision = 0.851562 (30.847 sec) +Saved checkpoint after 82 epoch(s) to ../data/resnet56/checkpoints/00082... +INFO:tensorflow:global_step/sec: 3.15295 +INFO:tensorflow:step = 32101, loss = 0.601262, precision = 0.90625 (31.716 sec) +INFO:tensorflow:global_step/sec: 3.23907 +INFO:tensorflow:step = 32201, loss = 0.615827, precision = 0.867188 (30.873 sec) +INFO:tensorflow:global_step/sec: 3.24048 +INFO:tensorflow:step = 32301, loss = 0.597451, precision = 0.882812 (30.860 sec) +INFO:tensorflow:global_step/sec: 3.23851 +INFO:tensorflow:step = 32401, loss = 0.64879, precision = 0.851562 (30.878 sec) +Saved checkpoint after 83 epoch(s) to ../data/resnet56/checkpoints/00083... +INFO:tensorflow:global_step/sec: 3.15928 +INFO:tensorflow:step = 32501, loss = 0.640822, precision = 0.867188 (31.653 sec) +INFO:tensorflow:global_step/sec: 3.24103 +INFO:tensorflow:step = 32601, loss = 0.708399, precision = 0.851562 (30.854 sec) +INFO:tensorflow:global_step/sec: 3.24049 +INFO:tensorflow:step = 32701, loss = 0.582245, precision = 0.898438 (30.859 sec) +INFO:tensorflow:global_step/sec: 3.24133 +INFO:tensorflow:step = 32801, loss = 0.77683, precision = 0.851562 (30.852 sec) +Saved checkpoint after 84 epoch(s) to ../data/resnet56/checkpoints/00084... +INFO:tensorflow:global_step/sec: 3.1648 +INFO:tensorflow:step = 32901, loss = 0.556718, precision = 0.914062 (31.598 sec) +INFO:tensorflow:global_step/sec: 3.23939 +INFO:tensorflow:step = 33001, loss = 0.670322, precision = 0.835938 (30.870 sec) +INFO:tensorflow:global_step/sec: 3.24069 +INFO:tensorflow:step = 33101, loss = 0.73616, precision = 0.828125 (30.858 sec) +INFO:tensorflow:global_step/sec: 3.24732 +INFO:tensorflow:step = 33201, loss = 0.628339, precision = 0.882812 (30.795 sec) +Saved checkpoint after 85 epoch(s) to ../data/resnet56/checkpoints/00085... +INFO:tensorflow:global_step/sec: 3.16304 +INFO:tensorflow:step = 33301, loss = 0.689889, precision = 0.875 (31.615 sec) +INFO:tensorflow:global_step/sec: 3.24013 +INFO:tensorflow:step = 33401, loss = 0.608276, precision = 0.898438 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.24125 +INFO:tensorflow:step = 33501, loss = 0.643935, precision = 0.898438 (30.852 sec) +INFO:tensorflow:global_step/sec: 3.24055 +INFO:tensorflow:step = 33601, loss = 0.647876, precision = 0.898438 (30.859 sec) +Saved checkpoint after 86 epoch(s) to ../data/resnet56/checkpoints/00086... +INFO:tensorflow:global_step/sec: 3.16152 +INFO:tensorflow:step = 33701, loss = 0.696604, precision = 0.859375 (31.630 sec) +INFO:tensorflow:global_step/sec: 3.24126 +INFO:tensorflow:step = 33801, loss = 0.622812, precision = 0.890625 (30.852 sec) +INFO:tensorflow:global_step/sec: 3.24142 +INFO:tensorflow:step = 33901, loss = 0.691204, precision = 0.867188 (30.851 sec) +INFO:tensorflow:global_step/sec: 3.24242 +INFO:tensorflow:step = 34001, loss = 0.666581, precision = 0.875 (30.841 sec) +Saved checkpoint after 87 epoch(s) to ../data/resnet56/checkpoints/00087... +INFO:tensorflow:global_step/sec: 3.16841 +INFO:tensorflow:step = 34101, loss = 0.755758, precision = 0.851562 (31.561 sec) +INFO:tensorflow:global_step/sec: 3.24054 +INFO:tensorflow:step = 34201, loss = 0.660613, precision = 0.898438 (30.859 sec) +INFO:tensorflow:global_step/sec: 3.24006 +INFO:tensorflow:step = 34301, loss = 0.629855, precision = 0.90625 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.2445 +INFO:tensorflow:step = 34401, loss = 0.705693, precision = 0.851562 (30.822 sec) +Saved checkpoint after 88 epoch(s) to ../data/resnet56/checkpoints/00088... +INFO:tensorflow:global_step/sec: 3.15997 +INFO:tensorflow:step = 34501, loss = 0.598873, precision = 0.890625 (31.646 sec) +INFO:tensorflow:global_step/sec: 3.23858 +INFO:tensorflow:step = 34601, loss = 0.675575, precision = 0.882812 (30.878 sec) +INFO:tensorflow:global_step/sec: 3.24101 +INFO:tensorflow:step = 34701, loss = 0.573516, precision = 0.921875 (30.854 sec) +Saved checkpoint after 89 epoch(s) to ../data/resnet56/checkpoints/00089... +INFO:tensorflow:global_step/sec: 3.16622 +INFO:tensorflow:step = 34801, loss = 0.538588, precision = 0.921875 (31.584 sec) +INFO:tensorflow:global_step/sec: 3.23778 +INFO:tensorflow:step = 34901, loss = 0.673319, precision = 0.898438 (30.885 sec) +INFO:tensorflow:global_step/sec: 3.23986 +INFO:tensorflow:step = 35001, loss = 0.6846, precision = 0.84375 (30.865 sec) +INFO:tensorflow:global_step/sec: 3.24045 +INFO:tensorflow:step = 35101, loss = 0.703716, precision = 0.820312 (30.860 sec) +Saved checkpoint after 90 epoch(s) to ../data/resnet56/checkpoints/00090... +INFO:tensorflow:global_step/sec: 3.16727 +INFO:tensorflow:step = 35201, loss = 0.56047, precision = 0.90625 (31.573 sec) +INFO:tensorflow:global_step/sec: 3.23722 +INFO:tensorflow:step = 35301, loss = 0.58823, precision = 0.921875 (30.891 sec) +INFO:tensorflow:global_step/sec: 3.24026 +INFO:tensorflow:step = 35401, loss = 0.591813, precision = 0.890625 (30.862 sec) +INFO:tensorflow:global_step/sec: 3.24091 +INFO:tensorflow:step = 35501, loss = 0.593875, precision = 0.898438 (30.856 sec) +Saved checkpoint after 91 epoch(s) to ../data/resnet56/checkpoints/00091... +INFO:tensorflow:global_step/sec: 3.16533 +INFO:tensorflow:step = 35601, loss = 0.613849, precision = 0.882812 (31.592 sec) +INFO:tensorflow:global_step/sec: 3.24078 +INFO:tensorflow:step = 35701, loss = 0.479265, precision = 0.929688 (30.856 sec) +INFO:tensorflow:global_step/sec: 3.24248 +INFO:tensorflow:step = 35801, loss = 0.539318, precision = 0.898438 (30.841 sec) +INFO:tensorflow:global_step/sec: 3.24048 +INFO:tensorflow:step = 35901, loss = 0.542818, precision = 0.929688 (30.860 sec) +Saved checkpoint after 92 epoch(s) to ../data/resnet56/checkpoints/00092... +INFO:tensorflow:global_step/sec: 3.16403 +INFO:tensorflow:step = 36001, loss = 0.471172, precision = 0.945312 (31.605 sec) +INFO:tensorflow:global_step/sec: 3.239 +INFO:tensorflow:step = 36101, loss = 0.461532, precision = 0.921875 (30.874 sec) +INFO:tensorflow:global_step/sec: 3.23851 +INFO:tensorflow:step = 36201, loss = 0.480311, precision = 0.953125 (30.878 sec) +INFO:tensorflow:global_step/sec: 3.24014 +INFO:tensorflow:step = 36301, loss = 0.45477, precision = 0.960938 (30.863 sec) +Saved checkpoint after 93 epoch(s) to ../data/resnet56/checkpoints/00093... +INFO:tensorflow:global_step/sec: 3.16344 +INFO:tensorflow:step = 36401, loss = 0.431867, precision = 0.9375 (31.611 sec) +INFO:tensorflow:global_step/sec: 3.23919 +INFO:tensorflow:step = 36501, loss = 0.458183, precision = 0.953125 (30.872 sec) +INFO:tensorflow:global_step/sec: 3.23947 +INFO:tensorflow:step = 36601, loss = 0.375476, precision = 0.96875 (30.869 sec) +INFO:tensorflow:global_step/sec: 3.23957 +INFO:tensorflow:step = 36701, loss = 0.430492, precision = 0.945312 (30.868 sec) +Saved checkpoint after 94 epoch(s) to ../data/resnet56/checkpoints/00094... +INFO:tensorflow:global_step/sec: 3.16052 +INFO:tensorflow:step = 36801, loss = 0.351233, precision = 0.984375 (31.641 sec) +INFO:tensorflow:global_step/sec: 3.23782 +INFO:tensorflow:step = 36901, loss = 0.383298, precision = 0.976562 (30.885 sec) +INFO:tensorflow:global_step/sec: 3.24099 +INFO:tensorflow:step = 37001, loss = 0.387388, precision = 0.96875 (30.855 sec) +INFO:tensorflow:global_step/sec: 3.24558 +INFO:tensorflow:step = 37101, loss = 0.356051, precision = 0.976562 (30.811 sec) +Saved checkpoint after 95 epoch(s) to ../data/resnet56/checkpoints/00095... +INFO:tensorflow:global_step/sec: 3.1569 +INFO:tensorflow:step = 37201, loss = 0.399721, precision = 0.960938 (31.676 sec) +INFO:tensorflow:global_step/sec: 3.24025 +INFO:tensorflow:step = 37301, loss = 0.394378, precision = 0.960938 (30.862 sec) +INFO:tensorflow:global_step/sec: 3.24254 +INFO:tensorflow:step = 37401, loss = 0.406494, precision = 0.945312 (30.840 sec) +INFO:tensorflow:global_step/sec: 3.24257 +INFO:tensorflow:step = 37501, loss = 0.394817, precision = 0.945312 (30.839 sec) +Saved checkpoint after 96 epoch(s) to ../data/resnet56/checkpoints/00096... +INFO:tensorflow:global_step/sec: 3.16266 +INFO:tensorflow:step = 37601, loss = 0.357393, precision = 0.96875 (31.619 sec) +INFO:tensorflow:global_step/sec: 3.24147 +INFO:tensorflow:step = 37701, loss = 0.325177, precision = 0.976562 (30.850 sec) +INFO:tensorflow:global_step/sec: 3.24167 +INFO:tensorflow:step = 37801, loss = 0.364041, precision = 0.960938 (30.848 sec) +INFO:tensorflow:global_step/sec: 3.24032 +INFO:tensorflow:step = 37901, loss = 0.312122, precision = 0.984375 (30.861 sec) +Saved checkpoint after 97 epoch(s) to ../data/resnet56/checkpoints/00097... +INFO:tensorflow:global_step/sec: 3.16042 +INFO:tensorflow:step = 38001, loss = 0.418578, precision = 0.953125 (31.641 sec) +INFO:tensorflow:global_step/sec: 3.24027 +INFO:tensorflow:step = 38101, loss = 0.39517, precision = 0.945312 (30.861 sec) +INFO:tensorflow:global_step/sec: 3.23877 +INFO:tensorflow:step = 38201, loss = 0.286413, precision = 0.992188 (30.876 sec) +INFO:tensorflow:global_step/sec: 3.24215 +INFO:tensorflow:step = 38301, loss = 0.319978, precision = 0.96875 (30.844 sec) +Saved checkpoint after 98 epoch(s) to ../data/resnet56/checkpoints/00098... +INFO:tensorflow:global_step/sec: 3.16286 +INFO:tensorflow:step = 38401, loss = 0.402525, precision = 0.953125 (31.617 sec) +INFO:tensorflow:global_step/sec: 3.23961 +INFO:tensorflow:step = 38501, loss = 0.360962, precision = 0.96875 (30.868 sec) +INFO:tensorflow:global_step/sec: 3.23931 +INFO:tensorflow:step = 38601, loss = 0.306361, precision = 0.984375 (30.871 sec) +INFO:tensorflow:global_step/sec: 3.24097 +INFO:tensorflow:step = 38701, loss = 0.277662, precision = 0.984375 (30.855 sec) +Saved checkpoint after 99 epoch(s) to ../data/resnet56/checkpoints/00099... +INFO:tensorflow:global_step/sec: 3.15888 +INFO:tensorflow:step = 38801, loss = 0.330869, precision = 0.984375 (31.657 sec) +INFO:tensorflow:global_step/sec: 3.23981 +INFO:tensorflow:step = 38901, loss = 0.371045, precision = 0.960938 (30.866 sec) +INFO:tensorflow:global_step/sec: 3.24581 +INFO:tensorflow:step = 39001, loss = 0.333956, precision = 0.960938 (30.809 sec) +Saved checkpoint after 100 epoch(s) to ../data/resnet56/checkpoints/00100... +INFO:tensorflow:global_step/sec: 3.15457 +INFO:tensorflow:step = 39101, loss = 0.285837, precision = 0.992188 (31.700 sec) +INFO:tensorflow:global_step/sec: 3.24009 +INFO:tensorflow:step = 39201, loss = 0.249009, precision = 1.0 (30.864 sec) +INFO:tensorflow:global_step/sec: 3.23882 +INFO:tensorflow:step = 39301, loss = 0.275424, precision = 0.976562 (30.875 sec) +INFO:tensorflow:global_step/sec: 3.24037 +INFO:tensorflow:step = 39401, loss = 0.294287, precision = 0.976562 (30.861 sec) +Saved checkpoint after 101 epoch(s) to ../data/resnet56/checkpoints/00101... +INFO:tensorflow:global_step/sec: 3.15425 +INFO:tensorflow:step = 39501, loss = 0.263849, precision = 1.0 (31.703 sec) +INFO:tensorflow:global_step/sec: 3.24118 +INFO:tensorflow:step = 39601, loss = 0.275091, precision = 0.984375 (30.853 sec) +INFO:tensorflow:global_step/sec: 3.23921 +INFO:tensorflow:step = 39701, loss = 0.312805, precision = 0.960938 (30.872 sec) +INFO:tensorflow:global_step/sec: 3.24094 +INFO:tensorflow:step = 39801, loss = 0.296475, precision = 0.960938 (30.855 sec) +Saved checkpoint after 102 epoch(s) to ../data/resnet56/checkpoints/00102... +INFO:tensorflow:global_step/sec: 3.15936 +INFO:tensorflow:step = 39901, loss = 0.275977, precision = 0.976562 (31.652 sec) +INFO:tensorflow:global_step/sec: 3.24259 +INFO:tensorflow:step = 40001, loss = 0.26825, precision = 0.976562 (30.840 sec) +INFO:tensorflow:global_step/sec: 3.23965 +INFO:tensorflow:step = 40101, loss = 0.297625, precision = 0.976562 (30.868 sec) +INFO:tensorflow:global_step/sec: 3.23916 +INFO:tensorflow:step = 40201, loss = 0.289002, precision = 0.960938 (30.872 sec) +Saved checkpoint after 103 epoch(s) to ../data/resnet56/checkpoints/00103... +INFO:tensorflow:global_step/sec: 3.16083 +INFO:tensorflow:step = 40301, loss = 0.280573, precision = 0.984375 (31.637 sec) +INFO:tensorflow:global_step/sec: 3.23895 +INFO:tensorflow:step = 40401, loss = 0.248995, precision = 0.992188 (30.874 sec) +INFO:tensorflow:global_step/sec: 3.24009 +INFO:tensorflow:step = 40501, loss = 0.285467, precision = 0.96875 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.24059 +INFO:tensorflow:step = 40601, loss = 0.302362, precision = 0.96875 (30.858 sec) +Saved checkpoint after 104 epoch(s) to ../data/resnet56/checkpoints/00104... +INFO:tensorflow:global_step/sec: 3.1586 +INFO:tensorflow:step = 40701, loss = 0.280053, precision = 0.976562 (31.660 sec) +INFO:tensorflow:global_step/sec: 3.23676 +INFO:tensorflow:step = 40801, loss = 0.318835, precision = 0.953125 (30.895 sec) +INFO:tensorflow:global_step/sec: 3.24212 +INFO:tensorflow:step = 40901, loss = 0.343092, precision = 0.953125 (30.844 sec) +INFO:tensorflow:global_step/sec: 3.24214 +INFO:tensorflow:step = 41001, loss = 0.234534, precision = 0.992188 (30.844 sec) +Saved checkpoint after 105 epoch(s) to ../data/resnet56/checkpoints/00105... +INFO:tensorflow:global_step/sec: 3.15857 +INFO:tensorflow:step = 41101, loss = 0.240854, precision = 0.984375 (31.660 sec) +INFO:tensorflow:global_step/sec: 3.24016 +INFO:tensorflow:step = 41201, loss = 0.25424, precision = 0.984375 (30.862 sec) +INFO:tensorflow:global_step/sec: 3.24147 +INFO:tensorflow:step = 41301, loss = 0.274741, precision = 0.984375 (30.850 sec) +INFO:tensorflow:global_step/sec: 3.23962 +INFO:tensorflow:step = 41401, loss = 0.239389, precision = 0.984375 (30.868 sec) +Saved checkpoint after 106 epoch(s) to ../data/resnet56/checkpoints/00106... +INFO:tensorflow:global_step/sec: 3.15819 +INFO:tensorflow:step = 41501, loss = 0.223727, precision = 0.992188 (31.664 sec) +INFO:tensorflow:global_step/sec: 3.23996 +INFO:tensorflow:step = 41601, loss = 0.225701, precision = 0.992188 (30.864 sec) +INFO:tensorflow:global_step/sec: 3.2428 +INFO:tensorflow:step = 41701, loss = 0.266907, precision = 0.96875 (30.837 sec) +INFO:tensorflow:global_step/sec: 3.24283 +INFO:tensorflow:step = 41801, loss = 0.234246, precision = 0.976562 (30.837 sec) +Saved checkpoint after 107 epoch(s) to ../data/resnet56/checkpoints/00107... +INFO:tensorflow:global_step/sec: 3.15927 +INFO:tensorflow:step = 41901, loss = 0.292504, precision = 0.976562 (31.653 sec) +INFO:tensorflow:global_step/sec: 3.2411 +INFO:tensorflow:step = 42001, loss = 0.237089, precision = 0.992188 (30.854 sec) +INFO:tensorflow:global_step/sec: 3.24204 +INFO:tensorflow:step = 42101, loss = 0.249975, precision = 0.992188 (30.845 sec) +INFO:tensorflow:global_step/sec: 3.24137 +INFO:tensorflow:step = 42201, loss = 0.223978, precision = 0.984375 (30.851 sec) +Saved checkpoint after 108 epoch(s) to ../data/resnet56/checkpoints/00108... +INFO:tensorflow:global_step/sec: 3.14604 +INFO:tensorflow:step = 42301, loss = 0.246362, precision = 0.96875 (31.786 sec) +INFO:tensorflow:global_step/sec: 3.24175 +INFO:tensorflow:step = 42401, loss = 0.212614, precision = 0.992188 (30.847 sec) +INFO:tensorflow:global_step/sec: 3.24182 +INFO:tensorflow:step = 42501, loss = 0.336555, precision = 0.921875 (30.847 sec) +INFO:tensorflow:global_step/sec: 3.24169 +INFO:tensorflow:step = 42601, loss = 0.242827, precision = 0.984375 (30.848 sec) +Saved checkpoint after 109 epoch(s) to ../data/resnet56/checkpoints/00109... +INFO:tensorflow:global_step/sec: 3.15779 +INFO:tensorflow:step = 42701, loss = 0.24367, precision = 0.984375 (31.668 sec) +INFO:tensorflow:global_step/sec: 3.24057 +INFO:tensorflow:step = 42801, loss = 0.240731, precision = 0.96875 (30.859 sec) +INFO:tensorflow:global_step/sec: 3.24494 +INFO:tensorflow:step = 42901, loss = 0.217032, precision = 0.992188 (30.817 sec) +INFO:tensorflow:global_step/sec: 3.23835 +INFO:tensorflow:step = 43001, loss = 0.221685, precision = 0.984375 (30.880 sec) +Saved checkpoint after 110 epoch(s) to ../data/resnet56/checkpoints/00110... +INFO:tensorflow:global_step/sec: 3.15892 +INFO:tensorflow:step = 43101, loss = 0.219649, precision = 0.992188 (31.656 sec) +INFO:tensorflow:global_step/sec: 3.24096 +INFO:tensorflow:step = 43201, loss = 0.211995, precision = 0.984375 (30.855 sec) +INFO:tensorflow:global_step/sec: 3.24364 +INFO:tensorflow:step = 43301, loss = 0.231768, precision = 0.976562 (30.830 sec) +Saved checkpoint after 111 epoch(s) to ../data/resnet56/checkpoints/00111... +INFO:tensorflow:global_step/sec: 3.16147 +INFO:tensorflow:step = 43401, loss = 0.256759, precision = 0.976562 (31.631 sec) +INFO:tensorflow:global_step/sec: 3.24101 +INFO:tensorflow:step = 43501, loss = 0.255642, precision = 0.960938 (30.855 sec) +INFO:tensorflow:global_step/sec: 3.2418 +INFO:tensorflow:step = 43601, loss = 0.264706, precision = 0.96875 (30.847 sec) +INFO:tensorflow:global_step/sec: 3.24145 +INFO:tensorflow:step = 43701, loss = 0.23828, precision = 0.984375 (30.850 sec) +Saved checkpoint after 112 epoch(s) to ../data/resnet56/checkpoints/00112... +INFO:tensorflow:global_step/sec: 3.16129 +INFO:tensorflow:step = 43801, loss = 0.248073, precision = 0.976562 (31.633 sec) +INFO:tensorflow:global_step/sec: 3.24031 +INFO:tensorflow:step = 43901, loss = 0.247252, precision = 0.96875 (30.861 sec) +INFO:tensorflow:global_step/sec: 3.23992 +INFO:tensorflow:step = 44001, loss = 0.210463, precision = 0.984375 (30.865 sec) +INFO:tensorflow:global_step/sec: 3.24391 +INFO:tensorflow:step = 44101, loss = 0.222474, precision = 0.984375 (30.827 sec) +Saved checkpoint after 113 epoch(s) to ../data/resnet56/checkpoints/00113... +INFO:tensorflow:global_step/sec: 3.1601 +INFO:tensorflow:step = 44201, loss = 0.247904, precision = 0.96875 (31.645 sec) +INFO:tensorflow:global_step/sec: 3.23801 +INFO:tensorflow:step = 44301, loss = 0.239128, precision = 0.96875 (30.883 sec) +INFO:tensorflow:global_step/sec: 3.2401 +INFO:tensorflow:step = 44401, loss = 0.226834, precision = 0.984375 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.24176 +INFO:tensorflow:step = 44501, loss = 0.199891, precision = 0.992188 (30.847 sec) +Saved checkpoint after 114 epoch(s) to ../data/resnet56/checkpoints/00114... +INFO:tensorflow:global_step/sec: 3.15999 +INFO:tensorflow:step = 44601, loss = 0.230208, precision = 0.984375 (31.645 sec) +INFO:tensorflow:global_step/sec: 3.2409 +INFO:tensorflow:step = 44701, loss = 0.265215, precision = 0.96875 (30.856 sec) +INFO:tensorflow:global_step/sec: 3.24332 +INFO:tensorflow:step = 44801, loss = 0.216664, precision = 0.984375 (30.833 sec) +INFO:tensorflow:global_step/sec: 3.24277 +INFO:tensorflow:step = 44901, loss = 0.209889, precision = 0.976562 (30.838 sec) +Saved checkpoint after 115 epoch(s) to ../data/resnet56/checkpoints/00115... +INFO:tensorflow:global_step/sec: 3.15732 +INFO:tensorflow:step = 45001, loss = 0.286685, precision = 0.960938 (31.673 sec) +INFO:tensorflow:global_step/sec: 3.2401 +INFO:tensorflow:step = 45101, loss = 0.226548, precision = 0.976562 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.23914 +INFO:tensorflow:step = 45201, loss = 0.238516, precision = 0.96875 (30.872 sec) +INFO:tensorflow:global_step/sec: 3.24267 +INFO:tensorflow:step = 45301, loss = 0.250804, precision = 0.984375 (30.839 sec) +Saved checkpoint after 116 epoch(s) to ../data/resnet56/checkpoints/00116... +INFO:tensorflow:global_step/sec: 3.15309 +INFO:tensorflow:step = 45401, loss = 0.247232, precision = 0.945312 (31.715 sec) +INFO:tensorflow:global_step/sec: 3.24333 +INFO:tensorflow:step = 45501, loss = 0.255961, precision = 0.976562 (30.832 sec) +INFO:tensorflow:global_step/sec: 3.24205 +INFO:tensorflow:step = 45601, loss = 0.27323, precision = 0.945312 (30.845 sec) +INFO:tensorflow:global_step/sec: 3.24096 +INFO:tensorflow:step = 45701, loss = 0.256394, precision = 0.96875 (30.855 sec) +Saved checkpoint after 117 epoch(s) to ../data/resnet56/checkpoints/00117... +INFO:tensorflow:global_step/sec: 3.1647 +INFO:tensorflow:step = 45801, loss = 0.230734, precision = 0.976562 (31.599 sec) +INFO:tensorflow:global_step/sec: 3.24458 +INFO:tensorflow:step = 45901, loss = 0.223337, precision = 0.96875 (30.820 sec) +INFO:tensorflow:global_step/sec: 3.24516 +INFO:tensorflow:step = 46001, loss = 0.203696, precision = 0.984375 (30.815 sec) +INFO:tensorflow:global_step/sec: 3.24079 +INFO:tensorflow:step = 46101, loss = 0.201563, precision = 0.960938 (30.856 sec) +Saved checkpoint after 118 epoch(s) to ../data/resnet56/checkpoints/00118... +INFO:tensorflow:global_step/sec: 3.16332 +INFO:tensorflow:step = 46201, loss = 0.250585, precision = 0.960938 (31.613 sec) +INFO:tensorflow:global_step/sec: 3.24307 +INFO:tensorflow:step = 46301, loss = 0.201716, precision = 0.984375 (30.835 sec) +INFO:tensorflow:global_step/sec: 3.24293 +INFO:tensorflow:step = 46401, loss = 0.254717, precision = 0.976562 (30.836 sec) +INFO:tensorflow:global_step/sec: 3.2388 +INFO:tensorflow:step = 46501, loss = 0.209073, precision = 0.976562 (30.876 sec) +Saved checkpoint after 119 epoch(s) to ../data/resnet56/checkpoints/00119... +INFO:tensorflow:global_step/sec: 3.16335 +INFO:tensorflow:step = 46601, loss = 0.219174, precision = 0.976562 (31.612 sec) +INFO:tensorflow:global_step/sec: 3.2424 +INFO:tensorflow:step = 46701, loss = 0.209297, precision = 0.984375 (30.841 sec) +INFO:tensorflow:global_step/sec: 3.24437 +INFO:tensorflow:step = 46801, loss = 0.24793, precision = 0.96875 (30.823 sec) +INFO:tensorflow:global_step/sec: 3.23916 +INFO:tensorflow:step = 46901, loss = 0.218209, precision = 0.96875 (30.872 sec) +Saved checkpoint after 120 epoch(s) to ../data/resnet56/checkpoints/00120... +INFO:tensorflow:global_step/sec: 3.15565 +INFO:tensorflow:step = 47001, loss = 0.244112, precision = 0.96875 (31.689 sec) +INFO:tensorflow:global_step/sec: 3.24202 +INFO:tensorflow:step = 47101, loss = 0.218789, precision = 0.976562 (30.845 sec) +INFO:tensorflow:global_step/sec: 3.23944 +INFO:tensorflow:step = 47201, loss = 0.193993, precision = 0.976562 (30.869 sec) +INFO:tensorflow:global_step/sec: 3.23699 +INFO:tensorflow:step = 47301, loss = 0.221786, precision = 0.992188 (30.893 sec) +Saved checkpoint after 121 epoch(s) to ../data/resnet56/checkpoints/00121... +INFO:tensorflow:global_step/sec: 3.16121 +INFO:tensorflow:step = 47401, loss = 0.177585, precision = 0.992188 (31.633 sec) +INFO:tensorflow:global_step/sec: 3.24155 +INFO:tensorflow:step = 47501, loss = 0.234776, precision = 0.960938 (30.849 sec) +INFO:tensorflow:global_step/sec: 3.24148 +INFO:tensorflow:step = 47601, loss = 0.180146, precision = 0.984375 (30.850 sec) +INFO:tensorflow:global_step/sec: 3.24219 +INFO:tensorflow:step = 47701, loss = 0.197518, precision = 0.984375 (30.843 sec) +Saved checkpoint after 122 epoch(s) to ../data/resnet56/checkpoints/00122... +INFO:tensorflow:global_step/sec: 3.1517 +INFO:tensorflow:step = 47801, loss = 0.189155, precision = 0.992188 (31.729 sec) +INFO:tensorflow:global_step/sec: 3.2404 +INFO:tensorflow:step = 47901, loss = 0.205248, precision = 0.984375 (30.860 sec) +INFO:tensorflow:global_step/sec: 3.24173 +INFO:tensorflow:step = 48001, loss = 0.194979, precision = 0.976562 (30.848 sec) +Saved checkpoint after 123 epoch(s) to ../data/resnet56/checkpoints/00123... +INFO:tensorflow:global_step/sec: 3.15996 +INFO:tensorflow:step = 48101, loss = 0.260111, precision = 0.960938 (31.646 sec) +INFO:tensorflow:global_step/sec: 3.24194 +INFO:tensorflow:step = 48201, loss = 0.217241, precision = 0.960938 (30.846 sec) +INFO:tensorflow:global_step/sec: 3.24443 +INFO:tensorflow:step = 48301, loss = 0.227325, precision = 0.96875 (30.822 sec) +INFO:tensorflow:global_step/sec: 3.23886 +INFO:tensorflow:step = 48401, loss = 0.263482, precision = 0.960938 (30.875 sec) +Saved checkpoint after 124 epoch(s) to ../data/resnet56/checkpoints/00124... +INFO:tensorflow:global_step/sec: 3.16376 +INFO:tensorflow:step = 48501, loss = 0.202074, precision = 0.984375 (31.608 sec) +INFO:tensorflow:global_step/sec: 3.242 +INFO:tensorflow:step = 48601, loss = 0.221365, precision = 0.976562 (30.845 sec) +INFO:tensorflow:global_step/sec: 3.24451 +INFO:tensorflow:step = 48701, loss = 0.243408, precision = 0.976562 (30.822 sec) +INFO:tensorflow:global_step/sec: 3.2389 +INFO:tensorflow:step = 48801, loss = 0.24078, precision = 0.953125 (30.875 sec) +Saved checkpoint after 125 epoch(s) to ../data/resnet56/checkpoints/00125... +INFO:tensorflow:global_step/sec: 3.15955 +INFO:tensorflow:step = 48901, loss = 0.220436, precision = 0.96875 (31.650 sec) +INFO:tensorflow:global_step/sec: 3.24089 +INFO:tensorflow:step = 49001, loss = 0.193414, precision = 0.992188 (30.856 sec) +INFO:tensorflow:global_step/sec: 3.24 +INFO:tensorflow:step = 49101, loss = 0.207859, precision = 0.976562 (30.864 sec) +INFO:tensorflow:global_step/sec: 3.2404 +INFO:tensorflow:step = 49201, loss = 0.26986, precision = 0.96875 (30.861 sec) +Saved checkpoint after 126 epoch(s) to ../data/resnet56/checkpoints/00126... +INFO:tensorflow:global_step/sec: 3.16266 +INFO:tensorflow:step = 49301, loss = 0.207721, precision = 0.992188 (31.619 sec) +INFO:tensorflow:global_step/sec: 3.24561 +INFO:tensorflow:step = 49401, loss = 0.171, precision = 1.0 (30.811 sec) +INFO:tensorflow:global_step/sec: 3.2424 +INFO:tensorflow:step = 49501, loss = 0.169523, precision = 0.992188 (30.841 sec) +INFO:tensorflow:global_step/sec: 3.24296 +INFO:tensorflow:step = 49601, loss = 0.181963, precision = 0.984375 (30.836 sec) +Saved checkpoint after 127 epoch(s) to ../data/resnet56/checkpoints/00127... +INFO:tensorflow:global_step/sec: 3.15889 +INFO:tensorflow:step = 49701, loss = 0.207253, precision = 0.976562 (31.657 sec) +INFO:tensorflow:global_step/sec: 3.23939 +INFO:tensorflow:step = 49801, loss = 0.233587, precision = 0.976562 (30.870 sec) +INFO:tensorflow:global_step/sec: 3.23985 +INFO:tensorflow:step = 49901, loss = 0.236415, precision = 0.984375 (30.866 sec) +INFO:tensorflow:global_step/sec: 3.23925 +INFO:tensorflow:step = 50001, loss = 0.229633, precision = 0.953125 (30.871 sec) +Saved checkpoint after 128 epoch(s) to ../data/resnet56/checkpoints/00128... +INFO:tensorflow:global_step/sec: 3.15961 +INFO:tensorflow:step = 50101, loss = 0.22086, precision = 0.976562 (31.649 sec) +INFO:tensorflow:global_step/sec: 3.24268 +INFO:tensorflow:step = 50201, loss = 0.216317, precision = 0.96875 (30.839 sec) +INFO:tensorflow:global_step/sec: 3.24456 +INFO:tensorflow:step = 50301, loss = 0.167479, precision = 1.0 (30.821 sec) +INFO:tensorflow:global_step/sec: 3.24094 +INFO:tensorflow:step = 50401, loss = 0.174982, precision = 0.984375 (30.855 sec) +Saved checkpoint after 129 epoch(s) to ../data/resnet56/checkpoints/00129... +INFO:tensorflow:global_step/sec: 3.16054 +INFO:tensorflow:step = 50501, loss = 0.187408, precision = 0.984375 (31.640 sec) +INFO:tensorflow:global_step/sec: 3.2452 +INFO:tensorflow:step = 50601, loss = 0.150864, precision = 1.0 (30.815 sec) +INFO:tensorflow:global_step/sec: 3.24249 +INFO:tensorflow:step = 50701, loss = 0.213886, precision = 0.976562 (30.841 sec) +INFO:tensorflow:global_step/sec: 3.24183 +INFO:tensorflow:step = 50801, loss = 0.163714, precision = 1.0 (30.847 sec) +Saved checkpoint after 130 epoch(s) to ../data/resnet56/checkpoints/00130... +INFO:tensorflow:global_step/sec: 3.16127 +INFO:tensorflow:step = 50901, loss = 0.173573, precision = 0.992188 (31.632 sec) +INFO:tensorflow:global_step/sec: 3.24435 +INFO:tensorflow:step = 51001, loss = 0.179063, precision = 0.992188 (30.823 sec) +INFO:tensorflow:global_step/sec: 3.24222 +INFO:tensorflow:step = 51101, loss = 0.210804, precision = 0.976562 (30.843 sec) +INFO:tensorflow:global_step/sec: 3.24209 +INFO:tensorflow:step = 51201, loss = 0.170578, precision = 0.992188 (30.844 sec) +Saved checkpoint after 131 epoch(s) to ../data/resnet56/checkpoints/00131... +INFO:tensorflow:global_step/sec: 3.16357 +INFO:tensorflow:step = 51301, loss = 0.174211, precision = 0.984375 (31.610 sec) +INFO:tensorflow:global_step/sec: 3.24098 +INFO:tensorflow:step = 51401, loss = 0.221685, precision = 0.976562 (30.855 sec) +INFO:tensorflow:global_step/sec: 3.24141 +INFO:tensorflow:step = 51501, loss = 0.166492, precision = 1.0 (30.851 sec) +INFO:tensorflow:global_step/sec: 3.24304 +INFO:tensorflow:step = 51601, loss = 0.161983, precision = 0.984375 (30.835 sec) +Saved checkpoint after 132 epoch(s) to ../data/resnet56/checkpoints/00132... +INFO:tensorflow:global_step/sec: 3.16617 +INFO:tensorflow:step = 51701, loss = 0.321535, precision = 0.9375 (31.584 sec) +INFO:tensorflow:global_step/sec: 3.24259 +INFO:tensorflow:step = 51801, loss = 0.186804, precision = 0.984375 (30.839 sec) +INFO:tensorflow:global_step/sec: 3.23976 +INFO:tensorflow:step = 51901, loss = 0.165188, precision = 0.984375 (30.867 sec) +INFO:tensorflow:global_step/sec: 3.23959 +INFO:tensorflow:step = 52001, loss = 0.231805, precision = 0.96875 (30.868 sec) +Saved checkpoint after 133 epoch(s) to ../data/resnet56/checkpoints/00133... +INFO:tensorflow:global_step/sec: 3.15239 +INFO:tensorflow:step = 52101, loss = 0.205167, precision = 0.976562 (31.722 sec) +INFO:tensorflow:global_step/sec: 3.24154 +INFO:tensorflow:step = 52201, loss = 0.192785, precision = 0.984375 (30.849 sec) +INFO:tensorflow:global_step/sec: 3.24275 +INFO:tensorflow:step = 52301, loss = 0.197572, precision = 0.976562 (30.838 sec) +Saved checkpoint after 134 epoch(s) to ../data/resnet56/checkpoints/00134... +INFO:tensorflow:global_step/sec: 3.16391 +INFO:tensorflow:step = 52401, loss = 0.209209, precision = 0.976562 (31.606 sec) +INFO:tensorflow:global_step/sec: 3.24428 +INFO:tensorflow:step = 52501, loss = 0.205791, precision = 0.976562 (30.823 sec) +INFO:tensorflow:global_step/sec: 3.23913 +INFO:tensorflow:step = 52601, loss = 0.230311, precision = 0.976562 (30.872 sec) +INFO:tensorflow:global_step/sec: 3.24096 +INFO:tensorflow:step = 52701, loss = 0.17285, precision = 0.992188 (30.855 sec) +Saved checkpoint after 135 epoch(s) to ../data/resnet56/checkpoints/00135... +INFO:tensorflow:global_step/sec: 3.16493 +INFO:tensorflow:step = 52801, loss = 0.169081, precision = 0.992188 (31.596 sec) +INFO:tensorflow:global_step/sec: 3.2422 +INFO:tensorflow:step = 52901, loss = 0.188583, precision = 0.976562 (30.843 sec) +INFO:tensorflow:global_step/sec: 3.24121 +INFO:tensorflow:step = 53001, loss = 0.163636, precision = 0.992188 (30.853 sec) +INFO:tensorflow:global_step/sec: 3.23711 +INFO:tensorflow:step = 53101, loss = 0.199181, precision = 0.96875 (30.892 sec) +Saved checkpoint after 136 epoch(s) to ../data/resnet56/checkpoints/00136... +INFO:tensorflow:global_step/sec: 3.16793 +INFO:tensorflow:step = 53201, loss = 0.182717, precision = 0.976562 (31.567 sec) +INFO:tensorflow:global_step/sec: 3.24073 +INFO:tensorflow:step = 53301, loss = 0.19892, precision = 0.976562 (30.857 sec) +INFO:tensorflow:global_step/sec: 3.24193 +INFO:tensorflow:step = 53401, loss = 0.185697, precision = 0.976562 (30.846 sec) +INFO:tensorflow:global_step/sec: 3.24198 +INFO:tensorflow:step = 53501, loss = 0.196406, precision = 0.984375 (30.845 sec) +Saved checkpoint after 137 epoch(s) to ../data/resnet56/checkpoints/00137... +INFO:tensorflow:global_step/sec: 3.16901 +INFO:tensorflow:step = 53601, loss = 0.159716, precision = 0.992188 (31.556 sec) +INFO:tensorflow:global_step/sec: 3.24242 +INFO:tensorflow:step = 53701, loss = 0.142702, precision = 1.0 (30.841 sec) +INFO:tensorflow:global_step/sec: 3.24082 +INFO:tensorflow:step = 53801, loss = 0.179521, precision = 0.976562 (30.857 sec) +INFO:tensorflow:global_step/sec: 3.23748 +INFO:tensorflow:step = 53901, loss = 0.141144, precision = 1.0 (30.888 sec) +Saved checkpoint after 138 epoch(s) to ../data/resnet56/checkpoints/00138... +INFO:tensorflow:global_step/sec: 3.1623 +INFO:tensorflow:step = 54001, loss = 0.144863, precision = 1.0 (31.623 sec) +INFO:tensorflow:global_step/sec: 3.24006 +INFO:tensorflow:step = 54101, loss = 0.15042, precision = 1.0 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.23996 +INFO:tensorflow:step = 54201, loss = 0.161521, precision = 0.992188 (30.865 sec) +INFO:tensorflow:global_step/sec: 3.23897 +INFO:tensorflow:step = 54301, loss = 0.161405, precision = 0.992188 (30.874 sec) +Saved checkpoint after 139 epoch(s) to ../data/resnet56/checkpoints/00139... +INFO:tensorflow:global_step/sec: 3.16593 +INFO:tensorflow:step = 54401, loss = 0.157539, precision = 0.992188 (31.587 sec) +INFO:tensorflow:global_step/sec: 3.24439 +INFO:tensorflow:step = 54501, loss = 0.162583, precision = 0.992188 (30.822 sec) +INFO:tensorflow:global_step/sec: 3.2415 +INFO:tensorflow:step = 54601, loss = 0.147619, precision = 1.0 (30.850 sec) +INFO:tensorflow:global_step/sec: 3.24159 +INFO:tensorflow:step = 54701, loss = 0.161331, precision = 0.984375 (30.849 sec) +Saved checkpoint after 140 epoch(s) to ../data/resnet56/checkpoints/00140... +INFO:tensorflow:global_step/sec: 3.16783 +INFO:tensorflow:step = 54801, loss = 0.149496, precision = 1.0 (31.568 sec) +INFO:tensorflow:global_step/sec: 3.24147 +INFO:tensorflow:step = 54901, loss = 0.140041, precision = 1.0 (30.850 sec) +INFO:tensorflow:global_step/sec: 3.24133 +INFO:tensorflow:step = 55001, loss = 0.149844, precision = 1.0 (30.852 sec) +INFO:tensorflow:global_step/sec: 3.2382 +INFO:tensorflow:step = 55101, loss = 0.143082, precision = 1.0 (30.881 sec) +Saved checkpoint after 141 epoch(s) to ../data/resnet56/checkpoints/00141... +INFO:tensorflow:global_step/sec: 3.16332 +INFO:tensorflow:step = 55201, loss = 0.147142, precision = 1.0 (31.613 sec) +INFO:tensorflow:global_step/sec: 3.24424 +INFO:tensorflow:step = 55301, loss = 0.147143, precision = 1.0 (30.824 sec) +INFO:tensorflow:global_step/sec: 3.24315 +INFO:tensorflow:step = 55401, loss = 0.147955, precision = 0.992188 (30.834 sec) +INFO:tensorflow:global_step/sec: 3.24031 +INFO:tensorflow:step = 55501, loss = 0.154814, precision = 0.992188 (30.861 sec) +Saved checkpoint after 142 epoch(s) to ../data/resnet56/checkpoints/00142... +INFO:tensorflow:global_step/sec: 3.16517 +INFO:tensorflow:step = 55601, loss = 0.14057, precision = 1.0 (31.594 sec) +INFO:tensorflow:global_step/sec: 3.24215 +INFO:tensorflow:step = 55701, loss = 0.146011, precision = 1.0 (30.844 sec) +INFO:tensorflow:global_step/sec: 3.24082 +INFO:tensorflow:step = 55801, loss = 0.155291, precision = 0.992188 (30.856 sec) +INFO:tensorflow:global_step/sec: 3.23798 +INFO:tensorflow:step = 55901, loss = 0.169045, precision = 0.984375 (30.883 sec) +Saved checkpoint after 143 epoch(s) to ../data/resnet56/checkpoints/00143... +INFO:tensorflow:global_step/sec: 3.16162 +INFO:tensorflow:step = 56001, loss = 0.156302, precision = 0.992188 (31.630 sec) +INFO:tensorflow:global_step/sec: 3.24467 +INFO:tensorflow:step = 56101, loss = 0.15467, precision = 0.992188 (30.819 sec) +INFO:tensorflow:global_step/sec: 3.24137 +INFO:tensorflow:step = 56201, loss = 0.14595, precision = 1.0 (30.851 sec) +INFO:tensorflow:global_step/sec: 3.24152 +INFO:tensorflow:step = 56301, loss = 0.136951, precision = 1.0 (30.850 sec) +Saved checkpoint after 144 epoch(s) to ../data/resnet56/checkpoints/00144... +INFO:tensorflow:global_step/sec: 3.16986 +INFO:tensorflow:step = 56401, loss = 0.1405, precision = 1.0 (31.547 sec) +INFO:tensorflow:global_step/sec: 3.24011 +INFO:tensorflow:step = 56501, loss = 0.147862, precision = 1.0 (30.864 sec) +INFO:tensorflow:global_step/sec: 3.23966 +INFO:tensorflow:step = 56601, loss = 0.143113, precision = 1.0 (30.867 sec) +Saved checkpoint after 145 epoch(s) to ../data/resnet56/checkpoints/00145... +INFO:tensorflow:global_step/sec: 3.15494 +INFO:tensorflow:step = 56701, loss = 0.137699, precision = 1.0 (31.696 sec) +INFO:tensorflow:global_step/sec: 3.24266 +INFO:tensorflow:step = 56801, loss = 0.144378, precision = 1.0 (30.839 sec) +INFO:tensorflow:global_step/sec: 3.24193 +INFO:tensorflow:step = 56901, loss = 0.142548, precision = 1.0 (30.846 sec) +INFO:tensorflow:global_step/sec: 3.23808 +INFO:tensorflow:step = 57001, loss = 0.140606, precision = 1.0 (30.883 sec) +Saved checkpoint after 146 epoch(s) to ../data/resnet56/checkpoints/00146... +INFO:tensorflow:global_step/sec: 3.16361 +INFO:tensorflow:step = 57101, loss = 0.144226, precision = 1.0 (31.609 sec) +INFO:tensorflow:global_step/sec: 3.24101 +INFO:tensorflow:step = 57201, loss = 0.158806, precision = 0.984375 (30.855 sec) +INFO:tensorflow:global_step/sec: 3.24041 +INFO:tensorflow:step = 57301, loss = 0.144294, precision = 0.992188 (30.860 sec) +INFO:tensorflow:global_step/sec: 3.24303 +INFO:tensorflow:step = 57401, loss = 0.151962, precision = 0.992188 (30.835 sec) +Saved checkpoint after 147 epoch(s) to ../data/resnet56/checkpoints/00147... +INFO:tensorflow:global_step/sec: 3.16643 +INFO:tensorflow:step = 57501, loss = 0.145278, precision = 0.992188 (31.581 sec) +INFO:tensorflow:global_step/sec: 3.24478 +INFO:tensorflow:step = 57601, loss = 0.141317, precision = 1.0 (30.819 sec) +INFO:tensorflow:global_step/sec: 3.24176 +INFO:tensorflow:step = 57701, loss = 0.141033, precision = 1.0 (30.848 sec) +INFO:tensorflow:global_step/sec: 3.23996 +INFO:tensorflow:step = 57801, loss = 0.133259, precision = 1.0 (30.864 sec) +Saved checkpoint after 148 epoch(s) to ../data/resnet56/checkpoints/00148... +INFO:tensorflow:global_step/sec: 3.16661 +INFO:tensorflow:step = 57901, loss = 0.140729, precision = 1.0 (31.579 sec) +INFO:tensorflow:global_step/sec: 3.24097 +INFO:tensorflow:step = 58001, loss = 0.150229, precision = 0.992188 (30.855 sec) +INFO:tensorflow:global_step/sec: 3.23985 +INFO:tensorflow:step = 58101, loss = 0.156091, precision = 0.984375 (30.866 sec) +INFO:tensorflow:global_step/sec: 3.2381 +INFO:tensorflow:step = 58201, loss = 0.135989, precision = 1.0 (30.882 sec) +Saved checkpoint after 149 epoch(s) to ../data/resnet56/checkpoints/00149... +INFO:tensorflow:global_step/sec: 3.16815 +INFO:tensorflow:step = 58301, loss = 0.140153, precision = 1.0 (31.564 sec) +INFO:tensorflow:global_step/sec: 3.24279 +INFO:tensorflow:step = 58401, loss = 0.174742, precision = 0.992188 (30.838 sec) +INFO:tensorflow:global_step/sec: 3.23745 +INFO:tensorflow:step = 58501, loss = 0.139429, precision = 1.0 (30.888 sec) +INFO:tensorflow:global_step/sec: 3.24043 +INFO:tensorflow:step = 58601, loss = 0.132285, precision = 1.0 (30.860 sec) +Saved checkpoint after 150 epoch(s) to ../data/resnet56/checkpoints/00150... +INFO:tensorflow:global_step/sec: 3.16835 +INFO:tensorflow:step = 58701, loss = 0.132163, precision = 1.0 (31.562 sec) +INFO:tensorflow:global_step/sec: 3.24174 +INFO:tensorflow:step = 58801, loss = 0.139866, precision = 1.0 (30.848 sec) +INFO:tensorflow:global_step/sec: 3.23888 +INFO:tensorflow:step = 58901, loss = 0.132751, precision = 1.0 (30.875 sec) +INFO:tensorflow:global_step/sec: 3.2349 +INFO:tensorflow:step = 59001, loss = 0.133226, precision = 1.0 (30.913 sec) +Saved checkpoint after 151 epoch(s) to ../data/resnet56/checkpoints/00151... +INFO:tensorflow:global_step/sec: 3.16409 +INFO:tensorflow:step = 59101, loss = 0.134756, precision = 1.0 (31.605 sec) +INFO:tensorflow:global_step/sec: 3.24295 +INFO:tensorflow:step = 59201, loss = 0.133212, precision = 1.0 (30.836 sec) +INFO:tensorflow:global_step/sec: 3.24375 +INFO:tensorflow:step = 59301, loss = 0.131778, precision = 1.0 (30.828 sec) +INFO:tensorflow:global_step/sec: 3.24126 +INFO:tensorflow:step = 59401, loss = 0.131302, precision = 1.0 (30.852 sec) +Saved checkpoint after 152 epoch(s) to ../data/resnet56/checkpoints/00152... +INFO:tensorflow:global_step/sec: 3.16387 +INFO:tensorflow:step = 59501, loss = 0.139452, precision = 0.992188 (31.607 sec) +INFO:tensorflow:global_step/sec: 3.23964 +INFO:tensorflow:step = 59601, loss = 0.140884, precision = 0.992188 (30.868 sec) +INFO:tensorflow:global_step/sec: 3.23937 +INFO:tensorflow:step = 59701, loss = 0.135945, precision = 1.0 (30.870 sec) +INFO:tensorflow:global_step/sec: 3.23887 +INFO:tensorflow:step = 59801, loss = 0.131728, precision = 1.0 (30.875 sec) +Saved checkpoint after 153 epoch(s) to ../data/resnet56/checkpoints/00153... +INFO:tensorflow:global_step/sec: 3.16134 +INFO:tensorflow:step = 59901, loss = 0.128841, precision = 1.0 (31.632 sec) +INFO:tensorflow:global_step/sec: 3.23712 +INFO:tensorflow:step = 60001, loss = 0.133811, precision = 1.0 (30.892 sec) +INFO:tensorflow:global_step/sec: 3.23804 +INFO:tensorflow:step = 60101, loss = 0.131281, precision = 1.0 (30.883 sec) +INFO:tensorflow:global_step/sec: 3.23843 +INFO:tensorflow:step = 60201, loss = 0.139922, precision = 0.992188 (30.879 sec) +Saved checkpoint after 154 epoch(s) to ../data/resnet56/checkpoints/00154... +INFO:tensorflow:global_step/sec: 3.16501 +INFO:tensorflow:step = 60301, loss = 0.131831, precision = 1.0 (31.595 sec) +INFO:tensorflow:global_step/sec: 3.23936 +INFO:tensorflow:step = 60401, loss = 0.135698, precision = 1.0 (30.870 sec) +INFO:tensorflow:global_step/sec: 3.23812 +INFO:tensorflow:step = 60501, loss = 0.13436, precision = 1.0 (30.882 sec) +INFO:tensorflow:global_step/sec: 3.2405 +INFO:tensorflow:step = 60601, loss = 0.134066, precision = 1.0 (30.859 sec) +Saved checkpoint after 155 epoch(s) to ../data/resnet56/checkpoints/00155... +INFO:tensorflow:global_step/sec: 3.16253 +INFO:tensorflow:step = 60701, loss = 0.132113, precision = 1.0 (31.620 sec) +INFO:tensorflow:global_step/sec: 3.23841 +INFO:tensorflow:step = 60801, loss = 0.132252, precision = 1.0 (30.880 sec) +INFO:tensorflow:global_step/sec: 3.23923 +INFO:tensorflow:step = 60901, loss = 0.130565, precision = 1.0 (30.871 sec) +Saved checkpoint after 156 epoch(s) to ../data/resnet56/checkpoints/00156... +INFO:tensorflow:global_step/sec: 3.16412 +INFO:tensorflow:step = 61001, loss = 0.129386, precision = 1.0 (31.604 sec) +INFO:tensorflow:global_step/sec: 3.24217 +INFO:tensorflow:step = 61101, loss = 0.132344, precision = 1.0 (30.843 sec) +INFO:tensorflow:global_step/sec: 3.24043 +INFO:tensorflow:step = 61201, loss = 0.137972, precision = 0.992188 (30.860 sec) +INFO:tensorflow:global_step/sec: 3.23856 +INFO:tensorflow:step = 61301, loss = 0.136821, precision = 1.0 (30.878 sec) +Saved checkpoint after 157 epoch(s) to ../data/resnet56/checkpoints/00157... +INFO:tensorflow:global_step/sec: 3.15545 +INFO:tensorflow:step = 61401, loss = 0.130387, precision = 1.0 (31.691 sec) +INFO:tensorflow:global_step/sec: 3.23919 +INFO:tensorflow:step = 61501, loss = 0.136834, precision = 1.0 (30.872 sec) +INFO:tensorflow:global_step/sec: 3.24071 +INFO:tensorflow:step = 61601, loss = 0.143692, precision = 0.992188 (30.858 sec) +INFO:tensorflow:global_step/sec: 3.24247 +INFO:tensorflow:step = 61701, loss = 0.12924, precision = 1.0 (30.840 sec) +Saved checkpoint after 158 epoch(s) to ../data/resnet56/checkpoints/00158... +INFO:tensorflow:global_step/sec: 3.16321 +INFO:tensorflow:step = 61801, loss = 0.128742, precision = 1.0 (31.614 sec) +INFO:tensorflow:global_step/sec: 3.24255 +INFO:tensorflow:step = 61901, loss = 0.131408, precision = 1.0 (30.840 sec) +INFO:tensorflow:global_step/sec: 3.23839 +INFO:tensorflow:step = 62001, loss = 0.129278, precision = 1.0 (30.880 sec) +INFO:tensorflow:global_step/sec: 3.23938 +INFO:tensorflow:step = 62101, loss = 0.127918, precision = 1.0 (30.870 sec) +Saved checkpoint after 159 epoch(s) to ../data/resnet56/checkpoints/00159... +INFO:tensorflow:global_step/sec: 3.16294 +INFO:tensorflow:step = 62201, loss = 0.130575, precision = 1.0 (31.616 sec) +INFO:tensorflow:global_step/sec: 3.24062 +INFO:tensorflow:step = 62301, loss = 0.134417, precision = 0.992188 (30.858 sec) +INFO:tensorflow:global_step/sec: 3.23855 +INFO:tensorflow:step = 62401, loss = 0.150861, precision = 0.992188 (30.878 sec) +INFO:tensorflow:global_step/sec: 3.23948 +INFO:tensorflow:step = 62501, loss = 0.127187, precision = 1.0 (30.869 sec) +Saved checkpoint after 160 epoch(s) to ../data/resnet56/checkpoints/00160... +INFO:tensorflow:global_step/sec: 3.16449 +INFO:tensorflow:step = 62601, loss = 0.126784, precision = 1.0 (31.601 sec) +INFO:tensorflow:global_step/sec: 3.2423 +INFO:tensorflow:step = 62701, loss = 0.127741, precision = 1.0 (30.842 sec) +INFO:tensorflow:global_step/sec: 3.23945 +INFO:tensorflow:step = 62801, loss = 0.128523, precision = 1.0 (30.869 sec) +INFO:tensorflow:global_step/sec: 3.24093 +INFO:tensorflow:step = 62901, loss = 0.128228, precision = 1.0 (30.856 sec) +Saved checkpoint after 161 epoch(s) to ../data/resnet56/checkpoints/00161... +INFO:tensorflow:global_step/sec: 3.16383 +INFO:tensorflow:step = 63001, loss = 0.127853, precision = 1.0 (31.607 sec) +INFO:tensorflow:global_step/sec: 3.23965 +INFO:tensorflow:step = 63101, loss = 0.126015, precision = 1.0 (30.868 sec) +INFO:tensorflow:global_step/sec: 3.24192 +INFO:tensorflow:step = 63201, loss = 0.131208, precision = 1.0 (30.845 sec) +INFO:tensorflow:global_step/sec: 3.2408 +INFO:tensorflow:step = 63301, loss = 0.126065, precision = 1.0 (30.857 sec) +Saved checkpoint after 162 epoch(s) to ../data/resnet56/checkpoints/00162... +INFO:tensorflow:global_step/sec: 3.16466 +INFO:tensorflow:step = 63401, loss = 0.12784, precision = 1.0 (31.599 sec) +INFO:tensorflow:global_step/sec: 3.24178 +INFO:tensorflow:step = 63501, loss = 0.131252, precision = 1.0 (30.847 sec) +INFO:tensorflow:global_step/sec: 3.24155 +INFO:tensorflow:step = 63601, loss = 0.126438, precision = 1.0 (30.850 sec) +INFO:tensorflow:global_step/sec: 3.242 +INFO:tensorflow:step = 63701, loss = 0.131, precision = 1.0 (30.845 sec) +Saved checkpoint after 163 epoch(s) to ../data/resnet56/checkpoints/00163... +INFO:tensorflow:global_step/sec: 3.17138 +INFO:tensorflow:step = 63801, loss = 0.130663, precision = 1.0 (31.532 sec) +INFO:tensorflow:global_step/sec: 3.24408 +INFO:tensorflow:step = 63901, loss = 0.126621, precision = 1.0 (30.825 sec) +INFO:tensorflow:global_step/sec: 3.24212 +INFO:tensorflow:step = 64001, loss = 0.126514, precision = 1.0 (30.844 sec) +INFO:tensorflow:global_step/sec: 3.24471 +INFO:tensorflow:step = 64101, loss = 0.12593, precision = 1.0 (30.820 sec) +Saved checkpoint after 164 epoch(s) to ../data/resnet56/checkpoints/00164... +INFO:tensorflow:global_step/sec: 3.16809 +INFO:tensorflow:step = 64201, loss = 0.125951, precision = 1.0 (31.565 sec) +INFO:tensorflow:global_step/sec: 3.24303 +INFO:tensorflow:step = 64301, loss = 0.12604, precision = 1.0 (30.836 sec) +INFO:tensorflow:global_step/sec: 3.23863 +INFO:tensorflow:step = 64401, loss = 0.126105, precision = 1.0 (30.877 sec) +INFO:tensorflow:global_step/sec: 3.24582 +INFO:tensorflow:step = 64501, loss = 0.128598, precision = 1.0 (30.809 sec) +Saved checkpoint after 165 epoch(s) to ../data/resnet56/checkpoints/00165... +INFO:tensorflow:global_step/sec: 3.16674 +INFO:tensorflow:step = 64601, loss = 0.128971, precision = 1.0 (31.579 sec) +INFO:tensorflow:global_step/sec: 3.24066 +INFO:tensorflow:step = 64701, loss = 0.130334, precision = 1.0 (30.858 sec) +INFO:tensorflow:global_step/sec: 3.24014 +INFO:tensorflow:step = 64801, loss = 0.129712, precision = 1.0 (30.863 sec) +INFO:tensorflow:global_step/sec: 3.24342 +INFO:tensorflow:step = 64901, loss = 0.126277, precision = 1.0 (30.831 sec) +Saved checkpoint after 166 epoch(s) to ../data/resnet56/checkpoints/00166... +INFO:tensorflow:global_step/sec: 3.16586 +INFO:tensorflow:step = 65001, loss = 0.12866, precision = 1.0 (31.587 sec) +INFO:tensorflow:global_step/sec: 3.24253 +INFO:tensorflow:step = 65101, loss = 0.129372, precision = 1.0 (30.840 sec) +INFO:tensorflow:global_step/sec: 3.24405 +INFO:tensorflow:step = 65201, loss = 0.12816, precision = 1.0 (30.826 sec) +Saved checkpoint after 167 epoch(s) to ../data/resnet56/checkpoints/00167... +INFO:tensorflow:global_step/sec: 3.16548 +INFO:tensorflow:step = 65301, loss = 0.123479, precision = 1.0 (31.591 sec) +INFO:tensorflow:global_step/sec: 3.24052 +INFO:tensorflow:step = 65401, loss = 0.124052, precision = 1.0 (30.859 sec) +INFO:tensorflow:global_step/sec: 3.2414 +INFO:tensorflow:step = 65501, loss = 0.125829, precision = 1.0 (30.851 sec) +INFO:tensorflow:global_step/sec: 3.24189 +INFO:tensorflow:step = 65601, loss = 0.132243, precision = 1.0 (30.846 sec) +Saved checkpoint after 168 epoch(s) to ../data/resnet56/checkpoints/00168... +INFO:tensorflow:global_step/sec: 3.16262 +INFO:tensorflow:step = 65701, loss = 0.12364, precision = 1.0 (31.619 sec) +INFO:tensorflow:global_step/sec: 3.24379 +INFO:tensorflow:step = 65801, loss = 0.127372, precision = 1.0 (30.828 sec) +INFO:tensorflow:global_step/sec: 3.2403 +INFO:tensorflow:step = 65901, loss = 0.122571, precision = 1.0 (30.861 sec) +INFO:tensorflow:global_step/sec: 3.24208 +INFO:tensorflow:step = 66001, loss = 0.126882, precision = 1.0 (30.845 sec) +Saved checkpoint after 169 epoch(s) to ../data/resnet56/checkpoints/00169... +INFO:tensorflow:global_step/sec: 3.1689 +INFO:tensorflow:step = 66101, loss = 0.125702, precision = 1.0 (31.557 sec) +INFO:tensorflow:global_step/sec: 3.24075 +INFO:tensorflow:step = 66201, loss = 0.12379, precision = 1.0 (30.857 sec) +INFO:tensorflow:global_step/sec: 3.23983 +INFO:tensorflow:step = 66301, loss = 0.127193, precision = 1.0 (30.866 sec) +INFO:tensorflow:global_step/sec: 3.2414 +INFO:tensorflow:step = 66401, loss = 0.125619, precision = 1.0 (30.851 sec) +Saved checkpoint after 170 epoch(s) to ../data/resnet56/checkpoints/00170... +INFO:tensorflow:global_step/sec: 3.16118 +INFO:tensorflow:step = 66501, loss = 0.122894, precision = 1.0 (31.634 sec) +INFO:tensorflow:global_step/sec: 3.24161 +INFO:tensorflow:step = 66601, loss = 0.121865, precision = 1.0 (30.849 sec) +INFO:tensorflow:global_step/sec: 3.23879 +INFO:tensorflow:step = 66701, loss = 0.138981, precision = 0.992188 (30.876 sec) +INFO:tensorflow:global_step/sec: 3.24032 +INFO:tensorflow:step = 66801, loss = 0.123134, precision = 1.0 (30.861 sec) +Saved checkpoint after 171 epoch(s) to ../data/resnet56/checkpoints/00171... +INFO:tensorflow:global_step/sec: 3.16305 +INFO:tensorflow:step = 66901, loss = 0.125654, precision = 1.0 (31.615 sec) +INFO:tensorflow:global_step/sec: 3.24128 +INFO:tensorflow:step = 67001, loss = 0.122699, precision = 1.0 (30.852 sec) +INFO:tensorflow:global_step/sec: 3.23921 +INFO:tensorflow:step = 67101, loss = 0.122814, precision = 1.0 (30.872 sec) +INFO:tensorflow:global_step/sec: 3.24135 +INFO:tensorflow:step = 67201, loss = 0.127184, precision = 1.0 (30.851 sec) +Saved checkpoint after 172 epoch(s) to ../data/resnet56/checkpoints/00172... +INFO:tensorflow:global_step/sec: 3.16913 +INFO:tensorflow:step = 67301, loss = 0.121107, precision = 1.0 (31.554 sec) +INFO:tensorflow:global_step/sec: 3.23987 +INFO:tensorflow:step = 67401, loss = 0.128079, precision = 1.0 (30.866 sec) +INFO:tensorflow:global_step/sec: 3.24206 +INFO:tensorflow:step = 67501, loss = 0.121451, precision = 1.0 (30.844 sec) +INFO:tensorflow:global_step/sec: 3.24273 +INFO:tensorflow:step = 67601, loss = 0.124922, precision = 1.0 (30.838 sec) +Saved checkpoint after 173 epoch(s) to ../data/resnet56/checkpoints/00173... +INFO:tensorflow:global_step/sec: 3.16187 +INFO:tensorflow:step = 67701, loss = 0.123114, precision = 1.0 (31.627 sec) +INFO:tensorflow:global_step/sec: 3.2386 +INFO:tensorflow:step = 67801, loss = 0.132863, precision = 1.0 (30.878 sec) +INFO:tensorflow:global_step/sec: 3.23998 +INFO:tensorflow:step = 67901, loss = 0.122244, precision = 1.0 (30.864 sec) +INFO:tensorflow:global_step/sec: 3.24509 +INFO:tensorflow:step = 68001, loss = 0.121527, precision = 1.0 (30.816 sec) +Saved checkpoint after 174 epoch(s) to ../data/resnet56/checkpoints/00174... +INFO:tensorflow:global_step/sec: 3.16542 +INFO:tensorflow:step = 68101, loss = 0.127519, precision = 1.0 (31.591 sec) +INFO:tensorflow:global_step/sec: 3.24149 +INFO:tensorflow:step = 68201, loss = 0.127829, precision = 1.0 (30.850 sec) +INFO:tensorflow:global_step/sec: 3.24198 +INFO:tensorflow:step = 68301, loss = 0.12042, precision = 1.0 (30.845 sec) +INFO:tensorflow:global_step/sec: 3.24397 +INFO:tensorflow:step = 68401, loss = 0.148026, precision = 0.984375 (30.827 sec) +Saved checkpoint after 175 epoch(s) to ../data/resnet56/checkpoints/00175... +INFO:tensorflow:global_step/sec: 3.16282 +INFO:tensorflow:step = 68501, loss = 0.123977, precision = 1.0 (31.617 sec) +INFO:tensorflow:global_step/sec: 3.24065 +INFO:tensorflow:step = 68601, loss = 0.121186, precision = 1.0 (30.858 sec) +INFO:tensorflow:global_step/sec: 3.2417 +INFO:tensorflow:step = 68701, loss = 0.124215, precision = 1.0 (30.848 sec) +INFO:tensorflow:global_step/sec: 3.24242 +INFO:tensorflow:step = 68801, loss = 0.119475, precision = 1.0 (30.841 sec) +Saved checkpoint after 176 epoch(s) to ../data/resnet56/checkpoints/00176... +INFO:tensorflow:global_step/sec: 3.15408 +INFO:tensorflow:step = 68901, loss = 0.11976, precision = 1.0 (31.705 sec) +INFO:tensorflow:global_step/sec: 3.24243 +INFO:tensorflow:step = 69001, loss = 0.126662, precision = 1.0 (30.841 sec) +INFO:tensorflow:global_step/sec: 3.2432 +INFO:tensorflow:step = 69101, loss = 0.121164, precision = 1.0 (30.834 sec) +INFO:tensorflow:global_step/sec: 3.24244 +INFO:tensorflow:step = 69201, loss = 0.121864, precision = 1.0 (30.841 sec) +Saved checkpoint after 177 epoch(s) to ../data/resnet56/checkpoints/00177... +INFO:tensorflow:global_step/sec: 3.1615 +INFO:tensorflow:step = 69301, loss = 0.121023, precision = 1.0 (31.630 sec) +INFO:tensorflow:global_step/sec: 3.23969 +INFO:tensorflow:step = 69401, loss = 0.121221, precision = 1.0 (30.867 sec) +INFO:tensorflow:global_step/sec: 3.23927 +INFO:tensorflow:step = 69501, loss = 0.119854, precision = 1.0 (30.871 sec) +Saved checkpoint after 178 epoch(s) to ../data/resnet56/checkpoints/00178... +INFO:tensorflow:global_step/sec: 3.16158 +INFO:tensorflow:step = 69601, loss = 0.118799, precision = 1.0 (31.630 sec) +INFO:tensorflow:global_step/sec: 3.24171 +INFO:tensorflow:step = 69701, loss = 0.119236, precision = 1.0 (30.848 sec) +INFO:tensorflow:global_step/sec: 3.24149 +INFO:tensorflow:step = 69801, loss = 0.118191, precision = 1.0 (30.850 sec) +INFO:tensorflow:global_step/sec: 3.24339 +INFO:tensorflow:step = 69901, loss = 0.118904, precision = 1.0 (30.832 sec) +Saved checkpoint after 179 epoch(s) to ../data/resnet56/checkpoints/00179... +INFO:tensorflow:global_step/sec: 3.16051 +INFO:tensorflow:step = 70001, loss = 0.118634, precision = 1.0 (31.641 sec) +INFO:tensorflow:global_step/sec: 3.24158 +INFO:tensorflow:step = 70101, loss = 0.119189, precision = 1.0 (30.849 sec) +INFO:tensorflow:global_step/sec: 3.23911 +INFO:tensorflow:step = 70201, loss = 0.119474, precision = 1.0 (30.872 sec) +INFO:tensorflow:global_step/sec: 3.24063 +INFO:tensorflow:step = 70301, loss = 0.11997, precision = 1.0 (30.858 sec) +Saved checkpoint after 180 epoch(s) to ../data/resnet56/checkpoints/00180... +INFO:tensorflow:global_step/sec: 3.15848 +INFO:tensorflow:step = 70401, loss = 0.118213, precision = 1.0 (31.661 sec) +INFO:tensorflow:global_step/sec: 3.24156 +INFO:tensorflow:step = 70501, loss = 0.117607, precision = 1.0 (30.849 sec) +INFO:tensorflow:global_step/sec: 3.24007 +INFO:tensorflow:step = 70601, loss = 0.124073, precision = 1.0 (30.864 sec) +INFO:tensorflow:global_step/sec: 3.24329 +INFO:tensorflow:step = 70701, loss = 0.117987, precision = 1.0 (30.833 sec) +Saved checkpoint after 181 epoch(s) to ../data/resnet56/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1p100_dawn/resnet164_b_train.log b/tensorflow/CIFAR10/logs/1p100_dawn/resnet164_b_train.log new file mode 100644 index 0000000..9721be9 --- /dev/null +++ b/tensorflow/CIFAR10/logs/1p100_dawn/resnet164_b_train.log @@ -0,0 +1,2221 @@ +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 64, 32, 32) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 128, 16, 16) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +INFO:tensorflow:image after unit (128, 256, 8, 8) +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/1.69m params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (256x10, 2.56k/2.56k params) + logit/biases (10, 10/10 params) + unit_1_0/common_bn_relu/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (1x1x16x16, 256/256 params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub3/bn3/beta (16, 16/16 params) + unit_1_0/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_0/sub_add/project/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_1/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_1/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub3/bn3/beta (16, 16/16 params) + unit_1_1/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_10/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_10/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_10/sub2/bn2/beta (16, 16/16 params) + unit_1_10/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/sub3/bn3/beta (16, 16/16 params) + unit_1_10/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_11/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_11/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_11/sub2/bn2/beta (16, 16/16 params) + unit_1_11/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/sub3/bn3/beta (16, 16/16 params) + unit_1_11/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_12/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_12/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_12/sub2/bn2/beta (16, 16/16 params) + unit_1_12/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/sub3/bn3/beta (16, 16/16 params) + unit_1_12/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_13/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_13/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_13/sub2/bn2/beta (16, 16/16 params) + unit_1_13/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/sub3/bn3/beta (16, 16/16 params) + unit_1_13/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_14/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_14/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_14/sub2/bn2/beta (16, 16/16 params) + unit_1_14/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/sub3/bn3/beta (16, 16/16 params) + unit_1_14/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_15/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_15/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_15/sub2/bn2/beta (16, 16/16 params) + unit_1_15/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/sub3/bn3/beta (16, 16/16 params) + unit_1_15/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_16/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_16/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_16/sub2/bn2/beta (16, 16/16 params) + unit_1_16/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/sub3/bn3/beta (16, 16/16 params) + unit_1_16/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_17/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_17/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_17/sub2/bn2/beta (16, 16/16 params) + unit_1_17/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/sub3/bn3/beta (16, 16/16 params) + unit_1_17/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_2/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_2/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub3/bn3/beta (16, 16/16 params) + unit_1_2/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_3/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_3/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub3/bn3/beta (16, 16/16 params) + unit_1_3/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_4/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_4/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub3/bn3/beta (16, 16/16 params) + unit_1_4/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_5/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_5/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub3/bn3/beta (16, 16/16 params) + unit_1_5/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_6/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_6/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub3/bn3/beta (16, 16/16 params) + unit_1_6/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_7/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_7/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub3/bn3/beta (16, 16/16 params) + unit_1_7/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_8/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_8/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub3/bn3/beta (16, 16/16 params) + unit_1_8/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_1_9/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_1_9/sub1/conv1/DW (1x1x64x16, 1.02k/1.02k params) + unit_1_9/sub2/bn2/beta (16, 16/16 params) + unit_1_9/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/sub3/bn3/beta (16, 16/16 params) + unit_1_9/sub3/conv3/DW (1x1x16x64, 1.02k/1.02k params) + unit_2_0/residual_bn_relu/init_bn/beta (64, 64/64 params) + unit_2_0/sub1/conv1/DW (1x1x64x32, 2.05k/2.05k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_0/sub3/bn3/beta (32, 32/32 params) + unit_2_0/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_0/sub_add/project/DW (1x1x64x128, 8.19k/8.19k params) + unit_2_1/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_1/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub3/bn3/beta (32, 32/32 params) + unit_2_1/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_10/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_10/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_10/sub2/bn2/beta (32, 32/32 params) + unit_2_10/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/sub3/bn3/beta (32, 32/32 params) + unit_2_10/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_11/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_11/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_11/sub2/bn2/beta (32, 32/32 params) + unit_2_11/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/sub3/bn3/beta (32, 32/32 params) + unit_2_11/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_12/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_12/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_12/sub2/bn2/beta (32, 32/32 params) + unit_2_12/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/sub3/bn3/beta (32, 32/32 params) + unit_2_12/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_13/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_13/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_13/sub2/bn2/beta (32, 32/32 params) + unit_2_13/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/sub3/bn3/beta (32, 32/32 params) + unit_2_13/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_14/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_14/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_14/sub2/bn2/beta (32, 32/32 params) + unit_2_14/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/sub3/bn3/beta (32, 32/32 params) + unit_2_14/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_15/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_15/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_15/sub2/bn2/beta (32, 32/32 params) + unit_2_15/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/sub3/bn3/beta (32, 32/32 params) + unit_2_15/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_16/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_16/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_16/sub2/bn2/beta (32, 32/32 params) + unit_2_16/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/sub3/bn3/beta (32, 32/32 params) + unit_2_16/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_17/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_17/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_17/sub2/bn2/beta (32, 32/32 params) + unit_2_17/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/sub3/bn3/beta (32, 32/32 params) + unit_2_17/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_2/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_2/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub3/bn3/beta (32, 32/32 params) + unit_2_2/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_3/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_3/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub3/bn3/beta (32, 32/32 params) + unit_2_3/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_4/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_4/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub3/bn3/beta (32, 32/32 params) + unit_2_4/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_5/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_5/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub3/bn3/beta (32, 32/32 params) + unit_2_5/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_6/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_6/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub3/bn3/beta (32, 32/32 params) + unit_2_6/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_7/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_7/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub3/bn3/beta (32, 32/32 params) + unit_2_7/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_8/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_8/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub3/bn3/beta (32, 32/32 params) + unit_2_8/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_2_9/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_2_9/sub1/conv1/DW (1x1x128x32, 4.10k/4.10k params) + unit_2_9/sub2/bn2/beta (32, 32/32 params) + unit_2_9/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/sub3/bn3/beta (32, 32/32 params) + unit_2_9/sub3/conv3/DW (1x1x32x128, 4.10k/4.10k params) + unit_3_0/residual_bn_relu/init_bn/beta (128, 128/128 params) + unit_3_0/sub1/conv1/DW (1x1x128x64, 8.19k/8.19k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_0/sub3/bn3/beta (64, 64/64 params) + unit_3_0/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_0/sub_add/project/DW (1x1x128x256, 32.77k/32.77k params) + unit_3_1/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_1/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub3/bn3/beta (64, 64/64 params) + unit_3_1/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_10/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_10/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_10/sub2/bn2/beta (64, 64/64 params) + unit_3_10/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/sub3/bn3/beta (64, 64/64 params) + unit_3_10/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_11/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_11/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_11/sub2/bn2/beta (64, 64/64 params) + unit_3_11/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/sub3/bn3/beta (64, 64/64 params) + unit_3_11/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_12/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_12/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_12/sub2/bn2/beta (64, 64/64 params) + unit_3_12/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/sub3/bn3/beta (64, 64/64 params) + unit_3_12/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_13/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_13/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_13/sub2/bn2/beta (64, 64/64 params) + unit_3_13/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/sub3/bn3/beta (64, 64/64 params) + unit_3_13/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_14/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_14/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_14/sub2/bn2/beta (64, 64/64 params) + unit_3_14/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/sub3/bn3/beta (64, 64/64 params) + unit_3_14/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_15/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_15/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_15/sub2/bn2/beta (64, 64/64 params) + unit_3_15/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/sub3/bn3/beta (64, 64/64 params) + unit_3_15/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_16/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_16/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_16/sub2/bn2/beta (64, 64/64 params) + unit_3_16/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/sub3/bn3/beta (64, 64/64 params) + unit_3_16/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_17/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_17/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_17/sub2/bn2/beta (64, 64/64 params) + unit_3_17/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/sub3/bn3/beta (64, 64/64 params) + unit_3_17/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_2/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_2/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub3/bn3/beta (64, 64/64 params) + unit_3_2/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_3/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_3/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub3/bn3/beta (64, 64/64 params) + unit_3_3/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_4/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_4/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub3/bn3/beta (64, 64/64 params) + unit_3_4/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_5/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_5/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub3/bn3/beta (64, 64/64 params) + unit_3_5/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_6/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_6/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub3/bn3/beta (64, 64/64 params) + unit_3_6/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_7/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_7/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub3/bn3/beta (64, 64/64 params) + unit_3_7/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_8/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_8/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub3/bn3/beta (64, 64/64 params) + unit_3_8/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_3_9/residual_bn_relu/init_bn/beta (256, 256/256 params) + unit_3_9/sub1/conv1/DW (1x1x256x64, 16.38k/16.38k params) + unit_3_9/sub2/bn2/beta (64, 64/64 params) + unit_3_9/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/sub3/bn3/beta (64, 64/64 params) + unit_3_9/sub3/conv3/DW (1x1x64x256, 16.38k/16.38k params) + unit_last/final_bn/beta (256, 256/256 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/62.59b flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub_add/project/Conv2D (536.87m/536.87m flops) + unit_2_0/sub_add/project/Conv2D (536.87m/536.87m flops) + unit_3_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_3_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_3_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_7/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_6/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_6/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_5/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_16/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_4/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_0/sub_add/project/Conv2D (268.44m/268.44m flops) + unit_1_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_8/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_5/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_4/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_3/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_3/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_2/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_2/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_17/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_17/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_16/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_15/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_15/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_14/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_14/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_13/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_12/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_7/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_8/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_1_9/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_1_9/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_0/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_1/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_1/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_10/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_10/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_11/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_11/sub3/conv3/Conv2D (268.44m/268.44m flops) + unit_2_12/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_13/sub1/conv1/Conv2D (268.44m/268.44m flops) + unit_2_0/sub1/conv1/Conv2D (134.22m/134.22m flops) + unit_3_0/sub1/conv1/Conv2D (134.22m/134.22m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + unit_1_0/sub1/conv1/Conv2D (67.11m/67.11m flops) + logit/xw_plus_b (1.28k/656.64k flops) + logit/xw_plus_b/MatMul (655.36k/655.36k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (655.36k/655.36k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (655.36k/655.36k flops) + +======================End of Report========================== +2017-08-03 12:51:05.622292: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla P100-PCIE-16GB +major: 6 minor: 0 memoryClockRate (GHz) 1.3285 +pciBusID 0000:05:00.0 +Total memory: 15.89GiB +Free memory: 15.61GiB +2017-08-03 12:51:05.622402: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-08-03 12:51:05.622419: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-08-03 12:51:05.622441: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:05:00.0) +2017-08-03 12:51:07.235773: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-08-03 12:51:07.235851: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 56 visible devices +2017-08-03 12:51:07.258990: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x7d679e0 executing computations on platform Host. Devices: +2017-08-03 12:51:07.259044: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-08-03 12:51:07.259446: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-08-03 12:51:07.259473: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 56 visible devices +2017-08-03 12:51:07.277990: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x89fa380 executing computations on platform CUDA. Devices: +2017-08-03 12:51:07.278078: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla P100-PCIE-16GB, Compute Capability 6.0 +2017-08-03 12:51:08.562368: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:247] PoolAllocator: After 1152 get requests, put_count=1100 evicted_count=1000 eviction_rate=0.909091 and unsatisfied allocation rate=1 +2017-08-03 12:51:08.562493: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:259] Raising pool_size_limit_ from 100 to 110 +INFO:tensorflow:step = 1, loss = 8.45936, precision = 0.09375 +INFO:tensorflow:global_step/sec: 4.15542 +INFO:tensorflow:step = 101, loss = 7.98056, precision = 0.273438 (24.067 sec) +INFO:tensorflow:global_step/sec: 4.51382 +INFO:tensorflow:step = 201, loss = 7.58357, precision = 0.429688 (22.156 sec) +INFO:tensorflow:global_step/sec: 4.51098 +INFO:tensorflow:step = 301, loss = 7.54678, precision = 0.40625 (22.166 sec) +total_params: 1691146 +Saved checkpoint after 1 epoch(s) to data/resnet164_b/checkpoints/00001... +INFO:tensorflow:global_step/sec: 4.02047 +INFO:tensorflow:step = 401, loss = 7.67127, precision = 0.34375 (24.873 sec) +INFO:tensorflow:global_step/sec: 4.5135 +INFO:tensorflow:step = 501, loss = 6.56055, precision = 0.523438 (22.156 sec) +INFO:tensorflow:global_step/sec: 4.50527 +INFO:tensorflow:step = 601, loss = 6.07623, precision = 0.507812 (22.196 sec) +INFO:tensorflow:global_step/sec: 4.48989 +INFO:tensorflow:step = 701, loss = 5.56229, precision = 0.554688 (22.272 sec) +Saved checkpoint after 2 epoch(s) to data/resnet164_b/checkpoints/00002... +INFO:tensorflow:global_step/sec: 3.98997 +INFO:tensorflow:step = 801, loss = 4.98226, precision = 0.632812 (25.063 sec) +INFO:tensorflow:global_step/sec: 4.50012 +INFO:tensorflow:step = 901, loss = 4.65794, precision = 0.625 (22.221 sec) +INFO:tensorflow:global_step/sec: 4.5048 +INFO:tensorflow:step = 1001, loss = 4.06321, precision = 0.710938 (22.198 sec) +INFO:tensorflow:global_step/sec: 4.48449 +INFO:tensorflow:step = 1101, loss = 3.88422, precision = 0.648438 (22.299 sec) +Saved checkpoint after 3 epoch(s) to data/resnet164_b/checkpoints/00003... +INFO:tensorflow:global_step/sec: 4.06993 +INFO:tensorflow:step = 1201, loss = 3.58265, precision = 0.703125 (24.572 sec) +INFO:tensorflow:global_step/sec: 4.48979 +INFO:tensorflow:step = 1301, loss = 3.36457, precision = 0.640625 (22.272 sec) +INFO:tensorflow:global_step/sec: 4.49126 +INFO:tensorflow:step = 1401, loss = 3.12786, precision = 0.695312 (22.265 sec) +INFO:tensorflow:global_step/sec: 4.49982 +INFO:tensorflow:step = 1501, loss = 2.82074, precision = 0.773438 (22.223 sec) +Saved checkpoint after 4 epoch(s) to data/resnet164_b/checkpoints/00004... +INFO:tensorflow:global_step/sec: 4.06501 +INFO:tensorflow:step = 1601, loss = 2.44694, precision = 0.820312 (24.600 sec) +INFO:tensorflow:global_step/sec: 4.50994 +INFO:tensorflow:step = 1701, loss = 2.39917, precision = 0.773438 (22.174 sec) +INFO:tensorflow:global_step/sec: 4.50344 +INFO:tensorflow:step = 1801, loss = 2.32019, precision = 0.78125 (22.205 sec) +INFO:tensorflow:global_step/sec: 4.50779 +INFO:tensorflow:step = 1901, loss = 2.06069, precision = 0.78125 (22.185 sec) +Saved checkpoint after 5 epoch(s) to data/resnet164_b/checkpoints/00005... +INFO:tensorflow:global_step/sec: 4.0705 +INFO:tensorflow:step = 2001, loss = 2.20496, precision = 0.75 (24.566 sec) +INFO:tensorflow:global_step/sec: 4.51004 +INFO:tensorflow:step = 2101, loss = 1.85154, precision = 0.765625 (22.173 sec) +INFO:tensorflow:global_step/sec: 4.50544 +INFO:tensorflow:step = 2201, loss = 1.77052, precision = 0.789062 (22.195 sec) +INFO:tensorflow:global_step/sec: 4.50182 +INFO:tensorflow:step = 2301, loss = 1.75173, precision = 0.742188 (22.213 sec) +Saved checkpoint after 6 epoch(s) to data/resnet164_b/checkpoints/00006... +INFO:tensorflow:global_step/sec: 4.05766 +INFO:tensorflow:step = 2401, loss = 1.62026, precision = 0.757812 (24.645 sec) +INFO:tensorflow:global_step/sec: 4.50986 +INFO:tensorflow:step = 2501, loss = 1.61222, precision = 0.734375 (22.174 sec) +INFO:tensorflow:global_step/sec: 4.50901 +INFO:tensorflow:step = 2601, loss = 1.44079, precision = 0.835938 (22.178 sec) +INFO:tensorflow:global_step/sec: 4.50078 +INFO:tensorflow:step = 2701, loss = 1.46287, precision = 0.789062 (22.218 sec) +Saved checkpoint after 7 epoch(s) to data/resnet164_b/checkpoints/00007... +INFO:tensorflow:global_step/sec: 4.09498 +INFO:tensorflow:step = 2801, loss = 1.42073, precision = 0.765625 (24.420 sec) +INFO:tensorflow:global_step/sec: 4.51317 +INFO:tensorflow:step = 2901, loss = 1.30534, precision = 0.796875 (22.157 sec) +INFO:tensorflow:global_step/sec: 4.50843 +INFO:tensorflow:step = 3001, loss = 1.24513, precision = 0.804688 (22.181 sec) +INFO:tensorflow:global_step/sec: 4.50887 +INFO:tensorflow:step = 3101, loss = 1.20949, precision = 0.851562 (22.178 sec) +Saved checkpoint after 8 epoch(s) to data/resnet164_b/checkpoints/00008... +INFO:tensorflow:global_step/sec: 4.06085 +INFO:tensorflow:step = 3201, loss = 1.17141, precision = 0.789062 (24.626 sec) +INFO:tensorflow:global_step/sec: 4.50715 +INFO:tensorflow:step = 3301, loss = 1.03851, precision = 0.859375 (22.187 sec) +INFO:tensorflow:global_step/sec: 4.50217 +INFO:tensorflow:step = 3401, loss = 1.13459, precision = 0.8125 (22.211 sec) +INFO:tensorflow:global_step/sec: 4.51073 +INFO:tensorflow:step = 3501, loss = 0.925007, precision = 0.867188 (22.171 sec) +Saved checkpoint after 9 epoch(s) to data/resnet164_b/checkpoints/00009... +INFO:tensorflow:global_step/sec: 4.05726 +INFO:tensorflow:step = 3601, loss = 1.0374, precision = 0.835938 (24.646 sec) +INFO:tensorflow:global_step/sec: 4.49725 +INFO:tensorflow:step = 3701, loss = 1.0823, precision = 0.796875 (22.237 sec) +INFO:tensorflow:global_step/sec: 4.50137 +INFO:tensorflow:step = 3801, loss = 0.857402, precision = 0.867188 (22.215 sec) +INFO:tensorflow:global_step/sec: 4.50626 +INFO:tensorflow:step = 3901, loss = 0.936576, precision = 0.84375 (22.190 sec) +Saved checkpoint after 10 epoch(s) to data/resnet164_b/checkpoints/00010... +INFO:tensorflow:global_step/sec: 4.01011 +INFO:tensorflow:step = 4001, loss = 1.06812, precision = 0.75 (24.938 sec) +INFO:tensorflow:global_step/sec: 4.50395 +INFO:tensorflow:step = 4101, loss = 0.880035, precision = 0.875 (22.202 sec) +INFO:tensorflow:global_step/sec: 4.50674 +INFO:tensorflow:step = 4201, loss = 0.895845, precision = 0.8125 (22.188 sec) +Saved checkpoint after 11 epoch(s) to data/resnet164_b/checkpoints/00011... +INFO:tensorflow:global_step/sec: 4.06934 +INFO:tensorflow:step = 4301, loss = 0.920511, precision = 0.796875 (24.573 sec) +INFO:tensorflow:global_step/sec: 4.50676 +INFO:tensorflow:step = 4401, loss = 0.969711, precision = 0.820312 (22.190 sec) +INFO:tensorflow:global_step/sec: 4.50661 +INFO:tensorflow:step = 4501, loss = 0.919952, precision = 0.8125 (22.190 sec) +INFO:tensorflow:global_step/sec: 4.49915 +INFO:tensorflow:step = 4601, loss = 0.885688, precision = 0.828125 (22.226 sec) +Saved checkpoint after 12 epoch(s) to data/resnet164_b/checkpoints/00012... +INFO:tensorflow:global_step/sec: 4.07409 +INFO:tensorflow:step = 4701, loss = 0.850987, precision = 0.851562 (24.546 sec) +INFO:tensorflow:global_step/sec: 4.49702 +INFO:tensorflow:step = 4801, loss = 0.923633, precision = 0.78125 (22.237 sec) +INFO:tensorflow:global_step/sec: 4.50615 +INFO:tensorflow:step = 4901, loss = 0.838534, precision = 0.867188 (22.192 sec) +INFO:tensorflow:global_step/sec: 4.50468 +INFO:tensorflow:step = 5001, loss = 0.769052, precision = 0.875 (22.199 sec) +Saved checkpoint after 13 epoch(s) to data/resnet164_b/checkpoints/00013... +INFO:tensorflow:global_step/sec: 4.09018 +INFO:tensorflow:step = 5101, loss = 0.796331, precision = 0.851562 (24.449 sec) +INFO:tensorflow:global_step/sec: 4.50278 +INFO:tensorflow:step = 5201, loss = 0.68574, precision = 0.898438 (22.208 sec) +INFO:tensorflow:global_step/sec: 4.50912 +INFO:tensorflow:step = 5301, loss = 0.721821, precision = 0.90625 (22.177 sec) +INFO:tensorflow:global_step/sec: 4.51307 +INFO:tensorflow:step = 5401, loss = 0.6806, precision = 0.875 (22.158 sec) +Saved checkpoint after 14 epoch(s) to data/resnet164_b/checkpoints/00014... +INFO:tensorflow:global_step/sec: 4.04046 +INFO:tensorflow:step = 5501, loss = 0.728001, precision = 0.851562 (24.750 sec) +INFO:tensorflow:global_step/sec: 4.50462 +INFO:tensorflow:step = 5601, loss = 0.819345, precision = 0.835938 (22.199 sec) +INFO:tensorflow:global_step/sec: 4.51043 +INFO:tensorflow:step = 5701, loss = 0.747886, precision = 0.859375 (22.171 sec) +INFO:tensorflow:global_step/sec: 4.50537 +INFO:tensorflow:step = 5801, loss = 0.718496, precision = 0.882812 (22.196 sec) +Saved checkpoint after 15 epoch(s) to data/resnet164_b/checkpoints/00015... +INFO:tensorflow:global_step/sec: 4.0715 +INFO:tensorflow:step = 5901, loss = 0.911161, precision = 0.789062 (24.561 sec) +INFO:tensorflow:global_step/sec: 4.50559 +INFO:tensorflow:step = 6001, loss = 0.799322, precision = 0.8125 (22.194 sec) +INFO:tensorflow:global_step/sec: 4.50429 +INFO:tensorflow:step = 6101, loss = 0.821712, precision = 0.820312 (22.201 sec) +INFO:tensorflow:global_step/sec: 4.50584 +INFO:tensorflow:step = 6201, loss = 0.829305, precision = 0.859375 (22.193 sec) +Saved checkpoint after 16 epoch(s) to data/resnet164_b/checkpoints/00016... +INFO:tensorflow:global_step/sec: 4.05562 +INFO:tensorflow:step = 6301, loss = 0.893874, precision = 0.75 (24.657 sec) +INFO:tensorflow:global_step/sec: 4.50213 +INFO:tensorflow:step = 6401, loss = 0.774855, precision = 0.835938 (22.212 sec) +INFO:tensorflow:global_step/sec: 4.50159 +INFO:tensorflow:step = 6501, loss = 0.784524, precision = 0.804688 (22.214 sec) +INFO:tensorflow:global_step/sec: 4.50106 +INFO:tensorflow:step = 6601, loss = 0.77931, precision = 0.867188 (22.217 sec) +Saved checkpoint after 17 epoch(s) to data/resnet164_b/checkpoints/00017... +INFO:tensorflow:global_step/sec: 4.05769 +INFO:tensorflow:step = 6701, loss = 0.788777, precision = 0.867188 (24.644 sec) +INFO:tensorflow:global_step/sec: 4.50997 +INFO:tensorflow:step = 6801, loss = 0.669884, precision = 0.851562 (22.173 sec) +INFO:tensorflow:global_step/sec: 4.49815 +INFO:tensorflow:step = 6901, loss = 0.906916, precision = 0.8125 (22.231 sec) +INFO:tensorflow:global_step/sec: 4.50533 +INFO:tensorflow:step = 7001, loss = 0.803581, precision = 0.851562 (22.196 sec) +Saved checkpoint after 18 epoch(s) to data/resnet164_b/checkpoints/00018... +INFO:tensorflow:global_step/sec: 4.05415 +INFO:tensorflow:step = 7101, loss = 0.646206, precision = 0.890625 (24.666 sec) +INFO:tensorflow:global_step/sec: 4.49873 +INFO:tensorflow:step = 7201, loss = 0.743718, precision = 0.851562 (22.228 sec) +INFO:tensorflow:global_step/sec: 4.50543 +INFO:tensorflow:step = 7301, loss = 0.698545, precision = 0.875 (22.195 sec) +INFO:tensorflow:global_step/sec: 4.50737 +INFO:tensorflow:step = 7401, loss = 0.627842, precision = 0.898438 (22.186 sec) +Saved checkpoint after 19 epoch(s) to data/resnet164_b/checkpoints/00019... +INFO:tensorflow:global_step/sec: 4.06967 +INFO:tensorflow:step = 7501, loss = 0.84832, precision = 0.820312 (24.572 sec) +INFO:tensorflow:global_step/sec: 4.50616 +INFO:tensorflow:step = 7601, loss = 0.742103, precision = 0.859375 (22.192 sec) +INFO:tensorflow:global_step/sec: 4.50765 +INFO:tensorflow:step = 7701, loss = 0.878042, precision = 0.765625 (22.184 sec) +INFO:tensorflow:global_step/sec: 4.5062 +INFO:tensorflow:step = 7801, loss = 0.757645, precision = 0.851562 (22.192 sec) +Saved checkpoint after 20 epoch(s) to data/resnet164_b/checkpoints/00020... +INFO:tensorflow:global_step/sec: 4.00758 +INFO:tensorflow:step = 7901, loss = 0.595212, precision = 0.898438 (24.953 sec) +INFO:tensorflow:global_step/sec: 4.50354 +INFO:tensorflow:step = 8001, loss = 0.812804, precision = 0.859375 (22.204 sec) +INFO:tensorflow:global_step/sec: 4.49639 +INFO:tensorflow:step = 8101, loss = 0.674644, precision = 0.921875 (22.240 sec) +INFO:tensorflow:global_step/sec: 4.50711 +INFO:tensorflow:step = 8201, loss = 0.787022, precision = 0.875 (22.187 sec) +Saved checkpoint after 21 epoch(s) to data/resnet164_b/checkpoints/00021... +INFO:tensorflow:global_step/sec: 4.0756 +INFO:tensorflow:step = 8301, loss = 0.566437, precision = 0.90625 (24.536 sec) +INFO:tensorflow:global_step/sec: 4.50253 +INFO:tensorflow:step = 8401, loss = 0.717454, precision = 0.851562 (22.210 sec) +INFO:tensorflow:global_step/sec: 4.50831 +INFO:tensorflow:step = 8501, loss = 0.901247, precision = 0.820312 (22.181 sec) +INFO:tensorflow:global_step/sec: 4.49882 +INFO:tensorflow:step = 8601, loss = 0.796378, precision = 0.835938 (22.228 sec) +Saved checkpoint after 22 epoch(s) to data/resnet164_b/checkpoints/00022... +INFO:tensorflow:global_step/sec: 4.02534 +INFO:tensorflow:step = 8701, loss = 0.647816, precision = 0.875 (24.843 sec) +INFO:tensorflow:global_step/sec: 4.50484 +INFO:tensorflow:step = 8801, loss = 0.705254, precision = 0.867188 (22.198 sec) +INFO:tensorflow:global_step/sec: 4.50566 +INFO:tensorflow:step = 8901, loss = 0.73573, precision = 0.867188 (22.195 sec) +Saved checkpoint after 23 epoch(s) to data/resnet164_b/checkpoints/00023... +INFO:tensorflow:global_step/sec: 4.07345 +INFO:tensorflow:step = 9001, loss = 0.645966, precision = 0.867188 (24.549 sec) +INFO:tensorflow:global_step/sec: 4.5114 +INFO:tensorflow:step = 9101, loss = 0.774796, precision = 0.851562 (22.167 sec) +INFO:tensorflow:global_step/sec: 4.50592 +INFO:tensorflow:step = 9201, loss = 0.803674, precision = 0.8125 (22.192 sec) +INFO:tensorflow:global_step/sec: 4.50822 +INFO:tensorflow:step = 9301, loss = 0.759733, precision = 0.828125 (22.182 sec) +Saved checkpoint after 24 epoch(s) to data/resnet164_b/checkpoints/00024... +INFO:tensorflow:global_step/sec: 4.05446 +INFO:tensorflow:step = 9401, loss = 0.751648, precision = 0.851562 (24.665 sec) +INFO:tensorflow:global_step/sec: 4.50549 +INFO:tensorflow:step = 9501, loss = 0.744819, precision = 0.859375 (22.194 sec) +INFO:tensorflow:global_step/sec: 4.50613 +INFO:tensorflow:step = 9601, loss = 0.678216, precision = 0.867188 (22.191 sec) +INFO:tensorflow:global_step/sec: 4.5018 +INFO:tensorflow:step = 9701, loss = 0.78034, precision = 0.859375 (22.213 sec) +Saved checkpoint after 25 epoch(s) to data/resnet164_b/checkpoints/00025... +INFO:tensorflow:global_step/sec: 4.06758 +INFO:tensorflow:step = 9801, loss = 0.658056, precision = 0.898438 (24.585 sec) +INFO:tensorflow:global_step/sec: 4.50891 +INFO:tensorflow:step = 9901, loss = 0.768032, precision = 0.820312 (22.178 sec) +INFO:tensorflow:global_step/sec: 4.50847 +INFO:tensorflow:step = 10001, loss = 0.742062, precision = 0.828125 (22.182 sec) +INFO:tensorflow:global_step/sec: 4.5046 +INFO:tensorflow:step = 10101, loss = 0.860589, precision = 0.804688 (22.198 sec) +Saved checkpoint after 26 epoch(s) to data/resnet164_b/checkpoints/00026... +INFO:tensorflow:global_step/sec: 4.08795 +INFO:tensorflow:step = 10201, loss = 0.653669, precision = 0.859375 (24.463 sec) +INFO:tensorflow:global_step/sec: 4.5083 +INFO:tensorflow:step = 10301, loss = 0.644489, precision = 0.890625 (22.181 sec) +INFO:tensorflow:global_step/sec: 4.49902 +INFO:tensorflow:step = 10401, loss = 0.738468, precision = 0.851562 (22.227 sec) +INFO:tensorflow:global_step/sec: 4.49655 +INFO:tensorflow:step = 10501, loss = 0.748106, precision = 0.867188 (22.241 sec) +Saved checkpoint after 27 epoch(s) to data/resnet164_b/checkpoints/00027... +INFO:tensorflow:global_step/sec: 4.0294 +INFO:tensorflow:step = 10601, loss = 0.655537, precision = 0.890625 (24.816 sec) +INFO:tensorflow:global_step/sec: 4.50105 +INFO:tensorflow:step = 10701, loss = 0.670334, precision = 0.90625 (22.217 sec) +INFO:tensorflow:global_step/sec: 4.5081 +INFO:tensorflow:step = 10801, loss = 0.683327, precision = 0.914062 (22.182 sec) +INFO:tensorflow:global_step/sec: 4.49973 +INFO:tensorflow:step = 10901, loss = 0.784045, precision = 0.84375 (22.225 sec) +Saved checkpoint after 28 epoch(s) to data/resnet164_b/checkpoints/00028... +INFO:tensorflow:global_step/sec: 4.06812 +INFO:tensorflow:step = 11001, loss = 0.703208, precision = 0.882812 (24.580 sec) +INFO:tensorflow:global_step/sec: 4.50445 +INFO:tensorflow:step = 11101, loss = 0.662254, precision = 0.859375 (22.200 sec) +INFO:tensorflow:global_step/sec: 4.50359 +INFO:tensorflow:step = 11201, loss = 0.726006, precision = 0.859375 (22.205 sec) +INFO:tensorflow:global_step/sec: 4.51304 +INFO:tensorflow:step = 11301, loss = 0.736108, precision = 0.835938 (22.158 sec) +Saved checkpoint after 29 epoch(s) to data/resnet164_b/checkpoints/00029... +INFO:tensorflow:global_step/sec: 4.05654 +INFO:tensorflow:step = 11401, loss = 0.712545, precision = 0.875 (24.652 sec) +INFO:tensorflow:global_step/sec: 4.50555 +INFO:tensorflow:step = 11501, loss = 0.812425, precision = 0.859375 (22.195 sec) +INFO:tensorflow:global_step/sec: 4.50745 +INFO:tensorflow:step = 11601, loss = 0.701415, precision = 0.890625 (22.187 sec) +INFO:tensorflow:global_step/sec: 4.50843 +INFO:tensorflow:step = 11701, loss = 0.654872, precision = 0.898438 (22.179 sec) +Saved checkpoint after 30 epoch(s) to data/resnet164_b/checkpoints/00030... +INFO:tensorflow:global_step/sec: 4.07359 +INFO:tensorflow:step = 11801, loss = 0.721868, precision = 0.859375 (24.549 sec) +INFO:tensorflow:global_step/sec: 4.50126 +INFO:tensorflow:step = 11901, loss = 0.660141, precision = 0.882812 (22.216 sec) +INFO:tensorflow:global_step/sec: 4.49724 +INFO:tensorflow:step = 12001, loss = 0.621428, precision = 0.882812 (22.236 sec) +INFO:tensorflow:global_step/sec: 4.50579 +INFO:tensorflow:step = 12101, loss = 0.770217, precision = 0.84375 (22.195 sec) +Saved checkpoint after 31 epoch(s) to data/resnet164_b/checkpoints/00031... +INFO:tensorflow:global_step/sec: 4.06888 +INFO:tensorflow:step = 12201, loss = 0.73752, precision = 0.851562 (24.577 sec) +INFO:tensorflow:global_step/sec: 4.49885 +INFO:tensorflow:step = 12301, loss = 0.649379, precision = 0.882812 (22.227 sec) +INFO:tensorflow:global_step/sec: 4.50122 +INFO:tensorflow:step = 12401, loss = 0.739503, precision = 0.835938 (22.218 sec) +INFO:tensorflow:global_step/sec: 4.50127 +INFO:tensorflow:step = 12501, loss = 0.626099, precision = 0.890625 (22.215 sec) +Saved checkpoint after 32 epoch(s) to data/resnet164_b/checkpoints/00032... +INFO:tensorflow:global_step/sec: 4.06379 +INFO:tensorflow:step = 12601, loss = 0.856466, precision = 0.84375 (24.607 sec) +INFO:tensorflow:global_step/sec: 4.50464 +INFO:tensorflow:step = 12701, loss = 0.658639, precision = 0.890625 (22.199 sec) +INFO:tensorflow:global_step/sec: 4.50916 +INFO:tensorflow:step = 12801, loss = 0.681697, precision = 0.851562 (22.177 sec) +INFO:tensorflow:global_step/sec: 4.5087 +INFO:tensorflow:step = 12901, loss = 0.690756, precision = 0.875 (22.179 sec) +Saved checkpoint after 33 epoch(s) to data/resnet164_b/checkpoints/00033... +INFO:tensorflow:global_step/sec: 4.07644 +INFO:tensorflow:step = 13001, loss = 0.703123, precision = 0.898438 (24.531 sec) +INFO:tensorflow:global_step/sec: 4.51005 +INFO:tensorflow:step = 13101, loss = 0.794392, precision = 0.851562 (22.173 sec) +INFO:tensorflow:global_step/sec: 4.51233 +INFO:tensorflow:step = 13201, loss = 0.728161, precision = 0.867188 (22.161 sec) +Saved checkpoint after 34 epoch(s) to data/resnet164_b/checkpoints/00034... +INFO:tensorflow:global_step/sec: 4.09065 +INFO:tensorflow:step = 13301, loss = 0.761782, precision = 0.890625 (24.446 sec) +INFO:tensorflow:global_step/sec: 4.50332 +INFO:tensorflow:step = 13401, loss = 0.813434, precision = 0.796875 (22.206 sec) +INFO:tensorflow:global_step/sec: 4.51369 +INFO:tensorflow:step = 13501, loss = 0.795882, precision = 0.835938 (22.155 sec) +INFO:tensorflow:global_step/sec: 4.50711 +INFO:tensorflow:step = 13601, loss = 0.759701, precision = 0.867188 (22.187 sec) +Saved checkpoint after 35 epoch(s) to data/resnet164_b/checkpoints/00035... +INFO:tensorflow:global_step/sec: 4.06402 +INFO:tensorflow:step = 13701, loss = 0.665975, precision = 0.859375 (24.606 sec) +INFO:tensorflow:global_step/sec: 4.49855 +INFO:tensorflow:step = 13801, loss = 0.621783, precision = 0.914062 (22.229 sec) +INFO:tensorflow:global_step/sec: 4.49988 +INFO:tensorflow:step = 13901, loss = 0.637466, precision = 0.859375 (22.223 sec) +INFO:tensorflow:global_step/sec: 4.50881 +INFO:tensorflow:step = 14001, loss = 0.798258, precision = 0.835938 (22.179 sec) +Saved checkpoint after 36 epoch(s) to data/resnet164_b/checkpoints/00036... +INFO:tensorflow:global_step/sec: 4.10168 +INFO:tensorflow:step = 14101, loss = 0.649534, precision = 0.875 (24.380 sec) +INFO:tensorflow:global_step/sec: 4.50761 +INFO:tensorflow:step = 14201, loss = 0.70353, precision = 0.882812 (22.185 sec) +INFO:tensorflow:global_step/sec: 4.50669 +INFO:tensorflow:step = 14301, loss = 0.719864, precision = 0.867188 (22.189 sec) +INFO:tensorflow:global_step/sec: 4.50312 +INFO:tensorflow:step = 14401, loss = 0.763483, precision = 0.835938 (22.207 sec) +Saved checkpoint after 37 epoch(s) to data/resnet164_b/checkpoints/00037... +INFO:tensorflow:global_step/sec: 4.04172 +INFO:tensorflow:step = 14501, loss = 0.672676, precision = 0.875 (24.742 sec) +INFO:tensorflow:global_step/sec: 4.51025 +INFO:tensorflow:step = 14601, loss = 0.696437, precision = 0.882812 (22.172 sec) +INFO:tensorflow:global_step/sec: 4.50582 +INFO:tensorflow:step = 14701, loss = 0.670782, precision = 0.90625 (22.194 sec) +INFO:tensorflow:global_step/sec: 4.50937 +INFO:tensorflow:step = 14801, loss = 0.665878, precision = 0.875 (22.176 sec) +Saved checkpoint after 38 epoch(s) to data/resnet164_b/checkpoints/00038... +INFO:tensorflow:global_step/sec: 4.09491 +INFO:tensorflow:step = 14901, loss = 0.586472, precision = 0.9375 (24.421 sec) +INFO:tensorflow:global_step/sec: 4.50903 +INFO:tensorflow:step = 15001, loss = 0.745632, precision = 0.84375 (22.177 sec) +INFO:tensorflow:global_step/sec: 4.49852 +INFO:tensorflow:step = 15101, loss = 0.569148, precision = 0.9375 (22.230 sec) +INFO:tensorflow:global_step/sec: 4.51249 +INFO:tensorflow:step = 15201, loss = 0.950446, precision = 0.804688 (22.161 sec) +Saved checkpoint after 39 epoch(s) to data/resnet164_b/checkpoints/00039... +INFO:tensorflow:global_step/sec: 4.0849 +INFO:tensorflow:step = 15301, loss = 0.64538, precision = 0.890625 (24.480 sec) +INFO:tensorflow:global_step/sec: 4.51101 +INFO:tensorflow:step = 15401, loss = 0.704231, precision = 0.882812 (22.168 sec) +INFO:tensorflow:global_step/sec: 4.50823 +INFO:tensorflow:step = 15501, loss = 0.745698, precision = 0.859375 (22.182 sec) +INFO:tensorflow:global_step/sec: 4.50547 +INFO:tensorflow:step = 15601, loss = 0.592679, precision = 0.898438 (22.195 sec) +Saved checkpoint after 40 epoch(s) to data/resnet164_b/checkpoints/00040... +INFO:tensorflow:global_step/sec: 4.08708 +INFO:tensorflow:step = 15701, loss = 0.751344, precision = 0.859375 (24.468 sec) +INFO:tensorflow:global_step/sec: 4.50671 +INFO:tensorflow:step = 15801, loss = 0.665273, precision = 0.867188 (22.189 sec) +INFO:tensorflow:global_step/sec: 4.49677 +INFO:tensorflow:step = 15901, loss = 0.72277, precision = 0.835938 (22.238 sec) +INFO:tensorflow:global_step/sec: 4.49831 +INFO:tensorflow:step = 16001, loss = 0.62272, precision = 0.90625 (22.231 sec) +Saved checkpoint after 41 epoch(s) to data/resnet164_b/checkpoints/00041... +INFO:tensorflow:global_step/sec: 4.02265 +INFO:tensorflow:step = 16101, loss = 0.629194, precision = 0.90625 (24.859 sec) +INFO:tensorflow:global_step/sec: 4.50745 +INFO:tensorflow:step = 16201, loss = 0.578666, precision = 0.929688 (22.185 sec) +INFO:tensorflow:global_step/sec: 4.51048 +INFO:tensorflow:step = 16301, loss = 0.656584, precision = 0.90625 (22.170 sec) +INFO:tensorflow:global_step/sec: 4.50184 +INFO:tensorflow:step = 16401, loss = 0.83328, precision = 0.84375 (22.213 sec) +Saved checkpoint after 42 epoch(s) to data/resnet164_b/checkpoints/00042... +INFO:tensorflow:global_step/sec: 4.01693 +INFO:tensorflow:step = 16501, loss = 0.792718, precision = 0.820312 (24.895 sec) +INFO:tensorflow:global_step/sec: 4.50522 +INFO:tensorflow:step = 16601, loss = 0.652825, precision = 0.890625 (22.196 sec) +INFO:tensorflow:global_step/sec: 4.50582 +INFO:tensorflow:step = 16701, loss = 0.664233, precision = 0.890625 (22.194 sec) +INFO:tensorflow:global_step/sec: 4.50598 +INFO:tensorflow:step = 16801, loss = 0.651725, precision = 0.882812 (22.192 sec) +Saved checkpoint after 43 epoch(s) to data/resnet164_b/checkpoints/00043... +INFO:tensorflow:global_step/sec: 4.06321 +INFO:tensorflow:step = 16901, loss = 0.706645, precision = 0.867188 (24.611 sec) +INFO:tensorflow:global_step/sec: 4.50604 +INFO:tensorflow:step = 17001, loss = 0.696293, precision = 0.867188 (22.192 sec) +INFO:tensorflow:global_step/sec: 4.51035 +INFO:tensorflow:step = 17101, loss = 0.60913, precision = 0.921875 (22.171 sec) +INFO:tensorflow:global_step/sec: 4.50621 +INFO:tensorflow:step = 17201, loss = 0.737007, precision = 0.867188 (22.192 sec) +Saved checkpoint after 44 epoch(s) to data/resnet164_b/checkpoints/00044... +INFO:tensorflow:global_step/sec: 4.06591 +INFO:tensorflow:step = 17301, loss = 0.858086, precision = 0.820312 (24.595 sec) +INFO:tensorflow:global_step/sec: 4.51086 +INFO:tensorflow:step = 17401, loss = 0.66333, precision = 0.882812 (22.169 sec) +INFO:tensorflow:global_step/sec: 4.49768 +INFO:tensorflow:step = 17501, loss = 0.740215, precision = 0.835938 (22.234 sec) +Saved checkpoint after 45 epoch(s) to data/resnet164_b/checkpoints/00045... +INFO:tensorflow:global_step/sec: 4.08281 +INFO:tensorflow:step = 17601, loss = 0.666182, precision = 0.875 (24.493 sec) +INFO:tensorflow:global_step/sec: 4.49464 +INFO:tensorflow:step = 17701, loss = 0.737983, precision = 0.882812 (22.249 sec) +INFO:tensorflow:global_step/sec: 4.49847 +INFO:tensorflow:step = 17801, loss = 0.631264, precision = 0.890625 (22.230 sec) +INFO:tensorflow:global_step/sec: 4.49907 +INFO:tensorflow:step = 17901, loss = 0.642643, precision = 0.914062 (22.229 sec) +Saved checkpoint after 46 epoch(s) to data/resnet164_b/checkpoints/00046... +INFO:tensorflow:global_step/sec: 4.01 +INFO:tensorflow:step = 18001, loss = 0.8227, precision = 0.835938 (24.936 sec) +INFO:tensorflow:global_step/sec: 4.51026 +INFO:tensorflow:step = 18101, loss = 0.779384, precision = 0.835938 (22.172 sec) +INFO:tensorflow:global_step/sec: 4.5054 +INFO:tensorflow:step = 18201, loss = 0.622156, precision = 0.921875 (22.195 sec) +INFO:tensorflow:global_step/sec: 4.49903 +INFO:tensorflow:step = 18301, loss = 0.767969, precision = 0.867188 (22.228 sec) +Saved checkpoint after 47 epoch(s) to data/resnet164_b/checkpoints/00047... +INFO:tensorflow:global_step/sec: 4.06104 +INFO:tensorflow:step = 18401, loss = 0.643326, precision = 0.90625 (24.624 sec) +INFO:tensorflow:global_step/sec: 4.51395 +INFO:tensorflow:step = 18501, loss = 0.70011, precision = 0.875 (22.153 sec) +INFO:tensorflow:global_step/sec: 4.49872 +INFO:tensorflow:step = 18601, loss = 0.54808, precision = 0.921875 (22.230 sec) +INFO:tensorflow:global_step/sec: 4.49873 +INFO:tensorflow:step = 18701, loss = 0.811486, precision = 0.84375 (22.228 sec) +Saved checkpoint after 48 epoch(s) to data/resnet164_b/checkpoints/00048... +INFO:tensorflow:global_step/sec: 4.00106 +INFO:tensorflow:step = 18801, loss = 0.691799, precision = 0.867188 (24.993 sec) +INFO:tensorflow:global_step/sec: 4.50862 +INFO:tensorflow:step = 18901, loss = 0.678525, precision = 0.875 (22.180 sec) +INFO:tensorflow:global_step/sec: 4.50883 +INFO:tensorflow:step = 19001, loss = 0.830017, precision = 0.8125 (22.179 sec) +INFO:tensorflow:global_step/sec: 4.50101 +INFO:tensorflow:step = 19101, loss = 0.693897, precision = 0.890625 (22.217 sec) +Saved checkpoint after 49 epoch(s) to data/resnet164_b/checkpoints/00049... +INFO:tensorflow:global_step/sec: 4.06646 +INFO:tensorflow:step = 19201, loss = 0.789796, precision = 0.820312 (24.591 sec) +INFO:tensorflow:global_step/sec: 4.50129 +INFO:tensorflow:step = 19301, loss = 0.728226, precision = 0.84375 (22.216 sec) +INFO:tensorflow:global_step/sec: 4.50547 +INFO:tensorflow:step = 19401, loss = 0.864713, precision = 0.851562 (22.195 sec) +INFO:tensorflow:global_step/sec: 4.49952 +INFO:tensorflow:step = 19501, loss = 0.704312, precision = 0.859375 (22.225 sec) +Saved checkpoint after 50 epoch(s) to data/resnet164_b/checkpoints/00050... +INFO:tensorflow:global_step/sec: 4.07581 +INFO:tensorflow:step = 19601, loss = 0.711815, precision = 0.867188 (24.535 sec) +INFO:tensorflow:global_step/sec: 4.50627 +INFO:tensorflow:step = 19701, loss = 0.664242, precision = 0.875 (22.191 sec) +INFO:tensorflow:global_step/sec: 4.50444 +INFO:tensorflow:step = 19801, loss = 0.685758, precision = 0.867188 (22.200 sec) +INFO:tensorflow:global_step/sec: 4.50536 +INFO:tensorflow:step = 19901, loss = 0.610662, precision = 0.90625 (22.196 sec) +Saved checkpoint after 51 epoch(s) to data/resnet164_b/checkpoints/00051... +INFO:tensorflow:global_step/sec: 4.0867 +INFO:tensorflow:step = 20001, loss = 0.71992, precision = 0.84375 (24.470 sec) +INFO:tensorflow:global_step/sec: 4.50824 +INFO:tensorflow:step = 20101, loss = 0.831592, precision = 0.851562 (22.183 sec) +INFO:tensorflow:global_step/sec: 4.51442 +INFO:tensorflow:step = 20201, loss = 0.597986, precision = 0.90625 (22.150 sec) +INFO:tensorflow:global_step/sec: 4.50826 +INFO:tensorflow:step = 20301, loss = 0.656492, precision = 0.875 (22.182 sec) +Saved checkpoint after 52 epoch(s) to data/resnet164_b/checkpoints/00052... +INFO:tensorflow:global_step/sec: 4.07086 +INFO:tensorflow:step = 20401, loss = 0.59692, precision = 0.914062 (24.565 sec) +INFO:tensorflow:global_step/sec: 4.51852 +INFO:tensorflow:step = 20501, loss = 0.727734, precision = 0.875 (22.131 sec) +INFO:tensorflow:global_step/sec: 4.50644 +INFO:tensorflow:step = 20601, loss = 0.644677, precision = 0.90625 (22.190 sec) +INFO:tensorflow:global_step/sec: 4.51209 +INFO:tensorflow:step = 20701, loss = 0.625977, precision = 0.898438 (22.163 sec) +Saved checkpoint after 53 epoch(s) to data/resnet164_b/checkpoints/00053... +INFO:tensorflow:global_step/sec: 4.06627 +INFO:tensorflow:step = 20801, loss = 0.64743, precision = 0.882812 (24.593 sec) +INFO:tensorflow:global_step/sec: 4.51167 +INFO:tensorflow:step = 20901, loss = 0.742494, precision = 0.867188 (22.165 sec) +INFO:tensorflow:global_step/sec: 4.50691 +INFO:tensorflow:step = 21001, loss = 0.759808, precision = 0.859375 (22.187 sec) +INFO:tensorflow:global_step/sec: 4.50191 +INFO:tensorflow:step = 21101, loss = 0.649122, precision = 0.921875 (22.213 sec) +Saved checkpoint after 54 epoch(s) to data/resnet164_b/checkpoints/00054... +INFO:tensorflow:global_step/sec: 4.0908 +INFO:tensorflow:step = 21201, loss = 0.68533, precision = 0.859375 (24.444 sec) +INFO:tensorflow:global_step/sec: 4.52627 +INFO:tensorflow:step = 21301, loss = 0.591892, precision = 0.914062 (22.093 sec) +INFO:tensorflow:global_step/sec: 4.52037 +INFO:tensorflow:step = 21401, loss = 0.72843, precision = 0.828125 (22.122 sec) +INFO:tensorflow:global_step/sec: 4.51042 +INFO:tensorflow:step = 21501, loss = 0.616918, precision = 0.898438 (22.171 sec) +Saved checkpoint after 55 epoch(s) to data/resnet164_b/checkpoints/00055... +INFO:tensorflow:global_step/sec: 4.09579 +INFO:tensorflow:step = 21601, loss = 0.733349, precision = 0.859375 (24.415 sec) +INFO:tensorflow:global_step/sec: 4.51659 +INFO:tensorflow:step = 21701, loss = 0.639513, precision = 0.914062 (22.141 sec) +INFO:tensorflow:global_step/sec: 4.51578 +INFO:tensorflow:step = 21801, loss = 0.597171, precision = 0.9375 (22.144 sec) +Saved checkpoint after 56 epoch(s) to data/resnet164_b/checkpoints/00056... +INFO:tensorflow:global_step/sec: 4.08825 +INFO:tensorflow:step = 21901, loss = 0.746544, precision = 0.882812 (24.460 sec) +INFO:tensorflow:global_step/sec: 4.51219 +INFO:tensorflow:step = 22001, loss = 0.629775, precision = 0.90625 (22.163 sec) +INFO:tensorflow:global_step/sec: 4.51482 +INFO:tensorflow:step = 22101, loss = 0.787723, precision = 0.820312 (22.149 sec) +INFO:tensorflow:global_step/sec: 4.51388 +INFO:tensorflow:step = 22201, loss = 0.582, precision = 0.90625 (22.154 sec) +Saved checkpoint after 57 epoch(s) to data/resnet164_b/checkpoints/00057... +INFO:tensorflow:global_step/sec: 4.04415 +INFO:tensorflow:step = 22301, loss = 0.71502, precision = 0.890625 (24.727 sec) +INFO:tensorflow:global_step/sec: 4.509 +INFO:tensorflow:step = 22401, loss = 0.747232, precision = 0.875 (22.178 sec) +INFO:tensorflow:global_step/sec: 4.52172 +INFO:tensorflow:step = 22501, loss = 0.666782, precision = 0.898438 (22.115 sec) +INFO:tensorflow:global_step/sec: 4.5126 +INFO:tensorflow:step = 22601, loss = 0.81657, precision = 0.820312 (22.160 sec) +Saved checkpoint after 58 epoch(s) to data/resnet164_b/checkpoints/00058... +INFO:tensorflow:global_step/sec: 4.04855 +INFO:tensorflow:step = 22701, loss = 0.553689, precision = 0.929688 (24.700 sec) +INFO:tensorflow:global_step/sec: 4.51248 +INFO:tensorflow:step = 22801, loss = 0.6815, precision = 0.851562 (22.161 sec) +INFO:tensorflow:global_step/sec: 4.50652 +INFO:tensorflow:step = 22901, loss = 0.708033, precision = 0.867188 (22.190 sec) +INFO:tensorflow:global_step/sec: 4.51131 +INFO:tensorflow:step = 23001, loss = 0.651123, precision = 0.90625 (22.166 sec) +Saved checkpoint after 59 epoch(s) to data/resnet164_b/checkpoints/00059... +INFO:tensorflow:global_step/sec: 4.07952 +INFO:tensorflow:step = 23101, loss = 0.778045, precision = 0.867188 (24.513 sec) +INFO:tensorflow:global_step/sec: 4.5185 +INFO:tensorflow:step = 23201, loss = 0.710205, precision = 0.875 (22.136 sec) +INFO:tensorflow:global_step/sec: 4.51508 +INFO:tensorflow:step = 23301, loss = 0.704065, precision = 0.859375 (22.144 sec) +INFO:tensorflow:global_step/sec: 4.51259 +INFO:tensorflow:step = 23401, loss = 0.707262, precision = 0.867188 (22.160 sec) +Saved checkpoint after 60 epoch(s) to data/resnet164_b/checkpoints/00060... +INFO:tensorflow:global_step/sec: 4.09388 +INFO:tensorflow:step = 23501, loss = 0.712215, precision = 0.882812 (24.427 sec) +INFO:tensorflow:global_step/sec: 4.51435 +INFO:tensorflow:step = 23601, loss = 0.720241, precision = 0.859375 (22.155 sec) +INFO:tensorflow:global_step/sec: 4.50287 +INFO:tensorflow:step = 23701, loss = 0.662755, precision = 0.890625 (22.204 sec) +INFO:tensorflow:global_step/sec: 4.51653 +INFO:tensorflow:step = 23801, loss = 0.615506, precision = 0.914062 (22.141 sec) +Saved checkpoint after 61 epoch(s) to data/resnet164_b/checkpoints/00061... +INFO:tensorflow:global_step/sec: 4.03833 +INFO:tensorflow:step = 23901, loss = 0.560069, precision = 0.921875 (24.764 sec) +INFO:tensorflow:global_step/sec: 4.5082 +INFO:tensorflow:step = 24001, loss = 0.673791, precision = 0.890625 (22.181 sec) +INFO:tensorflow:global_step/sec: 4.50431 +INFO:tensorflow:step = 24101, loss = 0.664941, precision = 0.875 (22.201 sec) +INFO:tensorflow:global_step/sec: 4.5047 +INFO:tensorflow:step = 24201, loss = 0.596371, precision = 0.898438 (22.198 sec) +Saved checkpoint after 62 epoch(s) to data/resnet164_b/checkpoints/00062... +INFO:tensorflow:global_step/sec: 4.07458 +INFO:tensorflow:step = 24301, loss = 0.640613, precision = 0.882812 (24.544 sec) +INFO:tensorflow:global_step/sec: 4.51088 +INFO:tensorflow:step = 24401, loss = 0.55379, precision = 0.914062 (22.167 sec) +INFO:tensorflow:global_step/sec: 4.51044 +INFO:tensorflow:step = 24501, loss = 0.663465, precision = 0.859375 (22.171 sec) +INFO:tensorflow:global_step/sec: 4.51124 +INFO:tensorflow:step = 24601, loss = 0.600945, precision = 0.914062 (22.172 sec) +Saved checkpoint after 63 epoch(s) to data/resnet164_b/checkpoints/00063... +INFO:tensorflow:global_step/sec: 4.06719 +INFO:tensorflow:step = 24701, loss = 0.728484, precision = 0.90625 (24.582 sec) +INFO:tensorflow:global_step/sec: 4.50901 +INFO:tensorflow:step = 24801, loss = 0.679687, precision = 0.890625 (22.182 sec) +INFO:tensorflow:global_step/sec: 4.50601 +INFO:tensorflow:step = 24901, loss = 0.778772, precision = 0.851562 (22.189 sec) +INFO:tensorflow:global_step/sec: 4.51485 +INFO:tensorflow:step = 25001, loss = 0.721681, precision = 0.851562 (22.148 sec) +Saved checkpoint after 64 epoch(s) to data/resnet164_b/checkpoints/00064... +INFO:tensorflow:global_step/sec: 4.08883 +INFO:tensorflow:step = 25101, loss = 0.693595, precision = 0.867188 (24.457 sec) +INFO:tensorflow:global_step/sec: 4.51247 +INFO:tensorflow:step = 25201, loss = 0.8501, precision = 0.8125 (22.160 sec) +INFO:tensorflow:global_step/sec: 4.51184 +INFO:tensorflow:step = 25301, loss = 0.7418, precision = 0.867188 (22.164 sec) +INFO:tensorflow:global_step/sec: 4.5085 +INFO:tensorflow:step = 25401, loss = 0.58426, precision = 0.90625 (22.180 sec) +Saved checkpoint after 65 epoch(s) to data/resnet164_b/checkpoints/00065... +INFO:tensorflow:global_step/sec: 4.09215 +INFO:tensorflow:step = 25501, loss = 0.65639, precision = 0.898438 (24.438 sec) +INFO:tensorflow:global_step/sec: 4.50352 +INFO:tensorflow:step = 25601, loss = 0.739135, precision = 0.859375 (22.205 sec) +INFO:tensorflow:global_step/sec: 4.51254 +INFO:tensorflow:step = 25701, loss = 0.628259, precision = 0.90625 (22.160 sec) +INFO:tensorflow:global_step/sec: 4.51612 +INFO:tensorflow:step = 25801, loss = 0.668352, precision = 0.875 (22.143 sec) +Saved checkpoint after 66 epoch(s) to data/resnet164_b/checkpoints/00066... +INFO:tensorflow:global_step/sec: 4.0933 +INFO:tensorflow:step = 25901, loss = 0.732186, precision = 0.8125 (24.433 sec) +INFO:tensorflow:global_step/sec: 4.51495 +INFO:tensorflow:step = 26001, loss = 0.653579, precision = 0.90625 (22.146 sec) +INFO:tensorflow:global_step/sec: 4.52012 +INFO:tensorflow:step = 26101, loss = 0.709435, precision = 0.851562 (22.126 sec) +Saved checkpoint after 67 epoch(s) to data/resnet164_b/checkpoints/00067... +INFO:tensorflow:global_step/sec: 4.10623 +INFO:tensorflow:step = 26201, loss = 0.73947, precision = 0.867188 (24.350 sec) +INFO:tensorflow:global_step/sec: 4.51303 +INFO:tensorflow:step = 26301, loss = 0.575626, precision = 0.929688 (22.158 sec) +INFO:tensorflow:global_step/sec: 4.51276 +INFO:tensorflow:step = 26401, loss = 0.687858, precision = 0.90625 (22.160 sec) +INFO:tensorflow:global_step/sec: 4.5058 +INFO:tensorflow:step = 26501, loss = 0.720351, precision = 0.890625 (22.193 sec) +Saved checkpoint after 68 epoch(s) to data/resnet164_b/checkpoints/00068... +INFO:tensorflow:global_step/sec: 4.06322 +INFO:tensorflow:step = 26601, loss = 0.618126, precision = 0.898438 (24.611 sec) +INFO:tensorflow:global_step/sec: 4.51648 +INFO:tensorflow:step = 26701, loss = 0.662525, precision = 0.890625 (22.141 sec) +INFO:tensorflow:global_step/sec: 4.51721 +INFO:tensorflow:step = 26801, loss = 1.01582, precision = 0.796875 (22.138 sec) +INFO:tensorflow:global_step/sec: 4.5158 +INFO:tensorflow:step = 26901, loss = 0.736794, precision = 0.867188 (22.145 sec) +Saved checkpoint after 69 epoch(s) to data/resnet164_b/checkpoints/00069... +INFO:tensorflow:global_step/sec: 4.10828 +INFO:tensorflow:step = 27001, loss = 0.709404, precision = 0.867188 (24.341 sec) +INFO:tensorflow:global_step/sec: 4.52345 +INFO:tensorflow:step = 27101, loss = 0.604096, precision = 0.921875 (22.107 sec) +INFO:tensorflow:global_step/sec: 4.51291 +INFO:tensorflow:step = 27201, loss = 0.633058, precision = 0.890625 (22.159 sec) +INFO:tensorflow:global_step/sec: 4.51775 +INFO:tensorflow:step = 27301, loss = 0.679945, precision = 0.890625 (22.135 sec) +Saved checkpoint after 70 epoch(s) to data/resnet164_b/checkpoints/00070... +INFO:tensorflow:global_step/sec: 4.08763 +INFO:tensorflow:step = 27401, loss = 0.571532, precision = 0.921875 (24.464 sec) +INFO:tensorflow:global_step/sec: 4.51895 +INFO:tensorflow:step = 27501, loss = 0.68779, precision = 0.859375 (22.129 sec) +INFO:tensorflow:global_step/sec: 4.52257 +INFO:tensorflow:step = 27601, loss = 0.687335, precision = 0.875 (22.111 sec) +INFO:tensorflow:global_step/sec: 4.51231 +INFO:tensorflow:step = 27701, loss = 0.585981, precision = 0.921875 (22.161 sec) +Saved checkpoint after 71 epoch(s) to data/resnet164_b/checkpoints/00071... +INFO:tensorflow:global_step/sec: 4.11668 +INFO:tensorflow:step = 27801, loss = 0.630532, precision = 0.90625 (24.292 sec) +INFO:tensorflow:global_step/sec: 4.51896 +INFO:tensorflow:step = 27901, loss = 0.725204, precision = 0.859375 (22.130 sec) +INFO:tensorflow:global_step/sec: 4.51276 +INFO:tensorflow:step = 28001, loss = 0.685199, precision = 0.898438 (22.158 sec) +INFO:tensorflow:global_step/sec: 4.52013 +INFO:tensorflow:step = 28101, loss = 0.658979, precision = 0.890625 (22.124 sec) +Saved checkpoint after 72 epoch(s) to data/resnet164_b/checkpoints/00072... +INFO:tensorflow:global_step/sec: 4.08865 +INFO:tensorflow:step = 28201, loss = 0.584161, precision = 0.90625 (24.459 sec) +INFO:tensorflow:global_step/sec: 4.51432 +INFO:tensorflow:step = 28301, loss = 0.632554, precision = 0.90625 (22.151 sec) +INFO:tensorflow:global_step/sec: 4.51169 +INFO:tensorflow:step = 28401, loss = 0.704257, precision = 0.867188 (22.166 sec) +INFO:tensorflow:global_step/sec: 4.51838 +INFO:tensorflow:step = 28501, loss = 0.718569, precision = 0.867188 (22.130 sec) +Saved checkpoint after 73 epoch(s) to data/resnet164_b/checkpoints/00073... +INFO:tensorflow:global_step/sec: 4.1005 +INFO:tensorflow:step = 28601, loss = 0.649858, precision = 0.921875 (24.387 sec) +INFO:tensorflow:global_step/sec: 4.5169 +INFO:tensorflow:step = 28701, loss = 0.579453, precision = 0.90625 (22.139 sec) +INFO:tensorflow:global_step/sec: 4.52049 +INFO:tensorflow:step = 28801, loss = 0.824635, precision = 0.820312 (22.121 sec) +INFO:tensorflow:global_step/sec: 4.51323 +INFO:tensorflow:step = 28901, loss = 0.704588, precision = 0.875 (22.159 sec) +Saved checkpoint after 74 epoch(s) to data/resnet164_b/checkpoints/00074... +INFO:tensorflow:global_step/sec: 4.08126 +INFO:tensorflow:step = 29001, loss = 0.717963, precision = 0.84375 (24.501 sec) +INFO:tensorflow:global_step/sec: 4.526 +INFO:tensorflow:step = 29101, loss = 0.601547, precision = 0.914062 (22.096 sec) +INFO:tensorflow:global_step/sec: 4.51577 +INFO:tensorflow:step = 29201, loss = 0.656978, precision = 0.890625 (22.147 sec) +INFO:tensorflow:global_step/sec: 4.51473 +INFO:tensorflow:step = 29301, loss = 0.613037, precision = 0.90625 (22.146 sec) +Saved checkpoint after 75 epoch(s) to data/resnet164_b/checkpoints/00075... +INFO:tensorflow:global_step/sec: 4.0622 +INFO:tensorflow:step = 29401, loss = 0.588532, precision = 0.90625 (24.619 sec) +INFO:tensorflow:global_step/sec: 4.50805 +INFO:tensorflow:step = 29501, loss = 0.590555, precision = 0.929688 (22.183 sec) +INFO:tensorflow:global_step/sec: 4.51671 +INFO:tensorflow:step = 29601, loss = 0.644707, precision = 0.921875 (22.137 sec) +INFO:tensorflow:global_step/sec: 4.51024 +INFO:tensorflow:step = 29701, loss = 0.744603, precision = 0.859375 (22.172 sec) +Saved checkpoint after 76 epoch(s) to data/resnet164_b/checkpoints/00076... +INFO:tensorflow:global_step/sec: 4.02979 +INFO:tensorflow:step = 29801, loss = 0.770405, precision = 0.867188 (24.817 sec) +INFO:tensorflow:global_step/sec: 4.51994 +INFO:tensorflow:step = 29901, loss = 0.679798, precision = 0.867188 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.51777 +INFO:tensorflow:step = 30001, loss = 0.610904, precision = 0.898438 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.51061 +INFO:tensorflow:step = 30101, loss = 0.580307, precision = 0.898438 (22.170 sec) +Saved checkpoint after 77 epoch(s) to data/resnet164_b/checkpoints/00077... +INFO:tensorflow:global_step/sec: 4.09698 +INFO:tensorflow:step = 30201, loss = 0.606376, precision = 0.90625 (24.409 sec) +INFO:tensorflow:global_step/sec: 4.51315 +INFO:tensorflow:step = 30301, loss = 0.712133, precision = 0.875 (22.158 sec) +INFO:tensorflow:global_step/sec: 4.51478 +INFO:tensorflow:step = 30401, loss = 0.677545, precision = 0.867188 (22.149 sec) +Saved checkpoint after 78 epoch(s) to data/resnet164_b/checkpoints/00078... +INFO:tensorflow:global_step/sec: 4.09494 +INFO:tensorflow:step = 30501, loss = 0.59471, precision = 0.9375 (24.419 sec) +INFO:tensorflow:global_step/sec: 4.5159 +INFO:tensorflow:step = 30601, loss = 0.662802, precision = 0.890625 (22.144 sec) +INFO:tensorflow:global_step/sec: 4.50767 +INFO:tensorflow:step = 30701, loss = 0.636622, precision = 0.90625 (22.186 sec) +INFO:tensorflow:global_step/sec: 4.50309 +INFO:tensorflow:step = 30801, loss = 0.764695, precision = 0.867188 (22.212 sec) +Saved checkpoint after 79 epoch(s) to data/resnet164_b/checkpoints/00079... +INFO:tensorflow:global_step/sec: 3.99835 +INFO:tensorflow:step = 30901, loss = 0.805371, precision = 0.851562 (25.003 sec) +INFO:tensorflow:global_step/sec: 4.51339 +INFO:tensorflow:step = 31001, loss = 0.565036, precision = 0.9375 (22.161 sec) +INFO:tensorflow:global_step/sec: 4.51501 +INFO:tensorflow:step = 31101, loss = 0.728807, precision = 0.84375 (22.146 sec) +INFO:tensorflow:global_step/sec: 4.51699 +INFO:tensorflow:step = 31201, loss = 0.687781, precision = 0.890625 (22.137 sec) +Saved checkpoint after 80 epoch(s) to data/resnet164_b/checkpoints/00080... +INFO:tensorflow:global_step/sec: 4.08123 +INFO:tensorflow:step = 31301, loss = 0.787573, precision = 0.851562 (24.503 sec) +INFO:tensorflow:global_step/sec: 4.51538 +INFO:tensorflow:step = 31401, loss = 0.710667, precision = 0.859375 (22.146 sec) +INFO:tensorflow:global_step/sec: 4.51667 +INFO:tensorflow:step = 31501, loss = 0.706679, precision = 0.867188 (22.140 sec) +INFO:tensorflow:global_step/sec: 4.51433 +INFO:tensorflow:step = 31601, loss = 0.67832, precision = 0.890625 (22.151 sec) +Saved checkpoint after 81 epoch(s) to data/resnet164_b/checkpoints/00081... +INFO:tensorflow:global_step/sec: 4.09743 +INFO:tensorflow:step = 31701, loss = 0.589937, precision = 0.929688 (24.406 sec) +INFO:tensorflow:global_step/sec: 4.5224 +INFO:tensorflow:step = 31801, loss = 0.665428, precision = 0.898438 (22.112 sec) +INFO:tensorflow:global_step/sec: 4.51855 +INFO:tensorflow:step = 31901, loss = 0.680854, precision = 0.921875 (22.131 sec) +INFO:tensorflow:global_step/sec: 4.51191 +INFO:tensorflow:step = 32001, loss = 0.557993, precision = 0.929688 (22.163 sec) +Saved checkpoint after 82 epoch(s) to data/resnet164_b/checkpoints/00082... +INFO:tensorflow:global_step/sec: 4.08916 +INFO:tensorflow:step = 32101, loss = 0.589365, precision = 0.914062 (24.455 sec) +INFO:tensorflow:global_step/sec: 4.51713 +INFO:tensorflow:step = 32201, loss = 0.669479, precision = 0.875 (22.139 sec) +INFO:tensorflow:global_step/sec: 4.51569 +INFO:tensorflow:step = 32301, loss = 0.745459, precision = 0.882812 (22.144 sec) +INFO:tensorflow:global_step/sec: 4.51435 +INFO:tensorflow:step = 32401, loss = 0.647559, precision = 0.867188 (22.153 sec) +Saved checkpoint after 83 epoch(s) to data/resnet164_b/checkpoints/00083... +INFO:tensorflow:global_step/sec: 4.09504 +INFO:tensorflow:step = 32501, loss = 0.684583, precision = 0.890625 (24.420 sec) +INFO:tensorflow:global_step/sec: 4.5115 +INFO:tensorflow:step = 32601, loss = 0.740362, precision = 0.84375 (22.166 sec) +INFO:tensorflow:global_step/sec: 4.52589 +INFO:tensorflow:step = 32701, loss = 0.686008, precision = 0.867188 (22.094 sec) +INFO:tensorflow:global_step/sec: 4.50531 +INFO:tensorflow:step = 32801, loss = 0.637008, precision = 0.90625 (22.196 sec) +Saved checkpoint after 84 epoch(s) to data/resnet164_b/checkpoints/00084... +INFO:tensorflow:global_step/sec: 4.0863 +INFO:tensorflow:step = 32901, loss = 0.683596, precision = 0.875 (24.476 sec) +INFO:tensorflow:global_step/sec: 4.51867 +INFO:tensorflow:step = 33001, loss = 0.610996, precision = 0.914062 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.5179 +INFO:tensorflow:step = 33101, loss = 0.701944, precision = 0.859375 (22.135 sec) +INFO:tensorflow:global_step/sec: 4.52108 +INFO:tensorflow:step = 33201, loss = 0.692392, precision = 0.851562 (22.117 sec) +Saved checkpoint after 85 epoch(s) to data/resnet164_b/checkpoints/00085... +INFO:tensorflow:global_step/sec: 4.05591 +INFO:tensorflow:step = 33301, loss = 0.757311, precision = 0.875 (24.656 sec) +INFO:tensorflow:global_step/sec: 4.51689 +INFO:tensorflow:step = 33401, loss = 0.520617, precision = 0.953125 (22.142 sec) +INFO:tensorflow:global_step/sec: 4.51516 +INFO:tensorflow:step = 33501, loss = 0.589176, precision = 0.914062 (22.145 sec) +INFO:tensorflow:global_step/sec: 4.51504 +INFO:tensorflow:step = 33601, loss = 0.637124, precision = 0.90625 (22.150 sec) +Saved checkpoint after 86 epoch(s) to data/resnet164_b/checkpoints/00086... +INFO:tensorflow:global_step/sec: 4.0959 +INFO:tensorflow:step = 33701, loss = 0.913795, precision = 0.804688 (24.413 sec) +INFO:tensorflow:global_step/sec: 4.51733 +INFO:tensorflow:step = 33801, loss = 0.668349, precision = 0.859375 (22.137 sec) +INFO:tensorflow:global_step/sec: 4.51257 +INFO:tensorflow:step = 33901, loss = 0.722847, precision = 0.875 (22.160 sec) +INFO:tensorflow:global_step/sec: 4.50902 +INFO:tensorflow:step = 34001, loss = 0.562211, precision = 0.9375 (22.178 sec) +Saved checkpoint after 87 epoch(s) to data/resnet164_b/checkpoints/00087... +INFO:tensorflow:global_step/sec: 4.06523 +INFO:tensorflow:step = 34101, loss = 0.699409, precision = 0.851562 (24.599 sec) +INFO:tensorflow:global_step/sec: 4.52074 +INFO:tensorflow:step = 34201, loss = 0.632394, precision = 0.882812 (22.122 sec) +INFO:tensorflow:global_step/sec: 4.51649 +INFO:tensorflow:step = 34301, loss = 0.715297, precision = 0.890625 (22.139 sec) +INFO:tensorflow:global_step/sec: 4.51064 +INFO:tensorflow:step = 34401, loss = 0.667021, precision = 0.890625 (22.170 sec) +Saved checkpoint after 88 epoch(s) to data/resnet164_b/checkpoints/00088... +INFO:tensorflow:global_step/sec: 4.10061 +INFO:tensorflow:step = 34501, loss = 0.707795, precision = 0.859375 (24.387 sec) +INFO:tensorflow:global_step/sec: 4.51508 +INFO:tensorflow:step = 34601, loss = 0.717691, precision = 0.882812 (22.148 sec) +INFO:tensorflow:global_step/sec: 4.51585 +INFO:tensorflow:step = 34701, loss = 0.698484, precision = 0.867188 (22.144 sec) +Saved checkpoint after 89 epoch(s) to data/resnet164_b/checkpoints/00089... +INFO:tensorflow:global_step/sec: 3.96087 +INFO:tensorflow:step = 34801, loss = 0.602452, precision = 0.90625 (25.246 sec) +INFO:tensorflow:global_step/sec: 4.51928 +INFO:tensorflow:step = 34901, loss = 0.613141, precision = 0.890625 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.52462 +INFO:tensorflow:step = 35001, loss = 0.705613, precision = 0.84375 (22.101 sec) +INFO:tensorflow:global_step/sec: 4.51925 +INFO:tensorflow:step = 35101, loss = 0.650273, precision = 0.90625 (22.127 sec) +Saved checkpoint after 90 epoch(s) to data/resnet164_b/checkpoints/00090... +INFO:tensorflow:global_step/sec: 4.09843 +INFO:tensorflow:step = 35201, loss = 0.828333, precision = 0.8125 (24.400 sec) +INFO:tensorflow:global_step/sec: 4.5103 +INFO:tensorflow:step = 35301, loss = 0.642054, precision = 0.890625 (22.172 sec) +INFO:tensorflow:global_step/sec: 4.51185 +INFO:tensorflow:step = 35401, loss = 0.671067, precision = 0.867188 (22.163 sec) +INFO:tensorflow:global_step/sec: 4.51768 +INFO:tensorflow:step = 35501, loss = 0.793946, precision = 0.890625 (22.135 sec) +Saved checkpoint after 91 epoch(s) to data/resnet164_b/checkpoints/00091... +INFO:tensorflow:global_step/sec: 4.05502 +INFO:tensorflow:step = 35601, loss = 0.600973, precision = 0.914062 (24.663 sec) +INFO:tensorflow:global_step/sec: 4.52019 +INFO:tensorflow:step = 35701, loss = 0.607117, precision = 0.882812 (22.123 sec) +INFO:tensorflow:global_step/sec: 4.51773 +INFO:tensorflow:step = 35801, loss = 0.501333, precision = 0.9375 (22.133 sec) +INFO:tensorflow:global_step/sec: 4.51111 +INFO:tensorflow:step = 35901, loss = 0.469347, precision = 0.960938 (22.168 sec) +Saved checkpoint after 92 epoch(s) to data/resnet164_b/checkpoints/00092... +INFO:tensorflow:global_step/sec: 4.05613 +INFO:tensorflow:step = 36001, loss = 0.400758, precision = 0.992188 (24.654 sec) +INFO:tensorflow:global_step/sec: 4.50585 +INFO:tensorflow:step = 36101, loss = 0.422982, precision = 0.976562 (22.195 sec) +INFO:tensorflow:global_step/sec: 4.51691 +INFO:tensorflow:step = 36201, loss = 0.432193, precision = 0.96875 (22.138 sec) +INFO:tensorflow:global_step/sec: 4.51724 +INFO:tensorflow:step = 36301, loss = 0.422697, precision = 0.984375 (22.137 sec) +Saved checkpoint after 93 epoch(s) to data/resnet164_b/checkpoints/00093... +INFO:tensorflow:global_step/sec: 4.08563 +INFO:tensorflow:step = 36401, loss = 0.439661, precision = 0.96875 (24.477 sec) +INFO:tensorflow:global_step/sec: 4.51875 +INFO:tensorflow:step = 36501, loss = 0.420348, precision = 0.960938 (22.131 sec) +INFO:tensorflow:global_step/sec: 4.52175 +INFO:tensorflow:step = 36601, loss = 0.402664, precision = 0.976562 (22.113 sec) +INFO:tensorflow:global_step/sec: 4.51694 +INFO:tensorflow:step = 36701, loss = 0.452501, precision = 0.96875 (22.144 sec) +Saved checkpoint after 94 epoch(s) to data/resnet164_b/checkpoints/00094... +INFO:tensorflow:global_step/sec: 4.08587 +INFO:tensorflow:step = 36801, loss = 0.366505, precision = 0.992188 (24.470 sec) +INFO:tensorflow:global_step/sec: 4.51572 +INFO:tensorflow:step = 36901, loss = 0.404821, precision = 0.96875 (22.147 sec) +INFO:tensorflow:global_step/sec: 4.5194 +INFO:tensorflow:step = 37001, loss = 0.387982, precision = 0.984375 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.52014 +INFO:tensorflow:step = 37101, loss = 0.429755, precision = 0.945312 (22.121 sec) +Saved checkpoint after 95 epoch(s) to data/resnet164_b/checkpoints/00095... +INFO:tensorflow:global_step/sec: 4.05992 +INFO:tensorflow:step = 37201, loss = 0.410836, precision = 0.945312 (24.630 sec) +INFO:tensorflow:global_step/sec: 4.51356 +INFO:tensorflow:step = 37301, loss = 0.38949, precision = 0.960938 (22.155 sec) +INFO:tensorflow:global_step/sec: 4.51685 +INFO:tensorflow:step = 37401, loss = 0.384978, precision = 0.96875 (22.139 sec) +INFO:tensorflow:global_step/sec: 4.51694 +INFO:tensorflow:step = 37501, loss = 0.414405, precision = 0.953125 (22.139 sec) +Saved checkpoint after 96 epoch(s) to data/resnet164_b/checkpoints/00096... +INFO:tensorflow:global_step/sec: 4.06315 +INFO:tensorflow:step = 37601, loss = 0.331945, precision = 0.992188 (24.611 sec) +INFO:tensorflow:global_step/sec: 4.51345 +INFO:tensorflow:step = 37701, loss = 0.384437, precision = 0.96875 (22.156 sec) +INFO:tensorflow:global_step/sec: 4.51261 +INFO:tensorflow:step = 37801, loss = 0.337161, precision = 0.992188 (22.160 sec) +INFO:tensorflow:global_step/sec: 4.50616 +INFO:tensorflow:step = 37901, loss = 0.331096, precision = 0.992188 (22.195 sec) +Saved checkpoint after 97 epoch(s) to data/resnet164_b/checkpoints/00097... +INFO:tensorflow:global_step/sec: 4.09177 +INFO:tensorflow:step = 38001, loss = 0.337714, precision = 0.976562 (24.438 sec) +INFO:tensorflow:global_step/sec: 4.52216 +INFO:tensorflow:step = 38101, loss = 0.306939, precision = 0.992188 (22.114 sec) +INFO:tensorflow:global_step/sec: 4.51783 +INFO:tensorflow:step = 38201, loss = 0.4192, precision = 0.9375 (22.132 sec) +INFO:tensorflow:global_step/sec: 4.5166 +INFO:tensorflow:step = 38301, loss = 0.357518, precision = 0.96875 (22.144 sec) +Saved checkpoint after 98 epoch(s) to data/resnet164_b/checkpoints/00098... +INFO:tensorflow:global_step/sec: 4.06839 +INFO:tensorflow:step = 38401, loss = 0.320755, precision = 0.992188 (24.579 sec) +INFO:tensorflow:global_step/sec: 4.50803 +INFO:tensorflow:step = 38501, loss = 0.333695, precision = 0.96875 (22.180 sec) +INFO:tensorflow:global_step/sec: 4.51174 +INFO:tensorflow:step = 38601, loss = 0.339277, precision = 0.96875 (22.164 sec) +INFO:tensorflow:global_step/sec: 4.51501 +INFO:tensorflow:step = 38701, loss = 0.300563, precision = 0.992188 (22.151 sec) +Saved checkpoint after 99 epoch(s) to data/resnet164_b/checkpoints/00099... +INFO:tensorflow:global_step/sec: 4.02714 +INFO:tensorflow:step = 38801, loss = 0.361986, precision = 0.953125 (24.829 sec) +INFO:tensorflow:global_step/sec: 4.521 +INFO:tensorflow:step = 38901, loss = 0.322947, precision = 0.96875 (22.121 sec) +INFO:tensorflow:global_step/sec: 4.52136 +INFO:tensorflow:step = 39001, loss = 0.35883, precision = 0.976562 (22.115 sec) +Saved checkpoint after 100 epoch(s) to data/resnet164_b/checkpoints/00100... +INFO:tensorflow:global_step/sec: 4.08924 +INFO:tensorflow:step = 39101, loss = 0.292474, precision = 0.992188 (24.455 sec) +INFO:tensorflow:global_step/sec: 4.51635 +INFO:tensorflow:step = 39201, loss = 0.335952, precision = 0.976562 (22.147 sec) +INFO:tensorflow:global_step/sec: 4.51872 +INFO:tensorflow:step = 39301, loss = 0.307079, precision = 0.984375 (22.127 sec) +INFO:tensorflow:global_step/sec: 4.5157 +INFO:tensorflow:step = 39401, loss = 0.301554, precision = 0.984375 (22.143 sec) +Saved checkpoint after 101 epoch(s) to data/resnet164_b/checkpoints/00101... +INFO:tensorflow:global_step/sec: 4.10217 +INFO:tensorflow:step = 39501, loss = 0.28813, precision = 0.992188 (24.379 sec) +INFO:tensorflow:global_step/sec: 4.51322 +INFO:tensorflow:step = 39601, loss = 0.276685, precision = 0.992188 (22.160 sec) +INFO:tensorflow:global_step/sec: 4.51495 +INFO:tensorflow:step = 39701, loss = 0.304966, precision = 0.984375 (22.144 sec) +INFO:tensorflow:global_step/sec: 4.51627 +INFO:tensorflow:step = 39801, loss = 0.317501, precision = 0.96875 (22.143 sec) +Saved checkpoint after 102 epoch(s) to data/resnet164_b/checkpoints/00102... +INFO:tensorflow:global_step/sec: 4.06845 +INFO:tensorflow:step = 39901, loss = 0.310645, precision = 0.976562 (24.578 sec) +INFO:tensorflow:global_step/sec: 4.51912 +INFO:tensorflow:step = 40001, loss = 0.298421, precision = 0.984375 (22.129 sec) +INFO:tensorflow:global_step/sec: 4.51118 +INFO:tensorflow:step = 40101, loss = 0.288066, precision = 0.984375 (22.167 sec) +INFO:tensorflow:global_step/sec: 4.51716 +INFO:tensorflow:step = 40201, loss = 0.257359, precision = 1.0 (22.138 sec) +Saved checkpoint after 103 epoch(s) to data/resnet164_b/checkpoints/00103... +INFO:tensorflow:global_step/sec: 4.10511 +INFO:tensorflow:step = 40301, loss = 0.27457, precision = 0.992188 (24.360 sec) +INFO:tensorflow:global_step/sec: 4.5251 +INFO:tensorflow:step = 40401, loss = 0.328259, precision = 0.960938 (22.100 sec) +INFO:tensorflow:global_step/sec: 4.52106 +INFO:tensorflow:step = 40501, loss = 0.332834, precision = 0.960938 (22.119 sec) +INFO:tensorflow:global_step/sec: 4.52228 +INFO:tensorflow:step = 40601, loss = 0.272959, precision = 0.992188 (22.113 sec) +Saved checkpoint after 104 epoch(s) to data/resnet164_b/checkpoints/00104... +INFO:tensorflow:global_step/sec: 4.10081 +INFO:tensorflow:step = 40701, loss = 0.328076, precision = 0.960938 (24.383 sec) +INFO:tensorflow:global_step/sec: 4.51539 +INFO:tensorflow:step = 40801, loss = 0.31794, precision = 0.976562 (22.146 sec) +INFO:tensorflow:global_step/sec: 4.51234 +INFO:tensorflow:step = 40901, loss = 0.241218, precision = 1.0 (22.161 sec) +INFO:tensorflow:global_step/sec: 4.51543 +INFO:tensorflow:step = 41001, loss = 0.300959, precision = 0.984375 (22.146 sec) +Saved checkpoint after 105 epoch(s) to data/resnet164_b/checkpoints/00105... +INFO:tensorflow:global_step/sec: 4.07253 +INFO:tensorflow:step = 41101, loss = 0.27139, precision = 0.976562 (24.555 sec) +INFO:tensorflow:global_step/sec: 4.51828 +INFO:tensorflow:step = 41201, loss = 0.268063, precision = 0.976562 (22.133 sec) +INFO:tensorflow:global_step/sec: 4.51129 +INFO:tensorflow:step = 41301, loss = 0.263603, precision = 0.992188 (22.167 sec) +INFO:tensorflow:global_step/sec: 4.51135 +INFO:tensorflow:step = 41401, loss = 0.28335, precision = 0.984375 (22.165 sec) +Saved checkpoint after 106 epoch(s) to data/resnet164_b/checkpoints/00106... +INFO:tensorflow:global_step/sec: 4.08893 +INFO:tensorflow:step = 41501, loss = 0.26598, precision = 0.984375 (24.456 sec) +INFO:tensorflow:global_step/sec: 4.51483 +INFO:tensorflow:step = 41601, loss = 0.226438, precision = 1.0 (22.149 sec) +INFO:tensorflow:global_step/sec: 4.52014 +INFO:tensorflow:step = 41701, loss = 0.238179, precision = 0.984375 (22.123 sec) +INFO:tensorflow:global_step/sec: 4.51437 +INFO:tensorflow:step = 41801, loss = 0.268811, precision = 0.984375 (22.152 sec) +Saved checkpoint after 107 epoch(s) to data/resnet164_b/checkpoints/00107... +INFO:tensorflow:global_step/sec: 4.05243 +INFO:tensorflow:step = 41901, loss = 0.235401, precision = 0.992188 (24.678 sec) +INFO:tensorflow:global_step/sec: 4.51633 +INFO:tensorflow:step = 42001, loss = 0.237011, precision = 0.992188 (22.141 sec) +INFO:tensorflow:global_step/sec: 4.52108 +INFO:tensorflow:step = 42101, loss = 0.230473, precision = 0.992188 (22.119 sec) +INFO:tensorflow:global_step/sec: 4.51175 +INFO:tensorflow:step = 42201, loss = 0.262431, precision = 0.976562 (22.165 sec) +Saved checkpoint after 108 epoch(s) to data/resnet164_b/checkpoints/00108... +INFO:tensorflow:global_step/sec: 4.07404 +INFO:tensorflow:step = 42301, loss = 0.230401, precision = 0.992188 (24.545 sec) +INFO:tensorflow:global_step/sec: 4.51856 +INFO:tensorflow:step = 42401, loss = 0.268836, precision = 0.984375 (22.131 sec) +INFO:tensorflow:global_step/sec: 4.52219 +INFO:tensorflow:step = 42501, loss = 0.217173, precision = 1.0 (22.118 sec) +INFO:tensorflow:global_step/sec: 4.51993 +INFO:tensorflow:step = 42601, loss = 0.239828, precision = 0.992188 (22.120 sec) +Saved checkpoint after 109 epoch(s) to data/resnet164_b/checkpoints/00109... +INFO:tensorflow:global_step/sec: 4.09198 +INFO:tensorflow:step = 42701, loss = 0.303131, precision = 0.96875 (24.438 sec) +INFO:tensorflow:global_step/sec: 4.52128 +INFO:tensorflow:step = 42801, loss = 0.255921, precision = 0.984375 (22.119 sec) +INFO:tensorflow:global_step/sec: 4.51363 +INFO:tensorflow:step = 42901, loss = 0.243714, precision = 0.984375 (22.153 sec) +INFO:tensorflow:global_step/sec: 4.52191 +INFO:tensorflow:step = 43001, loss = 0.266359, precision = 0.976562 (22.115 sec) +Saved checkpoint after 110 epoch(s) to data/resnet164_b/checkpoints/00110... +INFO:tensorflow:global_step/sec: 3.96949 +INFO:tensorflow:step = 43101, loss = 0.233758, precision = 0.984375 (25.192 sec) +INFO:tensorflow:global_step/sec: 4.51734 +INFO:tensorflow:step = 43201, loss = 0.258458, precision = 0.992188 (22.138 sec) +INFO:tensorflow:global_step/sec: 4.51586 +INFO:tensorflow:step = 43301, loss = 0.220031, precision = 0.992188 (22.144 sec) +Saved checkpoint after 111 epoch(s) to data/resnet164_b/checkpoints/00111... +INFO:tensorflow:global_step/sec: 4.02318 +INFO:tensorflow:step = 43401, loss = 0.222772, precision = 0.992188 (24.855 sec) +INFO:tensorflow:global_step/sec: 4.51437 +INFO:tensorflow:step = 43501, loss = 0.227378, precision = 0.992188 (22.153 sec) +INFO:tensorflow:global_step/sec: 4.51667 +INFO:tensorflow:step = 43601, loss = 0.281506, precision = 0.96875 (22.139 sec) +INFO:tensorflow:global_step/sec: 4.51711 +INFO:tensorflow:step = 43701, loss = 0.27788, precision = 0.96875 (22.138 sec) +Saved checkpoint after 112 epoch(s) to data/resnet164_b/checkpoints/00112... +INFO:tensorflow:global_step/sec: 4.10912 +INFO:tensorflow:step = 43801, loss = 0.28473, precision = 0.953125 (24.336 sec) +INFO:tensorflow:global_step/sec: 4.51649 +INFO:tensorflow:step = 43901, loss = 0.224621, precision = 0.992188 (22.141 sec) +INFO:tensorflow:global_step/sec: 4.5198 +INFO:tensorflow:step = 44001, loss = 0.304335, precision = 0.960938 (22.126 sec) +INFO:tensorflow:global_step/sec: 4.51897 +INFO:tensorflow:step = 44101, loss = 0.30714, precision = 0.976562 (22.130 sec) +Saved checkpoint after 113 epoch(s) to data/resnet164_b/checkpoints/00113... +INFO:tensorflow:global_step/sec: 4.09965 +INFO:tensorflow:step = 44201, loss = 0.283379, precision = 0.96875 (24.391 sec) +INFO:tensorflow:global_step/sec: 4.51338 +INFO:tensorflow:step = 44301, loss = 0.315951, precision = 0.960938 (22.157 sec) +INFO:tensorflow:global_step/sec: 4.51429 +INFO:tensorflow:step = 44401, loss = 0.239134, precision = 0.984375 (22.152 sec) +INFO:tensorflow:global_step/sec: 4.52094 +INFO:tensorflow:step = 44501, loss = 0.219059, precision = 0.984375 (22.118 sec) +Saved checkpoint after 114 epoch(s) to data/resnet164_b/checkpoints/00114... +INFO:tensorflow:global_step/sec: 4.09553 +INFO:tensorflow:step = 44601, loss = 0.231558, precision = 0.976562 (24.416 sec) +INFO:tensorflow:global_step/sec: 4.51628 +INFO:tensorflow:step = 44701, loss = 0.236313, precision = 0.984375 (22.143 sec) +INFO:tensorflow:global_step/sec: 4.52231 +INFO:tensorflow:step = 44801, loss = 0.370998, precision = 0.953125 (22.113 sec) +INFO:tensorflow:global_step/sec: 4.52384 +INFO:tensorflow:step = 44901, loss = 0.271476, precision = 0.960938 (22.105 sec) +Saved checkpoint after 115 epoch(s) to data/resnet164_b/checkpoints/00115... +INFO:tensorflow:global_step/sec: 4.08869 +INFO:tensorflow:step = 45001, loss = 0.224308, precision = 0.976562 (24.458 sec) +INFO:tensorflow:global_step/sec: 4.52075 +INFO:tensorflow:step = 45101, loss = 0.235937, precision = 0.984375 (22.120 sec) +INFO:tensorflow:global_step/sec: 4.51801 +INFO:tensorflow:step = 45201, loss = 0.217463, precision = 0.984375 (22.133 sec) +INFO:tensorflow:global_step/sec: 4.51442 +INFO:tensorflow:step = 45301, loss = 0.252468, precision = 0.984375 (22.155 sec) +Saved checkpoint after 116 epoch(s) to data/resnet164_b/checkpoints/00116... +INFO:tensorflow:global_step/sec: 4.08015 +INFO:tensorflow:step = 45401, loss = 0.23228, precision = 0.984375 (24.507 sec) +INFO:tensorflow:global_step/sec: 4.51858 +INFO:tensorflow:step = 45501, loss = 0.220303, precision = 0.984375 (22.130 sec) +INFO:tensorflow:global_step/sec: 4.52027 +INFO:tensorflow:step = 45601, loss = 0.20629, precision = 0.992188 (22.126 sec) +INFO:tensorflow:global_step/sec: 4.52242 +INFO:tensorflow:step = 45701, loss = 0.219899, precision = 0.992188 (22.109 sec) +Saved checkpoint after 117 epoch(s) to data/resnet164_b/checkpoints/00117... +INFO:tensorflow:global_step/sec: 4.09568 +INFO:tensorflow:step = 45801, loss = 0.233624, precision = 0.984375 (24.416 sec) +INFO:tensorflow:global_step/sec: 4.52328 +INFO:tensorflow:step = 45901, loss = 0.196795, precision = 0.992188 (22.108 sec) +INFO:tensorflow:global_step/sec: 4.52001 +INFO:tensorflow:step = 46001, loss = 0.246035, precision = 0.976562 (22.124 sec) +INFO:tensorflow:global_step/sec: 4.52452 +INFO:tensorflow:step = 46101, loss = 0.185073, precision = 1.0 (22.104 sec) +Saved checkpoint after 118 epoch(s) to data/resnet164_b/checkpoints/00118... +INFO:tensorflow:global_step/sec: 4.0364 +INFO:tensorflow:step = 46201, loss = 0.269653, precision = 0.96875 (24.772 sec) +INFO:tensorflow:global_step/sec: 4.52519 +INFO:tensorflow:step = 46301, loss = 0.375372, precision = 0.945312 (22.099 sec) +INFO:tensorflow:global_step/sec: 4.51715 +INFO:tensorflow:step = 46401, loss = 0.205144, precision = 0.984375 (22.138 sec) +INFO:tensorflow:global_step/sec: 4.51958 +INFO:tensorflow:step = 46501, loss = 0.251552, precision = 0.960938 (22.126 sec) +Saved checkpoint after 119 epoch(s) to data/resnet164_b/checkpoints/00119... +INFO:tensorflow:global_step/sec: 4.07924 +INFO:tensorflow:step = 46601, loss = 0.210111, precision = 0.992188 (24.517 sec) +INFO:tensorflow:global_step/sec: 4.51529 +INFO:tensorflow:step = 46701, loss = 0.213859, precision = 0.984375 (22.147 sec) +INFO:tensorflow:global_step/sec: 4.52485 +INFO:tensorflow:step = 46801, loss = 0.188096, precision = 1.0 (22.098 sec) +INFO:tensorflow:global_step/sec: 4.51715 +INFO:tensorflow:step = 46901, loss = 0.208392, precision = 0.976562 (22.139 sec) +Saved checkpoint after 120 epoch(s) to data/resnet164_b/checkpoints/00120... +INFO:tensorflow:global_step/sec: 4.07391 +INFO:tensorflow:step = 47001, loss = 0.183283, precision = 1.0 (24.547 sec) +INFO:tensorflow:global_step/sec: 4.52073 +INFO:tensorflow:step = 47101, loss = 0.225986, precision = 0.960938 (22.122 sec) +INFO:tensorflow:global_step/sec: 4.52823 +INFO:tensorflow:step = 47201, loss = 0.226042, precision = 0.96875 (22.081 sec) +INFO:tensorflow:global_step/sec: 4.51809 +INFO:tensorflow:step = 47301, loss = 0.229989, precision = 0.976562 (22.133 sec) +Saved checkpoint after 121 epoch(s) to data/resnet164_b/checkpoints/00121... +INFO:tensorflow:global_step/sec: 4.09388 +INFO:tensorflow:step = 47401, loss = 0.194572, precision = 0.992188 (24.427 sec) +INFO:tensorflow:global_step/sec: 4.52006 +INFO:tensorflow:step = 47501, loss = 0.274492, precision = 0.96875 (22.125 sec) +INFO:tensorflow:global_step/sec: 4.52622 +INFO:tensorflow:step = 47601, loss = 0.194228, precision = 0.992188 (22.093 sec) +INFO:tensorflow:global_step/sec: 4.51728 +INFO:tensorflow:step = 47701, loss = 0.192708, precision = 0.992188 (22.140 sec) +Saved checkpoint after 122 epoch(s) to data/resnet164_b/checkpoints/00122... +INFO:tensorflow:global_step/sec: 4.10049 +INFO:tensorflow:step = 47801, loss = 0.221801, precision = 0.960938 (24.384 sec) +INFO:tensorflow:global_step/sec: 4.53115 +INFO:tensorflow:step = 47901, loss = 0.19113, precision = 0.992188 (22.068 sec) +INFO:tensorflow:global_step/sec: 4.52178 +INFO:tensorflow:step = 48001, loss = 0.188703, precision = 0.992188 (22.115 sec) +Saved checkpoint after 123 epoch(s) to data/resnet164_b/checkpoints/00123... +INFO:tensorflow:global_step/sec: 4.10732 +INFO:tensorflow:step = 48101, loss = 0.210477, precision = 0.984375 (24.349 sec) +INFO:tensorflow:global_step/sec: 4.52222 +INFO:tensorflow:step = 48201, loss = 0.183562, precision = 1.0 (22.111 sec) +INFO:tensorflow:global_step/sec: 4.51713 +INFO:tensorflow:step = 48301, loss = 0.189976, precision = 0.984375 (22.138 sec) +INFO:tensorflow:global_step/sec: 4.52112 +INFO:tensorflow:step = 48401, loss = 0.246666, precision = 0.96875 (22.120 sec) +Saved checkpoint after 124 epoch(s) to data/resnet164_b/checkpoints/00124... +INFO:tensorflow:global_step/sec: 4.0837 +INFO:tensorflow:step = 48501, loss = 0.176611, precision = 1.0 (24.489 sec) +INFO:tensorflow:global_step/sec: 4.52403 +INFO:tensorflow:step = 48601, loss = 0.195693, precision = 0.984375 (22.104 sec) +INFO:tensorflow:global_step/sec: 4.51896 +INFO:tensorflow:step = 48701, loss = 0.225445, precision = 0.976562 (22.127 sec) +INFO:tensorflow:global_step/sec: 4.51789 +INFO:tensorflow:step = 48801, loss = 0.189959, precision = 0.992188 (22.134 sec) +Saved checkpoint after 125 epoch(s) to data/resnet164_b/checkpoints/00125... +INFO:tensorflow:global_step/sec: 4.08246 +INFO:tensorflow:step = 48901, loss = 0.174218, precision = 1.0 (24.495 sec) +INFO:tensorflow:global_step/sec: 4.51124 +INFO:tensorflow:step = 49001, loss = 0.279873, precision = 0.960938 (22.167 sec) +INFO:tensorflow:global_step/sec: 4.52164 +INFO:tensorflow:step = 49101, loss = 0.193153, precision = 0.984375 (22.119 sec) +INFO:tensorflow:global_step/sec: 4.52397 +INFO:tensorflow:step = 49201, loss = 0.286435, precision = 0.953125 (22.101 sec) +Saved checkpoint after 126 epoch(s) to data/resnet164_b/checkpoints/00126... +INFO:tensorflow:global_step/sec: 4.08845 +INFO:tensorflow:step = 49301, loss = 0.19083, precision = 0.984375 (24.461 sec) +INFO:tensorflow:global_step/sec: 4.52295 +INFO:tensorflow:step = 49401, loss = 0.229453, precision = 0.960938 (22.107 sec) +INFO:tensorflow:global_step/sec: 4.52019 +INFO:tensorflow:step = 49501, loss = 0.203549, precision = 0.984375 (22.123 sec) +INFO:tensorflow:global_step/sec: 4.51968 +INFO:tensorflow:step = 49601, loss = 0.190845, precision = 0.976562 (22.127 sec) +Saved checkpoint after 127 epoch(s) to data/resnet164_b/checkpoints/00127... +INFO:tensorflow:global_step/sec: 4.06891 +INFO:tensorflow:step = 49701, loss = 0.233145, precision = 0.976562 (24.575 sec) +INFO:tensorflow:global_step/sec: 4.51811 +INFO:tensorflow:step = 49801, loss = 0.252058, precision = 0.960938 (22.133 sec) +INFO:tensorflow:global_step/sec: 4.52233 +INFO:tensorflow:step = 49901, loss = 0.200834, precision = 0.984375 (22.117 sec) +INFO:tensorflow:global_step/sec: 4.5275 +INFO:tensorflow:step = 50001, loss = 0.2442, precision = 0.960938 (22.083 sec) +Saved checkpoint after 128 epoch(s) to data/resnet164_b/checkpoints/00128... +INFO:tensorflow:global_step/sec: 4.11209 +INFO:tensorflow:step = 50101, loss = 0.175281, precision = 0.992188 (24.319 sec) +INFO:tensorflow:global_step/sec: 4.52429 +INFO:tensorflow:step = 50201, loss = 0.20853, precision = 0.984375 (22.105 sec) +INFO:tensorflow:global_step/sec: 4.52416 +INFO:tensorflow:step = 50301, loss = 0.184321, precision = 0.992188 (22.102 sec) +INFO:tensorflow:global_step/sec: 4.52519 +INFO:tensorflow:step = 50401, loss = 0.169599, precision = 0.992188 (22.101 sec) +Saved checkpoint after 129 epoch(s) to data/resnet164_b/checkpoints/00129... +INFO:tensorflow:global_step/sec: 4.10744 +INFO:tensorflow:step = 50501, loss = 0.19726, precision = 0.976562 (24.343 sec) +INFO:tensorflow:global_step/sec: 4.52677 +INFO:tensorflow:step = 50601, loss = 0.250402, precision = 0.976562 (22.091 sec) +INFO:tensorflow:global_step/sec: 4.52812 +INFO:tensorflow:step = 50701, loss = 0.208133, precision = 0.976562 (22.084 sec) +INFO:tensorflow:global_step/sec: 4.52469 +INFO:tensorflow:step = 50801, loss = 0.187626, precision = 0.984375 (22.102 sec) +Saved checkpoint after 130 epoch(s) to data/resnet164_b/checkpoints/00130... +INFO:tensorflow:global_step/sec: 4.09284 +INFO:tensorflow:step = 50901, loss = 0.197659, precision = 0.984375 (24.435 sec) +INFO:tensorflow:global_step/sec: 4.51786 +INFO:tensorflow:step = 51001, loss = 0.217086, precision = 0.976562 (22.133 sec) +INFO:tensorflow:global_step/sec: 4.52616 +INFO:tensorflow:step = 51101, loss = 0.162732, precision = 0.992188 (22.091 sec) +INFO:tensorflow:global_step/sec: 4.5214 +INFO:tensorflow:step = 51201, loss = 0.252131, precision = 0.976562 (22.124 sec) +Saved checkpoint after 131 epoch(s) to data/resnet164_b/checkpoints/00131... +INFO:tensorflow:global_step/sec: 4.07653 +INFO:tensorflow:step = 51301, loss = 0.200279, precision = 0.984375 (24.524 sec) +INFO:tensorflow:global_step/sec: 4.52315 +INFO:tensorflow:step = 51401, loss = 0.186016, precision = 0.984375 (22.108 sec) +INFO:tensorflow:global_step/sec: 4.52715 +INFO:tensorflow:step = 51501, loss = 0.212778, precision = 0.976562 (22.091 sec) +INFO:tensorflow:global_step/sec: 4.52337 +INFO:tensorflow:step = 51601, loss = 0.247976, precision = 0.960938 (22.105 sec) +Saved checkpoint after 132 epoch(s) to data/resnet164_b/checkpoints/00132... +INFO:tensorflow:global_step/sec: 4.06796 +INFO:tensorflow:step = 51701, loss = 0.204012, precision = 0.984375 (24.588 sec) +INFO:tensorflow:global_step/sec: 4.52535 +INFO:tensorflow:step = 51801, loss = 0.176756, precision = 0.984375 (22.094 sec) +INFO:tensorflow:global_step/sec: 4.52302 +INFO:tensorflow:step = 51901, loss = 0.184739, precision = 0.992188 (22.108 sec) +INFO:tensorflow:global_step/sec: 4.51981 +INFO:tensorflow:step = 52001, loss = 0.27646, precision = 0.945312 (22.124 sec) +Saved checkpoint after 133 epoch(s) to data/resnet164_b/checkpoints/00133... +INFO:tensorflow:global_step/sec: 4.08835 +INFO:tensorflow:step = 52101, loss = 0.220779, precision = 0.960938 (24.460 sec) +INFO:tensorflow:global_step/sec: 4.52162 +INFO:tensorflow:step = 52201, loss = 0.175804, precision = 0.992188 (22.116 sec) +INFO:tensorflow:global_step/sec: 4.52855 +INFO:tensorflow:step = 52301, loss = 0.233874, precision = 0.953125 (22.082 sec) +Saved checkpoint after 134 epoch(s) to data/resnet164_b/checkpoints/00134... +INFO:tensorflow:global_step/sec: 4.09367 +INFO:tensorflow:step = 52401, loss = 0.206991, precision = 0.984375 (24.429 sec) +INFO:tensorflow:global_step/sec: 4.51811 +INFO:tensorflow:step = 52501, loss = 0.168824, precision = 0.984375 (22.132 sec) +INFO:tensorflow:global_step/sec: 4.52493 +INFO:tensorflow:step = 52601, loss = 0.205242, precision = 0.96875 (22.101 sec) +INFO:tensorflow:global_step/sec: 4.52659 +INFO:tensorflow:step = 52701, loss = 0.205326, precision = 0.96875 (22.091 sec) +Saved checkpoint after 135 epoch(s) to data/resnet164_b/checkpoints/00135... +INFO:tensorflow:global_step/sec: 4.08445 +INFO:tensorflow:step = 52801, loss = 0.22534, precision = 0.960938 (24.485 sec) +INFO:tensorflow:global_step/sec: 4.52111 +INFO:tensorflow:step = 52901, loss = 0.198194, precision = 0.992188 (22.117 sec) +INFO:tensorflow:global_step/sec: 4.51908 +INFO:tensorflow:step = 53001, loss = 0.280661, precision = 0.945312 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.52184 +INFO:tensorflow:step = 53101, loss = 0.170026, precision = 1.0 (22.115 sec) +Saved checkpoint after 136 epoch(s) to data/resnet164_b/checkpoints/00136... +INFO:tensorflow:global_step/sec: 4.0943 +INFO:tensorflow:step = 53201, loss = 0.263585, precision = 0.96875 (24.428 sec) +INFO:tensorflow:global_step/sec: 4.51637 +INFO:tensorflow:step = 53301, loss = 0.227448, precision = 0.984375 (22.140 sec) +INFO:tensorflow:global_step/sec: 4.519 +INFO:tensorflow:step = 53401, loss = 0.181781, precision = 0.976562 (22.127 sec) +INFO:tensorflow:global_step/sec: 4.52242 +INFO:tensorflow:step = 53501, loss = 0.164651, precision = 0.992188 (22.112 sec) +Saved checkpoint after 137 epoch(s) to data/resnet164_b/checkpoints/00137... +INFO:tensorflow:global_step/sec: 4.09613 +INFO:tensorflow:step = 53601, loss = 0.162852, precision = 0.992188 (24.414 sec) +INFO:tensorflow:global_step/sec: 4.52809 +INFO:tensorflow:step = 53701, loss = 0.153436, precision = 1.0 (22.084 sec) +INFO:tensorflow:global_step/sec: 4.52061 +INFO:tensorflow:step = 53801, loss = 0.154461, precision = 1.0 (22.121 sec) +INFO:tensorflow:global_step/sec: 4.52291 +INFO:tensorflow:step = 53901, loss = 0.16919, precision = 0.992188 (22.111 sec) +Saved checkpoint after 138 epoch(s) to data/resnet164_b/checkpoints/00138... +INFO:tensorflow:global_step/sec: 4.08744 +INFO:tensorflow:step = 54001, loss = 0.154961, precision = 1.0 (24.464 sec) +INFO:tensorflow:global_step/sec: 4.51832 +INFO:tensorflow:step = 54101, loss = 0.153937, precision = 1.0 (22.132 sec) +INFO:tensorflow:global_step/sec: 4.5279 +INFO:tensorflow:step = 54201, loss = 0.166246, precision = 0.992188 (22.085 sec) +INFO:tensorflow:global_step/sec: 4.52005 +INFO:tensorflow:step = 54301, loss = 0.152041, precision = 1.0 (22.124 sec) +Saved checkpoint after 139 epoch(s) to data/resnet164_b/checkpoints/00139... +INFO:tensorflow:global_step/sec: 4.08275 +INFO:tensorflow:step = 54401, loss = 0.15066, precision = 1.0 (24.493 sec) +INFO:tensorflow:global_step/sec: 4.52187 +INFO:tensorflow:step = 54501, loss = 0.156446, precision = 1.0 (22.115 sec) +INFO:tensorflow:global_step/sec: 4.52289 +INFO:tensorflow:step = 54601, loss = 0.150625, precision = 1.0 (22.110 sec) +INFO:tensorflow:global_step/sec: 4.51654 +INFO:tensorflow:step = 54701, loss = 0.153544, precision = 1.0 (22.141 sec) +Saved checkpoint after 140 epoch(s) to data/resnet164_b/checkpoints/00140... +INFO:tensorflow:global_step/sec: 4.0952 +INFO:tensorflow:step = 54801, loss = 0.153345, precision = 0.992188 (24.419 sec) +INFO:tensorflow:global_step/sec: 4.51917 +INFO:tensorflow:step = 54901, loss = 0.176511, precision = 0.992188 (22.133 sec) +INFO:tensorflow:global_step/sec: 4.5237 +INFO:tensorflow:step = 55001, loss = 0.14918, precision = 1.0 (22.101 sec) +INFO:tensorflow:global_step/sec: 4.52179 +INFO:tensorflow:step = 55101, loss = 0.150396, precision = 1.0 (22.115 sec) +Saved checkpoint after 141 epoch(s) to data/resnet164_b/checkpoints/00141... +INFO:tensorflow:global_step/sec: 4.02645 +INFO:tensorflow:step = 55201, loss = 0.144563, precision = 1.0 (24.836 sec) +INFO:tensorflow:global_step/sec: 4.52073 +INFO:tensorflow:step = 55301, loss = 0.148293, precision = 1.0 (22.122 sec) +INFO:tensorflow:global_step/sec: 4.52283 +INFO:tensorflow:step = 55401, loss = 0.146145, precision = 1.0 (22.108 sec) +INFO:tensorflow:global_step/sec: 4.51945 +INFO:tensorflow:step = 55501, loss = 0.146648, precision = 1.0 (22.130 sec) +Saved checkpoint after 142 epoch(s) to data/resnet164_b/checkpoints/00142... +INFO:tensorflow:global_step/sec: 4.07742 +INFO:tensorflow:step = 55601, loss = 0.149798, precision = 1.0 (24.521 sec) +INFO:tensorflow:global_step/sec: 4.52179 +INFO:tensorflow:step = 55701, loss = 0.148795, precision = 1.0 (22.115 sec) +INFO:tensorflow:global_step/sec: 4.5231 +INFO:tensorflow:step = 55801, loss = 0.151477, precision = 1.0 (22.109 sec) +INFO:tensorflow:global_step/sec: 4.52018 +INFO:tensorflow:step = 55901, loss = 0.145441, precision = 1.0 (22.123 sec) +Saved checkpoint after 143 epoch(s) to data/resnet164_b/checkpoints/00143... +INFO:tensorflow:global_step/sec: 4.07607 +INFO:tensorflow:step = 56001, loss = 0.151915, precision = 0.992188 (24.534 sec) +INFO:tensorflow:global_step/sec: 4.5274 +INFO:tensorflow:step = 56101, loss = 0.144318, precision = 1.0 (22.088 sec) +INFO:tensorflow:global_step/sec: 4.523 +INFO:tensorflow:step = 56201, loss = 0.152579, precision = 1.0 (22.109 sec) +INFO:tensorflow:global_step/sec: 4.51379 +INFO:tensorflow:step = 56301, loss = 0.168564, precision = 0.992188 (22.159 sec) +Saved checkpoint after 144 epoch(s) to data/resnet164_b/checkpoints/00144... +INFO:tensorflow:global_step/sec: 4.07026 +INFO:tensorflow:step = 56401, loss = 0.147217, precision = 1.0 (24.564 sec) +INFO:tensorflow:global_step/sec: 4.51356 +INFO:tensorflow:step = 56501, loss = 0.151902, precision = 0.992188 (22.156 sec) +INFO:tensorflow:global_step/sec: 4.51967 +INFO:tensorflow:step = 56601, loss = 0.144333, precision = 1.0 (22.127 sec) +Saved checkpoint after 145 epoch(s) to data/resnet164_b/checkpoints/00145... +INFO:tensorflow:global_step/sec: 4.09642 +INFO:tensorflow:step = 56701, loss = 0.150701, precision = 1.0 (24.410 sec) +INFO:tensorflow:global_step/sec: 4.52809 +INFO:tensorflow:step = 56801, loss = 0.144524, precision = 1.0 (22.084 sec) +INFO:tensorflow:global_step/sec: 4.51674 +INFO:tensorflow:step = 56901, loss = 0.14413, precision = 1.0 (22.140 sec) +INFO:tensorflow:global_step/sec: 4.52594 +INFO:tensorflow:step = 57001, loss = 0.142282, precision = 1.0 (22.095 sec) +Saved checkpoint after 146 epoch(s) to data/resnet164_b/checkpoints/00146... +INFO:tensorflow:global_step/sec: 4.09489 +INFO:tensorflow:step = 57101, loss = 0.148299, precision = 1.0 (24.420 sec) +INFO:tensorflow:global_step/sec: 4.51749 +INFO:tensorflow:step = 57201, loss = 0.151571, precision = 0.992188 (22.140 sec) +INFO:tensorflow:global_step/sec: 4.52406 +INFO:tensorflow:step = 57301, loss = 0.145855, precision = 1.0 (22.101 sec) +INFO:tensorflow:global_step/sec: 4.5257 +INFO:tensorflow:step = 57401, loss = 0.154163, precision = 0.992188 (22.097 sec) +Saved checkpoint after 147 epoch(s) to data/resnet164_b/checkpoints/00147... +INFO:tensorflow:global_step/sec: 4.05946 +INFO:tensorflow:step = 57501, loss = 0.140875, precision = 1.0 (24.633 sec) +INFO:tensorflow:global_step/sec: 4.51182 +INFO:tensorflow:step = 57601, loss = 0.144406, precision = 1.0 (22.163 sec) +INFO:tensorflow:global_step/sec: 4.52013 +INFO:tensorflow:step = 57701, loss = 0.143095, precision = 1.0 (22.125 sec) +INFO:tensorflow:global_step/sec: 4.51604 +INFO:tensorflow:step = 57801, loss = 0.144782, precision = 1.0 (22.141 sec) +Saved checkpoint after 148 epoch(s) to data/resnet164_b/checkpoints/00148... +INFO:tensorflow:global_step/sec: 4.09397 +INFO:tensorflow:step = 57901, loss = 0.143724, precision = 1.0 (24.428 sec) +INFO:tensorflow:global_step/sec: 4.52125 +INFO:tensorflow:step = 58001, loss = 0.144173, precision = 1.0 (22.119 sec) +INFO:tensorflow:global_step/sec: 4.51456 +INFO:tensorflow:step = 58101, loss = 0.171087, precision = 0.992188 (22.150 sec) +INFO:tensorflow:global_step/sec: 4.52439 +INFO:tensorflow:step = 58201, loss = 0.140617, precision = 1.0 (22.100 sec) +Saved checkpoint after 149 epoch(s) to data/resnet164_b/checkpoints/00149... +INFO:tensorflow:global_step/sec: 4.10441 +INFO:tensorflow:step = 58301, loss = 0.141327, precision = 1.0 (24.368 sec) +INFO:tensorflow:global_step/sec: 4.50614 +INFO:tensorflow:step = 58401, loss = 0.143781, precision = 1.0 (22.187 sec) +INFO:tensorflow:global_step/sec: 4.51331 +INFO:tensorflow:step = 58501, loss = 0.139868, precision = 1.0 (22.159 sec) +INFO:tensorflow:global_step/sec: 4.52129 +INFO:tensorflow:step = 58601, loss = 0.144822, precision = 1.0 (22.115 sec) +Saved checkpoint after 150 epoch(s) to data/resnet164_b/checkpoints/00150... +INFO:tensorflow:global_step/sec: 4.09983 +INFO:tensorflow:step = 58701, loss = 0.143746, precision = 1.0 (24.394 sec) +INFO:tensorflow:global_step/sec: 4.51158 +INFO:tensorflow:step = 58801, loss = 0.141661, precision = 1.0 (22.165 sec) +INFO:tensorflow:global_step/sec: 4.51977 +INFO:tensorflow:step = 58901, loss = 0.14415, precision = 0.992188 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.51984 +INFO:tensorflow:step = 59001, loss = 0.142446, precision = 1.0 (22.123 sec) +Saved checkpoint after 151 epoch(s) to data/resnet164_b/checkpoints/00151... +INFO:tensorflow:global_step/sec: 4.10424 +INFO:tensorflow:step = 59101, loss = 0.139882, precision = 1.0 (24.361 sec) +INFO:tensorflow:global_step/sec: 4.51311 +INFO:tensorflow:step = 59201, loss = 0.145344, precision = 1.0 (22.161 sec) +INFO:tensorflow:global_step/sec: 4.51863 +INFO:tensorflow:step = 59301, loss = 0.144168, precision = 1.0 (22.127 sec) +INFO:tensorflow:global_step/sec: 4.51891 +INFO:tensorflow:step = 59401, loss = 0.139091, precision = 1.0 (22.129 sec) +Saved checkpoint after 152 epoch(s) to data/resnet164_b/checkpoints/00152... +INFO:tensorflow:global_step/sec: 4.06073 +INFO:tensorflow:step = 59501, loss = 0.139399, precision = 1.0 (24.626 sec) +INFO:tensorflow:global_step/sec: 4.51805 +INFO:tensorflow:step = 59601, loss = 0.140712, precision = 1.0 (22.139 sec) +INFO:tensorflow:global_step/sec: 4.52238 +INFO:tensorflow:step = 59701, loss = 0.138615, precision = 1.0 (22.106 sec) +INFO:tensorflow:global_step/sec: 4.52339 +INFO:tensorflow:step = 59801, loss = 0.140162, precision = 1.0 (22.107 sec) +Saved checkpoint after 153 epoch(s) to data/resnet164_b/checkpoints/00153... +INFO:tensorflow:global_step/sec: 4.04923 +INFO:tensorflow:step = 59901, loss = 0.141624, precision = 1.0 (24.696 sec) +INFO:tensorflow:global_step/sec: 4.51667 +INFO:tensorflow:step = 60001, loss = 0.141226, precision = 1.0 (22.140 sec) +INFO:tensorflow:global_step/sec: 4.51998 +INFO:tensorflow:step = 60101, loss = 0.143581, precision = 1.0 (22.124 sec) +INFO:tensorflow:global_step/sec: 4.51722 +INFO:tensorflow:step = 60201, loss = 0.137368, precision = 1.0 (22.137 sec) +Saved checkpoint after 154 epoch(s) to data/resnet164_b/checkpoints/00154... +INFO:tensorflow:global_step/sec: 4.09543 +INFO:tensorflow:step = 60301, loss = 0.138125, precision = 1.0 (24.418 sec) +INFO:tensorflow:global_step/sec: 4.52464 +INFO:tensorflow:step = 60401, loss = 0.13941, precision = 1.0 (22.101 sec) +INFO:tensorflow:global_step/sec: 4.5249 +INFO:tensorflow:step = 60501, loss = 0.136445, precision = 1.0 (22.100 sec) +INFO:tensorflow:global_step/sec: 4.51675 +INFO:tensorflow:step = 60601, loss = 0.137104, precision = 1.0 (22.140 sec) +Saved checkpoint after 155 epoch(s) to data/resnet164_b/checkpoints/00155... +INFO:tensorflow:global_step/sec: 4.08397 +INFO:tensorflow:step = 60701, loss = 0.13733, precision = 1.0 (24.486 sec) +INFO:tensorflow:global_step/sec: 4.52553 +INFO:tensorflow:step = 60801, loss = 0.138952, precision = 1.0 (22.098 sec) +INFO:tensorflow:global_step/sec: 4.52254 +INFO:tensorflow:step = 60901, loss = 0.140088, precision = 1.0 (22.117 sec) +Saved checkpoint after 156 epoch(s) to data/resnet164_b/checkpoints/00156... +INFO:tensorflow:global_step/sec: 4.12234 +INFO:tensorflow:step = 61001, loss = 0.140506, precision = 1.0 (24.251 sec) +INFO:tensorflow:global_step/sec: 4.52486 +INFO:tensorflow:step = 61101, loss = 0.137991, precision = 1.0 (22.101 sec) +INFO:tensorflow:global_step/sec: 4.52373 +INFO:tensorflow:step = 61201, loss = 0.143441, precision = 1.0 (22.106 sec) +INFO:tensorflow:global_step/sec: 4.52326 +INFO:tensorflow:step = 61301, loss = 0.139269, precision = 1.0 (22.108 sec) +Saved checkpoint after 157 epoch(s) to data/resnet164_b/checkpoints/00157... +INFO:tensorflow:global_step/sec: 4.09498 +INFO:tensorflow:step = 61401, loss = 0.135636, precision = 1.0 (24.420 sec) +INFO:tensorflow:global_step/sec: 4.51249 +INFO:tensorflow:step = 61501, loss = 0.13634, precision = 1.0 (22.162 sec) +INFO:tensorflow:global_step/sec: 4.51883 +INFO:tensorflow:step = 61601, loss = 0.13577, precision = 1.0 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.52005 +INFO:tensorflow:step = 61701, loss = 0.136482, precision = 1.0 (22.124 sec) +Saved checkpoint after 158 epoch(s) to data/resnet164_b/checkpoints/00158... +INFO:tensorflow:global_step/sec: 4.06348 +INFO:tensorflow:step = 61801, loss = 0.134962, precision = 1.0 (24.612 sec) +INFO:tensorflow:global_step/sec: 4.51871 +INFO:tensorflow:step = 61901, loss = 0.137663, precision = 1.0 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.52113 +INFO:tensorflow:step = 62001, loss = 0.144462, precision = 1.0 (22.118 sec) +INFO:tensorflow:global_step/sec: 4.51868 +INFO:tensorflow:step = 62101, loss = 0.135631, precision = 1.0 (22.136 sec) +Saved checkpoint after 159 epoch(s) to data/resnet164_b/checkpoints/00159... +INFO:tensorflow:global_step/sec: 4.08457 +INFO:tensorflow:step = 62201, loss = 0.137392, precision = 1.0 (24.477 sec) +INFO:tensorflow:global_step/sec: 4.52597 +INFO:tensorflow:step = 62301, loss = 0.135123, precision = 1.0 (22.097 sec) +INFO:tensorflow:global_step/sec: 4.51143 +INFO:tensorflow:step = 62401, loss = 0.13711, precision = 1.0 (22.164 sec) +INFO:tensorflow:global_step/sec: 4.51894 +INFO:tensorflow:step = 62501, loss = 0.134761, precision = 1.0 (22.137 sec) +Saved checkpoint after 160 epoch(s) to data/resnet164_b/checkpoints/00160... +INFO:tensorflow:global_step/sec: 4.07137 +INFO:tensorflow:step = 62601, loss = 0.135131, precision = 1.0 (24.554 sec) +INFO:tensorflow:global_step/sec: 4.52297 +INFO:tensorflow:step = 62701, loss = 0.135814, precision = 1.0 (22.109 sec) +INFO:tensorflow:global_step/sec: 4.52103 +INFO:tensorflow:step = 62801, loss = 0.136177, precision = 1.0 (22.119 sec) +INFO:tensorflow:global_step/sec: 4.51502 +INFO:tensorflow:step = 62901, loss = 0.134329, precision = 1.0 (22.154 sec) +Saved checkpoint after 161 epoch(s) to data/resnet164_b/checkpoints/00161... +INFO:tensorflow:global_step/sec: 4.07949 +INFO:tensorflow:step = 63001, loss = 0.134809, precision = 1.0 (24.508 sec) +INFO:tensorflow:global_step/sec: 4.52789 +INFO:tensorflow:step = 63101, loss = 0.164469, precision = 0.992188 (22.086 sec) +INFO:tensorflow:global_step/sec: 4.52091 +INFO:tensorflow:step = 63201, loss = 0.133652, precision = 1.0 (22.117 sec) +INFO:tensorflow:global_step/sec: 4.52504 +INFO:tensorflow:step = 63301, loss = 0.135038, precision = 1.0 (22.099 sec) +Saved checkpoint after 162 epoch(s) to data/resnet164_b/checkpoints/00162... +INFO:tensorflow:global_step/sec: 4.01226 +INFO:tensorflow:step = 63401, loss = 0.133073, precision = 1.0 (24.924 sec) +INFO:tensorflow:global_step/sec: 4.52508 +INFO:tensorflow:step = 63501, loss = 0.133122, precision = 1.0 (22.102 sec) +INFO:tensorflow:global_step/sec: 4.51659 +INFO:tensorflow:step = 63601, loss = 0.133894, precision = 1.0 (22.138 sec) +INFO:tensorflow:global_step/sec: 4.51663 +INFO:tensorflow:step = 63701, loss = 0.139313, precision = 1.0 (22.140 sec) +Saved checkpoint after 163 epoch(s) to data/resnet164_b/checkpoints/00163... +INFO:tensorflow:global_step/sec: 4.08993 +INFO:tensorflow:step = 63801, loss = 0.132333, precision = 1.0 (24.451 sec) +INFO:tensorflow:global_step/sec: 4.52282 +INFO:tensorflow:step = 63901, loss = 0.134185, precision = 1.0 (22.110 sec) +INFO:tensorflow:global_step/sec: 4.52634 +INFO:tensorflow:step = 64001, loss = 0.133299, precision = 1.0 (22.093 sec) +INFO:tensorflow:global_step/sec: 4.52639 +INFO:tensorflow:step = 64101, loss = 0.133776, precision = 1.0 (22.100 sec) +Saved checkpoint after 164 epoch(s) to data/resnet164_b/checkpoints/00164... +INFO:tensorflow:global_step/sec: 4.08433 +INFO:tensorflow:step = 64201, loss = 0.135383, precision = 1.0 (24.477 sec) +INFO:tensorflow:global_step/sec: 4.52059 +INFO:tensorflow:step = 64301, loss = 0.133727, precision = 1.0 (22.121 sec) +INFO:tensorflow:global_step/sec: 4.52243 +INFO:tensorflow:step = 64401, loss = 0.13335, precision = 1.0 (22.119 sec) +INFO:tensorflow:global_step/sec: 4.51265 +INFO:tensorflow:step = 64501, loss = 0.132995, precision = 1.0 (22.153 sec) +Saved checkpoint after 165 epoch(s) to data/resnet164_b/checkpoints/00165... +INFO:tensorflow:global_step/sec: 4.09027 +INFO:tensorflow:step = 64601, loss = 0.132204, precision = 1.0 (24.448 sec) +INFO:tensorflow:global_step/sec: 4.51938 +INFO:tensorflow:step = 64701, loss = 0.134973, precision = 1.0 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.51979 +INFO:tensorflow:step = 64801, loss = 0.137621, precision = 0.992188 (22.124 sec) +INFO:tensorflow:global_step/sec: 4.52641 +INFO:tensorflow:step = 64901, loss = 0.139335, precision = 1.0 (22.092 sec) +Saved checkpoint after 166 epoch(s) to data/resnet164_b/checkpoints/00166... +INFO:tensorflow:global_step/sec: 4.0712 +INFO:tensorflow:step = 65001, loss = 0.134148, precision = 1.0 (24.565 sec) +INFO:tensorflow:global_step/sec: 4.51929 +INFO:tensorflow:step = 65101, loss = 0.130537, precision = 1.0 (22.127 sec) +INFO:tensorflow:global_step/sec: 4.51953 +INFO:tensorflow:step = 65201, loss = 0.136398, precision = 1.0 (22.124 sec) +Saved checkpoint after 167 epoch(s) to data/resnet164_b/checkpoints/00167... +INFO:tensorflow:global_step/sec: 4.09983 +INFO:tensorflow:step = 65301, loss = 0.130525, precision = 1.0 (24.391 sec) +INFO:tensorflow:global_step/sec: 4.52116 +INFO:tensorflow:step = 65401, loss = 0.130265, precision = 1.0 (22.118 sec) +INFO:tensorflow:global_step/sec: 4.51786 +INFO:tensorflow:step = 65501, loss = 0.132704, precision = 1.0 (22.134 sec) +INFO:tensorflow:global_step/sec: 4.51332 +INFO:tensorflow:step = 65601, loss = 0.13717, precision = 1.0 (22.157 sec) +Saved checkpoint after 168 epoch(s) to data/resnet164_b/checkpoints/00168... +INFO:tensorflow:global_step/sec: 4.07003 +INFO:tensorflow:step = 65701, loss = 0.130234, precision = 1.0 (24.570 sec) +INFO:tensorflow:global_step/sec: 4.52005 +INFO:tensorflow:step = 65801, loss = 0.130366, precision = 1.0 (22.124 sec) +INFO:tensorflow:global_step/sec: 4.51876 +INFO:tensorflow:step = 65901, loss = 0.129191, precision = 1.0 (22.132 sec) +INFO:tensorflow:global_step/sec: 4.51477 +INFO:tensorflow:step = 66001, loss = 0.131959, precision = 1.0 (22.147 sec) +Saved checkpoint after 169 epoch(s) to data/resnet164_b/checkpoints/00169... +INFO:tensorflow:global_step/sec: 4.01508 +INFO:tensorflow:step = 66101, loss = 0.130111, precision = 1.0 (24.907 sec) +INFO:tensorflow:global_step/sec: 4.52041 +INFO:tensorflow:step = 66201, loss = 0.13013, precision = 1.0 (22.121 sec) +INFO:tensorflow:global_step/sec: 4.52624 +INFO:tensorflow:step = 66301, loss = 0.130194, precision = 1.0 (22.096 sec) +INFO:tensorflow:global_step/sec: 4.52318 +INFO:tensorflow:step = 66401, loss = 0.129713, precision = 1.0 (22.105 sec) +Saved checkpoint after 170 epoch(s) to data/resnet164_b/checkpoints/00170... +INFO:tensorflow:global_step/sec: 4.09439 +INFO:tensorflow:step = 66501, loss = 0.132716, precision = 1.0 (24.427 sec) +INFO:tensorflow:global_step/sec: 4.51666 +INFO:tensorflow:step = 66601, loss = 0.12865, precision = 1.0 (22.139 sec) +INFO:tensorflow:global_step/sec: 4.52211 +INFO:tensorflow:step = 66701, loss = 0.129472, precision = 1.0 (22.113 sec) +INFO:tensorflow:global_step/sec: 4.51701 +INFO:tensorflow:step = 66801, loss = 0.129192, precision = 1.0 (22.137 sec) +Saved checkpoint after 171 epoch(s) to data/resnet164_b/checkpoints/00171... +INFO:tensorflow:global_step/sec: 4.06846 +INFO:tensorflow:step = 66901, loss = 0.128074, precision = 1.0 (24.579 sec) +INFO:tensorflow:global_step/sec: 4.51077 +INFO:tensorflow:step = 67001, loss = 0.129603, precision = 1.0 (22.171 sec) +INFO:tensorflow:global_step/sec: 4.52042 +INFO:tensorflow:step = 67101, loss = 0.129657, precision = 1.0 (22.120 sec) +INFO:tensorflow:global_step/sec: 4.51842 +INFO:tensorflow:step = 67201, loss = 0.128599, precision = 1.0 (22.132 sec) +Saved checkpoint after 172 epoch(s) to data/resnet164_b/checkpoints/00172... +INFO:tensorflow:global_step/sec: 4.04888 +INFO:tensorflow:step = 67301, loss = 0.128223, precision = 1.0 (24.698 sec) +INFO:tensorflow:global_step/sec: 4.52305 +INFO:tensorflow:step = 67401, loss = 0.128834, precision = 1.0 (22.109 sec) +INFO:tensorflow:global_step/sec: 4.52386 +INFO:tensorflow:step = 67501, loss = 0.137341, precision = 0.992188 (22.105 sec) +INFO:tensorflow:global_step/sec: 4.52092 +INFO:tensorflow:step = 67601, loss = 0.136775, precision = 1.0 (22.125 sec) +Saved checkpoint after 173 epoch(s) to data/resnet164_b/checkpoints/00173... +INFO:tensorflow:global_step/sec: 4.08958 +INFO:tensorflow:step = 67701, loss = 0.127491, precision = 1.0 (24.451 sec) +INFO:tensorflow:global_step/sec: 4.51854 +INFO:tensorflow:step = 67801, loss = 0.129285, precision = 1.0 (22.128 sec) +INFO:tensorflow:global_step/sec: 4.52379 +INFO:tensorflow:step = 67901, loss = 0.131464, precision = 1.0 (22.109 sec) +INFO:tensorflow:global_step/sec: 4.51851 +INFO:tensorflow:step = 68001, loss = 0.13055, precision = 1.0 (22.128 sec) +Saved checkpoint after 174 epoch(s) to data/resnet164_b/checkpoints/00174... +INFO:tensorflow:global_step/sec: 4.08315 +INFO:tensorflow:step = 68101, loss = 0.128371, precision = 1.0 (24.495 sec) +INFO:tensorflow:global_step/sec: 4.52236 +INFO:tensorflow:step = 68201, loss = 0.126783, precision = 1.0 (22.112 sec) +INFO:tensorflow:global_step/sec: 4.52133 +INFO:tensorflow:step = 68301, loss = 0.127012, precision = 1.0 (22.116 sec) +INFO:tensorflow:global_step/sec: 4.52004 +INFO:tensorflow:step = 68401, loss = 0.126065, precision = 1.0 (22.122 sec) +Saved checkpoint after 175 epoch(s) to data/resnet164_b/checkpoints/00175... +INFO:tensorflow:global_step/sec: 4.05726 +INFO:tensorflow:step = 68501, loss = 0.129939, precision = 1.0 (24.647 sec) +INFO:tensorflow:global_step/sec: 4.5239 +INFO:tensorflow:step = 68601, loss = 0.125935, precision = 1.0 (22.105 sec) +INFO:tensorflow:global_step/sec: 4.52028 +INFO:tensorflow:step = 68701, loss = 0.128153, precision = 1.0 (22.123 sec) +INFO:tensorflow:global_step/sec: 4.52387 +INFO:tensorflow:step = 68801, loss = 0.125749, precision = 1.0 (22.105 sec) +Saved checkpoint after 176 epoch(s) to data/resnet164_b/checkpoints/00176... +INFO:tensorflow:global_step/sec: 4.05935 +INFO:tensorflow:step = 68901, loss = 0.126355, precision = 1.0 (24.635 sec) +INFO:tensorflow:global_step/sec: 4.52156 +INFO:tensorflow:step = 69001, loss = 0.125694, precision = 1.0 (22.116 sec) +INFO:tensorflow:global_step/sec: 4.51748 +INFO:tensorflow:step = 69101, loss = 0.126846, precision = 1.0 (22.136 sec) +INFO:tensorflow:global_step/sec: 4.51809 +INFO:tensorflow:step = 69201, loss = 0.126997, precision = 1.0 (22.136 sec) +Saved checkpoint after 177 epoch(s) to data/resnet164_b/checkpoints/00177... +INFO:tensorflow:global_step/sec: 4.08792 +INFO:tensorflow:step = 69301, loss = 0.125812, precision = 1.0 (24.459 sec) +INFO:tensorflow:global_step/sec: 4.5262 +INFO:tensorflow:step = 69401, loss = 0.125843, precision = 1.0 (22.095 sec) +INFO:tensorflow:global_step/sec: 4.5197 +INFO:tensorflow:step = 69501, loss = 0.124636, precision = 1.0 (22.123 sec) +Saved checkpoint after 178 epoch(s) to data/resnet164_b/checkpoints/00178... +INFO:tensorflow:global_step/sec: 4.09885 +INFO:tensorflow:step = 69601, loss = 0.125332, precision = 1.0 (24.398 sec) +INFO:tensorflow:global_step/sec: 4.5217 +INFO:tensorflow:step = 69701, loss = 0.125382, precision = 1.0 (22.115 sec) +INFO:tensorflow:global_step/sec: 4.52801 +INFO:tensorflow:step = 69801, loss = 0.124398, precision = 1.0 (22.085 sec) +INFO:tensorflow:global_step/sec: 4.5178 +INFO:tensorflow:step = 69901, loss = 0.1251, precision = 1.0 (22.140 sec) +Saved checkpoint after 179 epoch(s) to data/resnet164_b/checkpoints/00179... +INFO:tensorflow:global_step/sec: 4.10009 +INFO:tensorflow:step = 70001, loss = 0.124587, precision = 1.0 (24.384 sec) +INFO:tensorflow:global_step/sec: 4.52013 +INFO:tensorflow:step = 70101, loss = 0.124869, precision = 1.0 (22.123 sec) +INFO:tensorflow:global_step/sec: 4.52792 +INFO:tensorflow:step = 70201, loss = 0.124356, precision = 1.0 (22.085 sec) +INFO:tensorflow:global_step/sec: 4.51215 +INFO:tensorflow:step = 70301, loss = 0.127105, precision = 1.0 (22.162 sec) +Saved checkpoint after 180 epoch(s) to data/resnet164_b/checkpoints/00180... +INFO:tensorflow:global_step/sec: 4.10737 +INFO:tensorflow:step = 70401, loss = 0.125018, precision = 1.0 (24.349 sec) +INFO:tensorflow:global_step/sec: 4.51768 +INFO:tensorflow:step = 70501, loss = 0.124044, precision = 1.0 (22.133 sec) +INFO:tensorflow:global_step/sec: 4.52255 +INFO:tensorflow:step = 70601, loss = 0.123854, precision = 1.0 (22.111 sec) +INFO:tensorflow:global_step/sec: 4.52456 +INFO:tensorflow:step = 70701, loss = 0.123597, precision = 1.0 (22.107 sec) +Saved checkpoint after 181 epoch(s) to data/resnet164_b/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1p100_dawn/resnet164_nb_train.log b/tensorflow/CIFAR10/logs/1p100_dawn/resnet164_nb_train.log new file mode 100644 index 0000000..7281e8f --- /dev/null +++ b/tensorflow/CIFAR10/logs/1p100_dawn/resnet164_nb_train.log @@ -0,0 +1,2161 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/2.60m params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_10/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_10/sub2/bn2/beta (16, 16/16 params) + unit_1_10/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_11/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_11/sub2/bn2/beta (16, 16/16 params) + unit_1_11/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_12/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_12/sub2/bn2/beta (16, 16/16 params) + unit_1_12/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_13/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_13/sub2/bn2/beta (16, 16/16 params) + unit_1_13/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_14/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_14/sub2/bn2/beta (16, 16/16 params) + unit_1_14/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_15/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_15/sub2/bn2/beta (16, 16/16 params) + unit_1_15/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_16/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_16/sub2/bn2/beta (16, 16/16 params) + unit_1_16/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_17/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_17/sub2/bn2/beta (16, 16/16 params) + unit_1_17/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_18/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_18/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_18/sub2/bn2/beta (16, 16/16 params) + unit_1_18/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_19/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_19/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_19/sub2/bn2/beta (16, 16/16 params) + unit_1_19/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_20/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_20/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_20/sub2/bn2/beta (16, 16/16 params) + unit_1_20/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_21/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_21/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_21/sub2/bn2/beta (16, 16/16 params) + unit_1_21/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_22/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_22/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_22/sub2/bn2/beta (16, 16/16 params) + unit_1_22/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_23/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_23/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_23/sub2/bn2/beta (16, 16/16 params) + unit_1_23/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_24/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_24/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_24/sub2/bn2/beta (16, 16/16 params) + unit_1_24/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_25/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_25/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_25/sub2/bn2/beta (16, 16/16 params) + unit_1_25/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_26/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_26/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_26/sub2/bn2/beta (16, 16/16 params) + unit_1_26/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_3/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_4/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_5/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_6/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_7/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_8/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_9/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_9/sub2/bn2/beta (16, 16/16 params) + unit_1_9/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_10/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_10/sub2/bn2/beta (32, 32/32 params) + unit_2_10/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_11/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_11/sub2/bn2/beta (32, 32/32 params) + unit_2_11/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_12/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_12/sub2/bn2/beta (32, 32/32 params) + unit_2_12/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_13/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_13/sub2/bn2/beta (32, 32/32 params) + unit_2_13/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_14/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_14/sub2/bn2/beta (32, 32/32 params) + unit_2_14/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_15/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_15/sub2/bn2/beta (32, 32/32 params) + unit_2_15/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_16/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_16/sub2/bn2/beta (32, 32/32 params) + unit_2_16/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_17/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_17/sub2/bn2/beta (32, 32/32 params) + unit_2_17/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_18/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_18/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_18/sub2/bn2/beta (32, 32/32 params) + unit_2_18/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_19/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_19/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_19/sub2/bn2/beta (32, 32/32 params) + unit_2_19/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_20/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_20/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_20/sub2/bn2/beta (32, 32/32 params) + unit_2_20/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_21/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_21/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_21/sub2/bn2/beta (32, 32/32 params) + unit_2_21/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_22/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_22/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_22/sub2/bn2/beta (32, 32/32 params) + unit_2_22/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_23/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_23/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_23/sub2/bn2/beta (32, 32/32 params) + unit_2_23/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_24/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_24/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_24/sub2/bn2/beta (32, 32/32 params) + unit_2_24/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_25/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_25/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_25/sub2/bn2/beta (32, 32/32 params) + unit_2_25/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_26/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_26/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_26/sub2/bn2/beta (32, 32/32 params) + unit_2_26/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_3/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_4/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_5/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_6/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_7/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_8/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_9/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_9/sub2/bn2/beta (32, 32/32 params) + unit_2_9/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_10/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_10/sub2/bn2/beta (64, 64/64 params) + unit_3_10/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_11/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_11/sub2/bn2/beta (64, 64/64 params) + unit_3_11/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_12/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_12/sub2/bn2/beta (64, 64/64 params) + unit_3_12/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_13/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_13/sub2/bn2/beta (64, 64/64 params) + unit_3_13/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_14/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_14/sub2/bn2/beta (64, 64/64 params) + unit_3_14/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_15/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_15/sub2/bn2/beta (64, 64/64 params) + unit_3_15/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_16/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_16/sub2/bn2/beta (64, 64/64 params) + unit_3_16/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_17/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_17/sub2/bn2/beta (64, 64/64 params) + unit_3_17/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_18/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_18/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_18/sub2/bn2/beta (64, 64/64 params) + unit_3_18/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_19/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_19/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_19/sub2/bn2/beta (64, 64/64 params) + unit_3_19/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_20/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_20/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_20/sub2/bn2/beta (64, 64/64 params) + unit_3_20/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_21/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_21/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_21/sub2/bn2/beta (64, 64/64 params) + unit_3_21/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_22/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_22/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_22/sub2/bn2/beta (64, 64/64 params) + unit_3_22/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_23/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_23/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_23/sub2/bn2/beta (64, 64/64 params) + unit_3_23/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_24/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_24/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_24/sub2/bn2/beta (64, 64/64 params) + unit_3_24/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_25/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_25/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_25/sub2/bn2/beta (64, 64/64 params) + unit_3_25/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_26/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_26/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_26/sub2/bn2/beta (64, 64/64 params) + unit_3_26/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_3/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_4/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_5/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_6/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_7/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_8/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_9/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_9/sub2/bn2/beta (64, 64/64 params) + unit_3_9/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/97.35b flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_22/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_20/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_20/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_21/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_21/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_22/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_23/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_23/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_24/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_24/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_25/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_25/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_26/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_26/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_15/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_10/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_11/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_11/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_12/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_12/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_13/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_13/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_14/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_14/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_15/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_16/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_16/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_17/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_17/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_18/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_18/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_19/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_19/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_9/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_9/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_10/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + +======================End of Report========================== +2017-08-03 01:27:16.484756: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla P100-PCIE-16GB +major: 6 minor: 0 memoryClockRate (GHz) 1.3285 +pciBusID 0000:05:00.0 +Total memory: 15.89GiB +Free memory: 15.61GiB +2017-08-03 01:27:16.484829: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-08-03 01:27:16.484840: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-08-03 01:27:16.484853: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:05:00.0) +2017-08-03 01:27:18.115127: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-08-03 01:27:18.115215: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 56 visible devices +2017-08-03 01:27:18.136389: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x8c0c4b0 executing computations on platform Host. Devices: +2017-08-03 01:27:18.136422: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-08-03 01:27:18.136668: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-08-03 01:27:18.136683: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 56 visible devices +2017-08-03 01:27:18.153789: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x7641270 executing computations on platform CUDA. Devices: +2017-08-03 01:27:18.153871: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla P100-PCIE-16GB, Compute Capability 6.0 +2017-08-03 01:27:19.407133: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:247] PoolAllocator: After 1146 get requests, put_count=1100 evicted_count=1000 eviction_rate=0.909091 and unsatisfied allocation rate=1 +2017-08-03 01:27:19.407227: I tensorflow/core/common_runtime/gpu/pool_allocator.cc:259] Raising pool_size_limit_ from 100 to 110 +INFO:tensorflow:step = 1, loss = 5.42836, precision = 0.078125 +INFO:tensorflow:global_step/sec: 4.80302 +INFO:tensorflow:step = 101, loss = 4.74395, precision = 0.320312 (20.822 sec) +INFO:tensorflow:global_step/sec: 5.12673 +INFO:tensorflow:step = 201, loss = 4.46027, precision = 0.476562 (19.505 sec) +INFO:tensorflow:global_step/sec: 5.13927 +INFO:tensorflow:step = 301, loss = 4.30192, precision = 0.453125 (19.458 sec) +total_params: 2596842 +Saved checkpoint after 1 epoch(s) to data/resnet164_nb/checkpoints/00001... +INFO:tensorflow:global_step/sec: 4.37999 +INFO:tensorflow:step = 401, loss = 4.79669, precision = 0.257812 (22.831 sec) +INFO:tensorflow:global_step/sec: 5.13243 +INFO:tensorflow:step = 501, loss = 3.88147, precision = 0.5625 (19.484 sec) +INFO:tensorflow:global_step/sec: 5.11953 +INFO:tensorflow:step = 601, loss = 3.55445, precision = 0.601562 (19.533 sec) +INFO:tensorflow:global_step/sec: 5.11708 +INFO:tensorflow:step = 701, loss = 3.34082, precision = 0.5625 (19.542 sec) +Saved checkpoint after 2 epoch(s) to data/resnet164_nb/checkpoints/00002... +INFO:tensorflow:global_step/sec: 4.49414 +INFO:tensorflow:step = 801, loss = 2.91996, precision = 0.6875 (22.253 sec) +INFO:tensorflow:global_step/sec: 5.1125 +INFO:tensorflow:step = 901, loss = 2.67938, precision = 0.71875 (19.559 sec) +INFO:tensorflow:global_step/sec: 5.11435 +INFO:tensorflow:step = 1001, loss = 2.51412, precision = 0.640625 (19.553 sec) +INFO:tensorflow:global_step/sec: 5.12473 +INFO:tensorflow:step = 1101, loss = 2.42825, precision = 0.710938 (19.513 sec) +Saved checkpoint after 3 epoch(s) to data/resnet164_nb/checkpoints/00003... +INFO:tensorflow:global_step/sec: 4.54779 +INFO:tensorflow:step = 1201, loss = 2.13342, precision = 0.828125 (21.989 sec) +INFO:tensorflow:global_step/sec: 5.12818 +INFO:tensorflow:step = 1301, loss = 2.12535, precision = 0.71875 (19.500 sec) +INFO:tensorflow:global_step/sec: 5.12159 +INFO:tensorflow:step = 1401, loss = 2.02713, precision = 0.71875 (19.525 sec) +INFO:tensorflow:global_step/sec: 5.12321 +INFO:tensorflow:step = 1501, loss = 1.90736, precision = 0.742188 (19.519 sec) +Saved checkpoint after 4 epoch(s) to data/resnet164_nb/checkpoints/00004... +INFO:tensorflow:global_step/sec: 4.55446 +INFO:tensorflow:step = 1601, loss = 1.7789, precision = 0.765625 (21.956 sec) +INFO:tensorflow:global_step/sec: 5.12968 +INFO:tensorflow:step = 1701, loss = 1.68754, precision = 0.75 (19.494 sec) +INFO:tensorflow:global_step/sec: 5.1311 +INFO:tensorflow:step = 1801, loss = 1.63549, precision = 0.742188 (19.489 sec) +INFO:tensorflow:global_step/sec: 5.12001 +INFO:tensorflow:step = 1901, loss = 1.40213, precision = 0.820312 (19.531 sec) +Saved checkpoint after 5 epoch(s) to data/resnet164_nb/checkpoints/00005... +INFO:tensorflow:global_step/sec: 4.55985 +INFO:tensorflow:step = 2001, loss = 1.37763, precision = 0.78125 (21.930 sec) +INFO:tensorflow:global_step/sec: 5.12592 +INFO:tensorflow:step = 2101, loss = 1.38882, precision = 0.734375 (19.511 sec) +INFO:tensorflow:global_step/sec: 5.11793 +INFO:tensorflow:step = 2201, loss = 1.31692, precision = 0.773438 (19.537 sec) +INFO:tensorflow:global_step/sec: 5.12125 +INFO:tensorflow:step = 2301, loss = 1.26583, precision = 0.804688 (19.527 sec) +Saved checkpoint after 6 epoch(s) to data/resnet164_nb/checkpoints/00006... +INFO:tensorflow:global_step/sec: 4.52585 +INFO:tensorflow:step = 2401, loss = 1.146, precision = 0.8125 (22.095 sec) +INFO:tensorflow:global_step/sec: 5.12719 +INFO:tensorflow:step = 2501, loss = 1.26895, precision = 0.765625 (19.504 sec) +INFO:tensorflow:global_step/sec: 5.11749 +INFO:tensorflow:step = 2601, loss = 1.15217, precision = 0.804688 (19.546 sec) +INFO:tensorflow:global_step/sec: 5.12414 +INFO:tensorflow:step = 2701, loss = 1.08326, precision = 0.789062 (19.510 sec) +Saved checkpoint after 7 epoch(s) to data/resnet164_nb/checkpoints/00007... +INFO:tensorflow:global_step/sec: 4.53201 +INFO:tensorflow:step = 2801, loss = 0.982015, precision = 0.835938 (22.066 sec) +INFO:tensorflow:global_step/sec: 5.13069 +INFO:tensorflow:step = 2901, loss = 0.998714, precision = 0.804688 (19.489 sec) +INFO:tensorflow:global_step/sec: 5.13118 +INFO:tensorflow:step = 3001, loss = 1.01195, precision = 0.828125 (19.489 sec) +INFO:tensorflow:global_step/sec: 5.13706 +INFO:tensorflow:step = 3101, loss = 1.07809, precision = 0.773438 (19.466 sec) +Saved checkpoint after 8 epoch(s) to data/resnet164_nb/checkpoints/00008... +INFO:tensorflow:global_step/sec: 4.5865 +INFO:tensorflow:step = 3201, loss = 0.919807, precision = 0.8125 (21.803 sec) +INFO:tensorflow:global_step/sec: 5.13553 +INFO:tensorflow:step = 3301, loss = 1.21141, precision = 0.734375 (19.472 sec) +INFO:tensorflow:global_step/sec: 5.1296 +INFO:tensorflow:step = 3401, loss = 0.97525, precision = 0.820312 (19.494 sec) +INFO:tensorflow:global_step/sec: 5.1253 +INFO:tensorflow:step = 3501, loss = 0.748372, precision = 0.90625 (19.511 sec) +Saved checkpoint after 9 epoch(s) to data/resnet164_nb/checkpoints/00009... +INFO:tensorflow:global_step/sec: 4.47395 +INFO:tensorflow:step = 3601, loss = 0.888899, precision = 0.835938 (22.355 sec) +INFO:tensorflow:global_step/sec: 5.13971 +INFO:tensorflow:step = 3701, loss = 0.858912, precision = 0.84375 (19.454 sec) +INFO:tensorflow:global_step/sec: 5.14197 +INFO:tensorflow:step = 3801, loss = 0.833367, precision = 0.835938 (19.451 sec) +INFO:tensorflow:global_step/sec: 5.13346 +INFO:tensorflow:step = 3901, loss = 0.911276, precision = 0.8125 (19.476 sec) +Saved checkpoint after 10 epoch(s) to data/resnet164_nb/checkpoints/00010... +INFO:tensorflow:global_step/sec: 4.51059 +INFO:tensorflow:step = 4001, loss = 0.854369, precision = 0.851562 (22.170 sec) +INFO:tensorflow:global_step/sec: 5.13997 +INFO:tensorflow:step = 4101, loss = 0.822328, precision = 0.875 (19.455 sec) +INFO:tensorflow:global_step/sec: 5.13603 +INFO:tensorflow:step = 4201, loss = 0.796101, precision = 0.820312 (19.470 sec) +Saved checkpoint after 11 epoch(s) to data/resnet164_nb/checkpoints/00011... +INFO:tensorflow:global_step/sec: 4.54744 +INFO:tensorflow:step = 4301, loss = 0.838089, precision = 0.8125 (21.990 sec) +INFO:tensorflow:global_step/sec: 5.1297 +INFO:tensorflow:step = 4401, loss = 0.791188, precision = 0.8125 (19.495 sec) +INFO:tensorflow:global_step/sec: 5.13859 +INFO:tensorflow:step = 4501, loss = 0.73079, precision = 0.859375 (19.461 sec) +INFO:tensorflow:global_step/sec: 5.15496 +INFO:tensorflow:step = 4601, loss = 0.74776, precision = 0.835938 (19.399 sec) +Saved checkpoint after 12 epoch(s) to data/resnet164_nb/checkpoints/00012... +INFO:tensorflow:global_step/sec: 4.55968 +INFO:tensorflow:step = 4701, loss = 0.773669, precision = 0.84375 (21.931 sec) +INFO:tensorflow:global_step/sec: 5.13474 +INFO:tensorflow:step = 4801, loss = 0.661209, precision = 0.882812 (19.475 sec) +INFO:tensorflow:global_step/sec: 5.13137 +INFO:tensorflow:step = 4901, loss = 1.03065, precision = 0.75 (19.489 sec) +INFO:tensorflow:global_step/sec: 5.13443 +INFO:tensorflow:step = 5001, loss = 0.720849, precision = 0.890625 (19.475 sec) +Saved checkpoint after 13 epoch(s) to data/resnet164_nb/checkpoints/00013... +INFO:tensorflow:global_step/sec: 4.56782 +INFO:tensorflow:step = 5101, loss = 0.709803, precision = 0.835938 (21.894 sec) +INFO:tensorflow:global_step/sec: 5.12774 +INFO:tensorflow:step = 5201, loss = 0.893288, precision = 0.773438 (19.500 sec) +INFO:tensorflow:global_step/sec: 5.13415 +INFO:tensorflow:step = 5301, loss = 0.89816, precision = 0.804688 (19.477 sec) +INFO:tensorflow:global_step/sec: 5.13865 +INFO:tensorflow:step = 5401, loss = 0.800971, precision = 0.820312 (19.460 sec) +Saved checkpoint after 14 epoch(s) to data/resnet164_nb/checkpoints/00014... +INFO:tensorflow:global_step/sec: 4.56986 +INFO:tensorflow:step = 5501, loss = 0.841643, precision = 0.84375 (21.883 sec) +INFO:tensorflow:global_step/sec: 5.13247 +INFO:tensorflow:step = 5601, loss = 0.694691, precision = 0.851562 (19.484 sec) +INFO:tensorflow:global_step/sec: 5.13304 +INFO:tensorflow:step = 5701, loss = 0.675151, precision = 0.898438 (19.482 sec) +INFO:tensorflow:global_step/sec: 5.13106 +INFO:tensorflow:step = 5801, loss = 0.648192, precision = 0.867188 (19.489 sec) +Saved checkpoint after 15 epoch(s) to data/resnet164_nb/checkpoints/00015... +INFO:tensorflow:global_step/sec: 4.56756 +INFO:tensorflow:step = 5901, loss = 0.720236, precision = 0.851562 (21.894 sec) +INFO:tensorflow:global_step/sec: 5.13081 +INFO:tensorflow:step = 6001, loss = 0.659713, precision = 0.867188 (19.490 sec) +INFO:tensorflow:global_step/sec: 5.12588 +INFO:tensorflow:step = 6101, loss = 0.729953, precision = 0.84375 (19.509 sec) +INFO:tensorflow:global_step/sec: 5.12217 +INFO:tensorflow:step = 6201, loss = 0.869042, precision = 0.773438 (19.523 sec) +Saved checkpoint after 16 epoch(s) to data/resnet164_nb/checkpoints/00016... +INFO:tensorflow:global_step/sec: 4.58066 +INFO:tensorflow:step = 6301, loss = 0.732346, precision = 0.828125 (21.831 sec) +INFO:tensorflow:global_step/sec: 5.1245 +INFO:tensorflow:step = 6401, loss = 0.754362, precision = 0.851562 (19.514 sec) +INFO:tensorflow:global_step/sec: 5.13008 +INFO:tensorflow:step = 6501, loss = 0.731841, precision = 0.84375 (19.493 sec) +INFO:tensorflow:global_step/sec: 5.13243 +INFO:tensorflow:step = 6601, loss = 0.906044, precision = 0.820312 (19.484 sec) +Saved checkpoint after 17 epoch(s) to data/resnet164_nb/checkpoints/00017... +INFO:tensorflow:global_step/sec: 4.50302 +INFO:tensorflow:step = 6701, loss = 0.660759, precision = 0.898438 (22.207 sec) +INFO:tensorflow:global_step/sec: 5.12001 +INFO:tensorflow:step = 6801, loss = 0.893288, precision = 0.773438 (19.531 sec) +INFO:tensorflow:global_step/sec: 5.12488 +INFO:tensorflow:step = 6901, loss = 0.707547, precision = 0.820312 (19.513 sec) +INFO:tensorflow:global_step/sec: 5.14267 +INFO:tensorflow:step = 7001, loss = 0.690679, precision = 0.859375 (19.450 sec) +Saved checkpoint after 18 epoch(s) to data/resnet164_nb/checkpoints/00018... +INFO:tensorflow:global_step/sec: 4.49701 +INFO:tensorflow:step = 7101, loss = 0.724672, precision = 0.882812 (22.236 sec) +INFO:tensorflow:global_step/sec: 5.1285 +INFO:tensorflow:step = 7201, loss = 0.66055, precision = 0.882812 (19.496 sec) +INFO:tensorflow:global_step/sec: 5.12615 +INFO:tensorflow:step = 7301, loss = 0.744526, precision = 0.84375 (19.508 sec) +INFO:tensorflow:global_step/sec: 5.12973 +INFO:tensorflow:step = 7401, loss = 0.804322, precision = 0.8125 (19.494 sec) +Saved checkpoint after 19 epoch(s) to data/resnet164_nb/checkpoints/00019... +INFO:tensorflow:global_step/sec: 4.55157 +INFO:tensorflow:step = 7501, loss = 0.60533, precision = 0.90625 (21.971 sec) +INFO:tensorflow:global_step/sec: 5.12738 +INFO:tensorflow:step = 7601, loss = 0.744891, precision = 0.84375 (19.502 sec) +INFO:tensorflow:global_step/sec: 5.134 +INFO:tensorflow:step = 7701, loss = 0.795824, precision = 0.835938 (19.478 sec) +INFO:tensorflow:global_step/sec: 5.13054 +INFO:tensorflow:step = 7801, loss = 0.823626, precision = 0.828125 (19.491 sec) +Saved checkpoint after 20 epoch(s) to data/resnet164_nb/checkpoints/00020... +INFO:tensorflow:global_step/sec: 4.56826 +INFO:tensorflow:step = 7901, loss = 0.852509, precision = 0.804688 (21.890 sec) +INFO:tensorflow:global_step/sec: 5.13765 +INFO:tensorflow:step = 8001, loss = 0.769712, precision = 0.851562 (19.464 sec) +INFO:tensorflow:global_step/sec: 5.13434 +INFO:tensorflow:step = 8101, loss = 0.815203, precision = 0.828125 (19.477 sec) +INFO:tensorflow:global_step/sec: 5.13038 +INFO:tensorflow:step = 8201, loss = 0.657938, precision = 0.882812 (19.492 sec) +Saved checkpoint after 21 epoch(s) to data/resnet164_nb/checkpoints/00021... +INFO:tensorflow:global_step/sec: 4.57849 +INFO:tensorflow:step = 8301, loss = 0.921166, precision = 0.789062 (21.842 sec) +INFO:tensorflow:global_step/sec: 5.13844 +INFO:tensorflow:step = 8401, loss = 0.724511, precision = 0.84375 (19.461 sec) +INFO:tensorflow:global_step/sec: 5.13889 +INFO:tensorflow:step = 8501, loss = 0.801763, precision = 0.835938 (19.459 sec) +INFO:tensorflow:global_step/sec: 5.1353 +INFO:tensorflow:step = 8601, loss = 0.642462, precision = 0.84375 (19.475 sec) +Saved checkpoint after 22 epoch(s) to data/resnet164_nb/checkpoints/00022... +INFO:tensorflow:global_step/sec: 4.56619 +INFO:tensorflow:step = 8701, loss = 0.621972, precision = 0.882812 (21.899 sec) +INFO:tensorflow:global_step/sec: 5.14263 +INFO:tensorflow:step = 8801, loss = 0.743347, precision = 0.835938 (19.447 sec) +INFO:tensorflow:global_step/sec: 5.1374 +INFO:tensorflow:step = 8901, loss = 0.619683, precision = 0.890625 (19.463 sec) +Saved checkpoint after 23 epoch(s) to data/resnet164_nb/checkpoints/00023... +INFO:tensorflow:global_step/sec: 4.51112 +INFO:tensorflow:step = 9001, loss = 0.731875, precision = 0.835938 (22.168 sec) +INFO:tensorflow:global_step/sec: 5.12663 +INFO:tensorflow:step = 9101, loss = 0.820192, precision = 0.828125 (19.505 sec) +INFO:tensorflow:global_step/sec: 5.12728 +INFO:tensorflow:step = 9201, loss = 0.80924, precision = 0.828125 (19.504 sec) +INFO:tensorflow:global_step/sec: 5.13041 +INFO:tensorflow:step = 9301, loss = 0.628841, precision = 0.890625 (19.492 sec) +Saved checkpoint after 24 epoch(s) to data/resnet164_nb/checkpoints/00024... +INFO:tensorflow:global_step/sec: 4.54615 +INFO:tensorflow:step = 9401, loss = 0.653239, precision = 0.867188 (21.997 sec) +INFO:tensorflow:global_step/sec: 5.12898 +INFO:tensorflow:step = 9501, loss = 0.597215, precision = 0.890625 (19.496 sec) +INFO:tensorflow:global_step/sec: 5.13391 +INFO:tensorflow:step = 9601, loss = 0.553443, precision = 0.921875 (19.478 sec) +INFO:tensorflow:global_step/sec: 5.13138 +INFO:tensorflow:step = 9701, loss = 0.732953, precision = 0.851562 (19.492 sec) +Saved checkpoint after 25 epoch(s) to data/resnet164_nb/checkpoints/00025... +INFO:tensorflow:global_step/sec: 4.56279 +INFO:tensorflow:step = 9801, loss = 0.56848, precision = 0.898438 (21.913 sec) +INFO:tensorflow:global_step/sec: 5.13773 +INFO:tensorflow:step = 9901, loss = 0.657257, precision = 0.867188 (19.464 sec) +INFO:tensorflow:global_step/sec: 5.1373 +INFO:tensorflow:step = 10001, loss = 0.65789, precision = 0.875 (19.468 sec) +INFO:tensorflow:global_step/sec: 5.14672 +INFO:tensorflow:step = 10101, loss = 0.655613, precision = 0.867188 (19.428 sec) +Saved checkpoint after 26 epoch(s) to data/resnet164_nb/checkpoints/00026... +INFO:tensorflow:global_step/sec: 4.55541 +INFO:tensorflow:step = 10201, loss = 0.657371, precision = 0.882812 (21.952 sec) +INFO:tensorflow:global_step/sec: 5.12589 +INFO:tensorflow:step = 10301, loss = 0.573779, precision = 0.914062 (19.509 sec) +INFO:tensorflow:global_step/sec: 5.1351 +INFO:tensorflow:step = 10401, loss = 0.670814, precision = 0.890625 (19.473 sec) +INFO:tensorflow:global_step/sec: 5.13332 +INFO:tensorflow:step = 10501, loss = 0.769034, precision = 0.851562 (19.482 sec) +Saved checkpoint after 27 epoch(s) to data/resnet164_nb/checkpoints/00027... +INFO:tensorflow:global_step/sec: 4.50665 +INFO:tensorflow:step = 10601, loss = 0.788741, precision = 0.835938 (22.188 sec) +INFO:tensorflow:global_step/sec: 5.13699 +INFO:tensorflow:step = 10701, loss = 0.67038, precision = 0.851562 (19.468 sec) +INFO:tensorflow:global_step/sec: 5.13189 +INFO:tensorflow:step = 10801, loss = 0.577851, precision = 0.90625 (19.485 sec) +INFO:tensorflow:global_step/sec: 5.13193 +INFO:tensorflow:step = 10901, loss = 0.649488, precision = 0.851562 (19.486 sec) +Saved checkpoint after 28 epoch(s) to data/resnet164_nb/checkpoints/00028... +INFO:tensorflow:global_step/sec: 4.56898 +INFO:tensorflow:step = 11001, loss = 0.702948, precision = 0.835938 (21.887 sec) +INFO:tensorflow:global_step/sec: 5.13495 +INFO:tensorflow:step = 11101, loss = 0.647718, precision = 0.898438 (19.474 sec) +INFO:tensorflow:global_step/sec: 5.13832 +INFO:tensorflow:step = 11201, loss = 0.623194, precision = 0.882812 (19.462 sec) +INFO:tensorflow:global_step/sec: 5.12601 +INFO:tensorflow:step = 11301, loss = 0.740656, precision = 0.851562 (19.508 sec) +Saved checkpoint after 29 epoch(s) to data/resnet164_nb/checkpoints/00029... +INFO:tensorflow:global_step/sec: 4.57495 +INFO:tensorflow:step = 11401, loss = 0.791753, precision = 0.8125 (21.858 sec) +INFO:tensorflow:global_step/sec: 5.13487 +INFO:tensorflow:step = 11501, loss = 0.669864, precision = 0.882812 (19.475 sec) +INFO:tensorflow:global_step/sec: 5.13827 +INFO:tensorflow:step = 11601, loss = 0.620103, precision = 0.890625 (19.462 sec) +INFO:tensorflow:global_step/sec: 5.13651 +INFO:tensorflow:step = 11701, loss = 0.674914, precision = 0.875 (19.469 sec) +Saved checkpoint after 30 epoch(s) to data/resnet164_nb/checkpoints/00030... +INFO:tensorflow:global_step/sec: 4.55347 +INFO:tensorflow:step = 11801, loss = 0.654215, precision = 0.851562 (21.961 sec) +INFO:tensorflow:global_step/sec: 5.12921 +INFO:tensorflow:step = 11901, loss = 0.61026, precision = 0.875 (19.496 sec) +INFO:tensorflow:global_step/sec: 5.12959 +INFO:tensorflow:step = 12001, loss = 0.726911, precision = 0.859375 (19.495 sec) +INFO:tensorflow:global_step/sec: 5.1326 +INFO:tensorflow:step = 12101, loss = 0.71643, precision = 0.875 (19.483 sec) +Saved checkpoint after 31 epoch(s) to data/resnet164_nb/checkpoints/00031... +INFO:tensorflow:global_step/sec: 4.5728 +INFO:tensorflow:step = 12201, loss = 0.566891, precision = 0.898438 (21.869 sec) +INFO:tensorflow:global_step/sec: 5.13392 +INFO:tensorflow:step = 12301, loss = 0.687493, precision = 0.851562 (19.478 sec) +INFO:tensorflow:global_step/sec: 5.13574 +INFO:tensorflow:step = 12401, loss = 0.703233, precision = 0.882812 (19.471 sec) +INFO:tensorflow:global_step/sec: 5.12697 +INFO:tensorflow:step = 12501, loss = 0.644073, precision = 0.90625 (19.505 sec) +Saved checkpoint after 32 epoch(s) to data/resnet164_nb/checkpoints/00032... +INFO:tensorflow:global_step/sec: 4.58604 +INFO:tensorflow:step = 12601, loss = 0.7251, precision = 0.867188 (21.805 sec) +INFO:tensorflow:global_step/sec: 5.12934 +INFO:tensorflow:step = 12701, loss = 0.701906, precision = 0.851562 (19.496 sec) +INFO:tensorflow:global_step/sec: 5.1233 +INFO:tensorflow:step = 12801, loss = 0.587264, precision = 0.90625 (19.519 sec) +INFO:tensorflow:global_step/sec: 5.13832 +INFO:tensorflow:step = 12901, loss = 0.681216, precision = 0.882812 (19.462 sec) +Saved checkpoint after 33 epoch(s) to data/resnet164_nb/checkpoints/00033... +INFO:tensorflow:global_step/sec: 4.57064 +INFO:tensorflow:step = 13001, loss = 0.785588, precision = 0.820312 (21.879 sec) +INFO:tensorflow:global_step/sec: 5.12504 +INFO:tensorflow:step = 13101, loss = 0.724269, precision = 0.859375 (19.512 sec) +INFO:tensorflow:global_step/sec: 5.13441 +INFO:tensorflow:step = 13201, loss = 0.752295, precision = 0.859375 (19.477 sec) +Saved checkpoint after 34 epoch(s) to data/resnet164_nb/checkpoints/00034... +INFO:tensorflow:global_step/sec: 4.49727 +INFO:tensorflow:step = 13301, loss = 0.62987, precision = 0.875 (22.236 sec) +INFO:tensorflow:global_step/sec: 5.13019 +INFO:tensorflow:step = 13401, loss = 0.692694, precision = 0.890625 (19.492 sec) +INFO:tensorflow:global_step/sec: 5.1321 +INFO:tensorflow:step = 13501, loss = 0.681913, precision = 0.882812 (19.487 sec) +INFO:tensorflow:global_step/sec: 5.13738 +INFO:tensorflow:step = 13601, loss = 0.694832, precision = 0.882812 (19.463 sec) +Saved checkpoint after 35 epoch(s) to data/resnet164_nb/checkpoints/00035... +INFO:tensorflow:global_step/sec: 4.56872 +INFO:tensorflow:step = 13701, loss = 0.631694, precision = 0.890625 (21.888 sec) +INFO:tensorflow:global_step/sec: 5.12784 +INFO:tensorflow:step = 13801, loss = 0.510267, precision = 0.9375 (19.502 sec) +INFO:tensorflow:global_step/sec: 5.13087 +INFO:tensorflow:step = 13901, loss = 0.892432, precision = 0.773438 (19.492 sec) +INFO:tensorflow:global_step/sec: 5.13414 +INFO:tensorflow:step = 14001, loss = 0.791691, precision = 0.796875 (19.475 sec) +Saved checkpoint after 36 epoch(s) to data/resnet164_nb/checkpoints/00036... +INFO:tensorflow:global_step/sec: 4.55053 +INFO:tensorflow:step = 14101, loss = 0.611401, precision = 0.898438 (21.976 sec) +INFO:tensorflow:global_step/sec: 5.13438 +INFO:tensorflow:step = 14201, loss = 0.573122, precision = 0.914062 (19.476 sec) +INFO:tensorflow:global_step/sec: 5.14596 +INFO:tensorflow:step = 14301, loss = 0.610775, precision = 0.882812 (19.433 sec) +INFO:tensorflow:global_step/sec: 5.13146 +INFO:tensorflow:step = 14401, loss = 0.657121, precision = 0.890625 (19.488 sec) +Saved checkpoint after 37 epoch(s) to data/resnet164_nb/checkpoints/00037... +INFO:tensorflow:global_step/sec: 4.53578 +INFO:tensorflow:step = 14501, loss = 0.580294, precision = 0.90625 (22.047 sec) +INFO:tensorflow:global_step/sec: 5.1373 +INFO:tensorflow:step = 14601, loss = 0.665138, precision = 0.859375 (19.465 sec) +INFO:tensorflow:global_step/sec: 5.12555 +INFO:tensorflow:step = 14701, loss = 0.708496, precision = 0.875 (19.510 sec) +INFO:tensorflow:global_step/sec: 5.12879 +INFO:tensorflow:step = 14801, loss = 0.629958, precision = 0.90625 (19.501 sec) +Saved checkpoint after 38 epoch(s) to data/resnet164_nb/checkpoints/00038... +INFO:tensorflow:global_step/sec: 4.43716 +INFO:tensorflow:step = 14901, loss = 0.515097, precision = 0.921875 (22.534 sec) +INFO:tensorflow:global_step/sec: 5.13331 +INFO:tensorflow:step = 15001, loss = 0.65199, precision = 0.882812 (19.480 sec) +INFO:tensorflow:global_step/sec: 5.12639 +INFO:tensorflow:step = 15101, loss = 0.720808, precision = 0.835938 (19.511 sec) +INFO:tensorflow:global_step/sec: 5.13162 +INFO:tensorflow:step = 15201, loss = 0.680973, precision = 0.84375 (19.483 sec) +Saved checkpoint after 39 epoch(s) to data/resnet164_nb/checkpoints/00039... +INFO:tensorflow:global_step/sec: 4.53927 +INFO:tensorflow:step = 15301, loss = 0.586867, precision = 0.914062 (22.032 sec) +INFO:tensorflow:global_step/sec: 5.13028 +INFO:tensorflow:step = 15401, loss = 0.685704, precision = 0.859375 (19.491 sec) +INFO:tensorflow:global_step/sec: 5.1299 +INFO:tensorflow:step = 15501, loss = 0.632224, precision = 0.898438 (19.494 sec) +INFO:tensorflow:global_step/sec: 5.13263 +INFO:tensorflow:step = 15601, loss = 0.599667, precision = 0.875 (19.482 sec) +Saved checkpoint after 40 epoch(s) to data/resnet164_nb/checkpoints/00040... +INFO:tensorflow:global_step/sec: 4.57141 +INFO:tensorflow:step = 15701, loss = 0.651918, precision = 0.882812 (21.875 sec) +INFO:tensorflow:global_step/sec: 5.13741 +INFO:tensorflow:step = 15801, loss = 0.694573, precision = 0.867188 (19.465 sec) +INFO:tensorflow:global_step/sec: 5.13127 +INFO:tensorflow:step = 15901, loss = 0.805515, precision = 0.8125 (19.489 sec) +INFO:tensorflow:global_step/sec: 5.13568 +INFO:tensorflow:step = 16001, loss = 0.66171, precision = 0.898438 (19.476 sec) +Saved checkpoint after 41 epoch(s) to data/resnet164_nb/checkpoints/00041... +INFO:tensorflow:global_step/sec: 4.57403 +INFO:tensorflow:step = 16101, loss = 0.75762, precision = 0.804688 (21.859 sec) +INFO:tensorflow:global_step/sec: 5.12918 +INFO:tensorflow:step = 16201, loss = 0.674681, precision = 0.84375 (19.496 sec) +INFO:tensorflow:global_step/sec: 5.13951 +INFO:tensorflow:step = 16301, loss = 0.677021, precision = 0.890625 (19.457 sec) +INFO:tensorflow:global_step/sec: 5.13427 +INFO:tensorflow:step = 16401, loss = 0.688102, precision = 0.835938 (19.477 sec) +Saved checkpoint after 42 epoch(s) to data/resnet164_nb/checkpoints/00042... +INFO:tensorflow:global_step/sec: 4.55775 +INFO:tensorflow:step = 16501, loss = 0.708009, precision = 0.867188 (21.941 sec) +INFO:tensorflow:global_step/sec: 5.13349 +INFO:tensorflow:step = 16601, loss = 0.692224, precision = 0.859375 (19.480 sec) +INFO:tensorflow:global_step/sec: 5.13535 +INFO:tensorflow:step = 16701, loss = 0.602669, precision = 0.921875 (19.473 sec) +INFO:tensorflow:global_step/sec: 5.13522 +INFO:tensorflow:step = 16801, loss = 0.681883, precision = 0.867188 (19.475 sec) +Saved checkpoint after 43 epoch(s) to data/resnet164_nb/checkpoints/00043... +INFO:tensorflow:global_step/sec: 4.58761 +INFO:tensorflow:step = 16901, loss = 0.733053, precision = 0.851562 (21.796 sec) +INFO:tensorflow:global_step/sec: 5.13699 +INFO:tensorflow:step = 17001, loss = 0.613993, precision = 0.898438 (19.467 sec) +INFO:tensorflow:global_step/sec: 5.13534 +INFO:tensorflow:step = 17101, loss = 0.61749, precision = 0.890625 (19.472 sec) +INFO:tensorflow:global_step/sec: 5.14155 +INFO:tensorflow:step = 17201, loss = 0.789378, precision = 0.84375 (19.449 sec) +Saved checkpoint after 44 epoch(s) to data/resnet164_nb/checkpoints/00044... +INFO:tensorflow:global_step/sec: 4.58431 +INFO:tensorflow:step = 17301, loss = 0.68753, precision = 0.882812 (21.818 sec) +INFO:tensorflow:global_step/sec: 5.13785 +INFO:tensorflow:step = 17401, loss = 0.722305, precision = 0.859375 (19.459 sec) +INFO:tensorflow:global_step/sec: 5.13421 +INFO:tensorflow:step = 17501, loss = 0.646154, precision = 0.890625 (19.478 sec) +Saved checkpoint after 45 epoch(s) to data/resnet164_nb/checkpoints/00045... +INFO:tensorflow:global_step/sec: 4.50421 +INFO:tensorflow:step = 17601, loss = 0.733671, precision = 0.835938 (22.200 sec) +INFO:tensorflow:global_step/sec: 5.1298 +INFO:tensorflow:step = 17701, loss = 0.5334, precision = 0.929688 (19.494 sec) +INFO:tensorflow:global_step/sec: 5.14219 +INFO:tensorflow:step = 17801, loss = 0.532953, precision = 0.929688 (19.447 sec) +INFO:tensorflow:global_step/sec: 5.14498 +INFO:tensorflow:step = 17901, loss = 0.679994, precision = 0.835938 (19.436 sec) +Saved checkpoint after 46 epoch(s) to data/resnet164_nb/checkpoints/00046... +INFO:tensorflow:global_step/sec: 4.49932 +INFO:tensorflow:step = 18001, loss = 0.720933, precision = 0.875 (22.226 sec) +INFO:tensorflow:global_step/sec: 5.13923 +INFO:tensorflow:step = 18101, loss = 0.62946, precision = 0.898438 (19.458 sec) +INFO:tensorflow:global_step/sec: 5.13789 +INFO:tensorflow:step = 18201, loss = 0.672176, precision = 0.898438 (19.468 sec) +INFO:tensorflow:global_step/sec: 5.13826 +INFO:tensorflow:step = 18301, loss = 0.760581, precision = 0.859375 (19.458 sec) +Saved checkpoint after 47 epoch(s) to data/resnet164_nb/checkpoints/00047... +INFO:tensorflow:global_step/sec: 4.53164 +INFO:tensorflow:step = 18401, loss = 0.667478, precision = 0.890625 (22.068 sec) +INFO:tensorflow:global_step/sec: 5.13237 +INFO:tensorflow:step = 18501, loss = 0.708469, precision = 0.84375 (19.486 sec) +INFO:tensorflow:global_step/sec: 5.13487 +INFO:tensorflow:step = 18601, loss = 0.610605, precision = 0.890625 (19.471 sec) +INFO:tensorflow:global_step/sec: 5.13525 +INFO:tensorflow:step = 18701, loss = 0.672294, precision = 0.90625 (19.473 sec) +Saved checkpoint after 48 epoch(s) to data/resnet164_nb/checkpoints/00048... +INFO:tensorflow:global_step/sec: 4.49731 +INFO:tensorflow:step = 18801, loss = 0.67159, precision = 0.890625 (22.236 sec) +INFO:tensorflow:global_step/sec: 5.13554 +INFO:tensorflow:step = 18901, loss = 0.638224, precision = 0.890625 (19.475 sec) +INFO:tensorflow:global_step/sec: 5.12609 +INFO:tensorflow:step = 19001, loss = 0.834279, precision = 0.828125 (19.505 sec) +INFO:tensorflow:global_step/sec: 5.14006 +INFO:tensorflow:step = 19101, loss = 0.664971, precision = 0.835938 (19.456 sec) +Saved checkpoint after 49 epoch(s) to data/resnet164_nb/checkpoints/00049... +INFO:tensorflow:global_step/sec: 4.57778 +INFO:tensorflow:step = 19201, loss = 0.678645, precision = 0.867188 (21.843 sec) +INFO:tensorflow:global_step/sec: 5.13874 +INFO:tensorflow:step = 19301, loss = 0.57741, precision = 0.914062 (19.461 sec) +INFO:tensorflow:global_step/sec: 5.13086 +INFO:tensorflow:step = 19401, loss = 0.73039, precision = 0.867188 (19.490 sec) +INFO:tensorflow:global_step/sec: 5.13638 +INFO:tensorflow:step = 19501, loss = 0.626255, precision = 0.898438 (19.473 sec) +Saved checkpoint after 50 epoch(s) to data/resnet164_nb/checkpoints/00050... +INFO:tensorflow:global_step/sec: 4.57141 +INFO:tensorflow:step = 19601, loss = 0.678141, precision = 0.851562 (21.871 sec) +INFO:tensorflow:global_step/sec: 5.13968 +INFO:tensorflow:step = 19701, loss = 0.540546, precision = 0.9375 (19.457 sec) +INFO:tensorflow:global_step/sec: 5.14042 +INFO:tensorflow:step = 19801, loss = 0.598814, precision = 0.921875 (19.454 sec) +INFO:tensorflow:global_step/sec: 5.13237 +INFO:tensorflow:step = 19901, loss = 0.702539, precision = 0.851562 (19.484 sec) +Saved checkpoint after 51 epoch(s) to data/resnet164_nb/checkpoints/00051... +INFO:tensorflow:global_step/sec: 4.54224 +INFO:tensorflow:step = 20001, loss = 0.67585, precision = 0.90625 (22.019 sec) +INFO:tensorflow:global_step/sec: 5.13615 +INFO:tensorflow:step = 20101, loss = 0.642693, precision = 0.875 (19.467 sec) +INFO:tensorflow:global_step/sec: 5.14267 +INFO:tensorflow:step = 20201, loss = 0.637615, precision = 0.898438 (19.445 sec) +INFO:tensorflow:global_step/sec: 5.13533 +INFO:tensorflow:step = 20301, loss = 0.586864, precision = 0.914062 (19.474 sec) +Saved checkpoint after 52 epoch(s) to data/resnet164_nb/checkpoints/00052... +INFO:tensorflow:global_step/sec: 4.54953 +INFO:tensorflow:step = 20401, loss = 0.619194, precision = 0.898438 (21.980 sec) +INFO:tensorflow:global_step/sec: 5.14718 +INFO:tensorflow:step = 20501, loss = 0.5293, precision = 0.929688 (19.428 sec) +INFO:tensorflow:global_step/sec: 5.13916 +INFO:tensorflow:step = 20601, loss = 0.650556, precision = 0.875 (19.458 sec) +INFO:tensorflow:global_step/sec: 5.1336 +INFO:tensorflow:step = 20701, loss = 0.673283, precision = 0.875 (19.480 sec) +Saved checkpoint after 53 epoch(s) to data/resnet164_nb/checkpoints/00053... +INFO:tensorflow:global_step/sec: 4.50233 +INFO:tensorflow:step = 20801, loss = 0.746603, precision = 0.851562 (22.211 sec) +INFO:tensorflow:global_step/sec: 5.12908 +INFO:tensorflow:step = 20901, loss = 0.647183, precision = 0.890625 (19.497 sec) +INFO:tensorflow:global_step/sec: 5.13079 +INFO:tensorflow:step = 21001, loss = 0.706764, precision = 0.882812 (19.494 sec) +INFO:tensorflow:global_step/sec: 5.13442 +INFO:tensorflow:step = 21101, loss = 0.678026, precision = 0.882812 (19.473 sec) +Saved checkpoint after 54 epoch(s) to data/resnet164_nb/checkpoints/00054... +INFO:tensorflow:global_step/sec: 4.56619 +INFO:tensorflow:step = 21201, loss = 0.621994, precision = 0.90625 (21.902 sec) +INFO:tensorflow:global_step/sec: 5.13844 +INFO:tensorflow:step = 21301, loss = 0.772093, precision = 0.835938 (19.459 sec) +INFO:tensorflow:global_step/sec: 5.13918 +INFO:tensorflow:step = 21401, loss = 0.636204, precision = 0.875 (19.458 sec) +INFO:tensorflow:global_step/sec: 5.13411 +INFO:tensorflow:step = 21501, loss = 0.642948, precision = 0.875 (19.478 sec) +Saved checkpoint after 55 epoch(s) to data/resnet164_nb/checkpoints/00055... +INFO:tensorflow:global_step/sec: 4.58165 +INFO:tensorflow:step = 21601, loss = 0.710044, precision = 0.882812 (21.829 sec) +INFO:tensorflow:global_step/sec: 5.13383 +INFO:tensorflow:step = 21701, loss = 0.863743, precision = 0.84375 (19.477 sec) +INFO:tensorflow:global_step/sec: 5.122 +INFO:tensorflow:step = 21801, loss = 0.757759, precision = 0.859375 (19.523 sec) +Saved checkpoint after 56 epoch(s) to data/resnet164_nb/checkpoints/00056... +INFO:tensorflow:global_step/sec: 4.50831 +INFO:tensorflow:step = 21901, loss = 0.60536, precision = 0.890625 (22.181 sec) +INFO:tensorflow:global_step/sec: 5.14284 +INFO:tensorflow:step = 22001, loss = 0.760857, precision = 0.84375 (19.445 sec) +INFO:tensorflow:global_step/sec: 5.14026 +INFO:tensorflow:step = 22101, loss = 0.64732, precision = 0.890625 (19.454 sec) +INFO:tensorflow:global_step/sec: 5.14813 +INFO:tensorflow:step = 22201, loss = 0.629414, precision = 0.898438 (19.425 sec) +Saved checkpoint after 57 epoch(s) to data/resnet164_nb/checkpoints/00057... +INFO:tensorflow:global_step/sec: 4.56262 +INFO:tensorflow:step = 22301, loss = 0.507122, precision = 0.945312 (21.917 sec) +INFO:tensorflow:global_step/sec: 5.13773 +INFO:tensorflow:step = 22401, loss = 0.575548, precision = 0.90625 (19.464 sec) +INFO:tensorflow:global_step/sec: 5.1424 +INFO:tensorflow:step = 22501, loss = 0.640427, precision = 0.921875 (19.446 sec) +INFO:tensorflow:global_step/sec: 5.14 +INFO:tensorflow:step = 22601, loss = 0.665333, precision = 0.921875 (19.457 sec) +Saved checkpoint after 58 epoch(s) to data/resnet164_nb/checkpoints/00058... +INFO:tensorflow:global_step/sec: 4.55032 +INFO:tensorflow:step = 22701, loss = 0.654365, precision = 0.882812 (21.976 sec) +INFO:tensorflow:global_step/sec: 5.14458 +INFO:tensorflow:step = 22801, loss = 0.692104, precision = 0.882812 (19.440 sec) +INFO:tensorflow:global_step/sec: 5.15018 +INFO:tensorflow:step = 22901, loss = 0.674659, precision = 0.859375 (19.414 sec) +INFO:tensorflow:global_step/sec: 5.15774 +INFO:tensorflow:step = 23001, loss = 0.699489, precision = 0.859375 (19.388 sec) +Saved checkpoint after 59 epoch(s) to data/resnet164_nb/checkpoints/00059... +INFO:tensorflow:global_step/sec: 4.59123 +INFO:tensorflow:step = 23101, loss = 0.599151, precision = 0.921875 (21.781 sec) +INFO:tensorflow:global_step/sec: 5.14143 +INFO:tensorflow:step = 23201, loss = 0.6344, precision = 0.875 (19.449 sec) +INFO:tensorflow:global_step/sec: 5.13046 +INFO:tensorflow:step = 23301, loss = 0.589787, precision = 0.9375 (19.494 sec) +INFO:tensorflow:global_step/sec: 5.13246 +INFO:tensorflow:step = 23401, loss = 0.646022, precision = 0.882812 (19.483 sec) +Saved checkpoint after 60 epoch(s) to data/resnet164_nb/checkpoints/00060... +INFO:tensorflow:global_step/sec: 4.53289 +INFO:tensorflow:step = 23501, loss = 0.670479, precision = 0.84375 (22.060 sec) +INFO:tensorflow:global_step/sec: 5.12871 +INFO:tensorflow:step = 23601, loss = 0.576521, precision = 0.929688 (19.501 sec) +INFO:tensorflow:global_step/sec: 5.12952 +INFO:tensorflow:step = 23701, loss = 0.767157, precision = 0.867188 (19.492 sec) +INFO:tensorflow:global_step/sec: 5.13646 +INFO:tensorflow:step = 23801, loss = 0.802574, precision = 0.835938 (19.469 sec) +Saved checkpoint after 61 epoch(s) to data/resnet164_nb/checkpoints/00061... +INFO:tensorflow:global_step/sec: 4.59234 +INFO:tensorflow:step = 23901, loss = 0.602564, precision = 0.914062 (21.776 sec) +INFO:tensorflow:global_step/sec: 5.13216 +INFO:tensorflow:step = 24001, loss = 0.703323, precision = 0.867188 (19.491 sec) +INFO:tensorflow:global_step/sec: 5.13435 +INFO:tensorflow:step = 24101, loss = 0.548478, precision = 0.929688 (19.471 sec) +INFO:tensorflow:global_step/sec: 5.13919 +INFO:tensorflow:step = 24201, loss = 0.745466, precision = 0.882812 (19.458 sec) +Saved checkpoint after 62 epoch(s) to data/resnet164_nb/checkpoints/00062... +INFO:tensorflow:global_step/sec: 4.57139 +INFO:tensorflow:step = 24301, loss = 0.637198, precision = 0.882812 (21.875 sec) +INFO:tensorflow:global_step/sec: 5.13911 +INFO:tensorflow:step = 24401, loss = 0.617806, precision = 0.882812 (19.459 sec) +INFO:tensorflow:global_step/sec: 5.14495 +INFO:tensorflow:step = 24501, loss = 0.661139, precision = 0.882812 (19.442 sec) +INFO:tensorflow:global_step/sec: 5.14199 +INFO:tensorflow:step = 24601, loss = 0.574448, precision = 0.9375 (19.443 sec) +Saved checkpoint after 63 epoch(s) to data/resnet164_nb/checkpoints/00063... +INFO:tensorflow:global_step/sec: 4.5909 +INFO:tensorflow:step = 24701, loss = 0.535309, precision = 0.929688 (21.787 sec) +INFO:tensorflow:global_step/sec: 5.13401 +INFO:tensorflow:step = 24801, loss = 0.779341, precision = 0.828125 (19.474 sec) +INFO:tensorflow:global_step/sec: 5.1417 +INFO:tensorflow:step = 24901, loss = 0.63561, precision = 0.898438 (19.449 sec) +INFO:tensorflow:global_step/sec: 5.13533 +INFO:tensorflow:step = 25001, loss = 0.531689, precision = 0.914062 (19.474 sec) +Saved checkpoint after 64 epoch(s) to data/resnet164_nb/checkpoints/00064... +INFO:tensorflow:global_step/sec: 4.58837 +INFO:tensorflow:step = 25101, loss = 0.617287, precision = 0.898438 (21.797 sec) +INFO:tensorflow:global_step/sec: 5.13479 +INFO:tensorflow:step = 25201, loss = 0.678888, precision = 0.851562 (19.476 sec) +INFO:tensorflow:global_step/sec: 5.14052 +INFO:tensorflow:step = 25301, loss = 0.624459, precision = 0.914062 (19.450 sec) +INFO:tensorflow:global_step/sec: 5.1408 +INFO:tensorflow:step = 25401, loss = 0.543654, precision = 0.929688 (19.451 sec) +Saved checkpoint after 65 epoch(s) to data/resnet164_nb/checkpoints/00065... +INFO:tensorflow:global_step/sec: 4.59602 +INFO:tensorflow:step = 25501, loss = 0.589975, precision = 0.898438 (21.759 sec) +INFO:tensorflow:global_step/sec: 5.14278 +INFO:tensorflow:step = 25601, loss = 0.614083, precision = 0.929688 (19.447 sec) +INFO:tensorflow:global_step/sec: 5.13666 +INFO:tensorflow:step = 25701, loss = 0.799694, precision = 0.828125 (19.466 sec) +INFO:tensorflow:global_step/sec: 5.14604 +INFO:tensorflow:step = 25801, loss = 0.674116, precision = 0.882812 (19.435 sec) +Saved checkpoint after 66 epoch(s) to data/resnet164_nb/checkpoints/00066... +INFO:tensorflow:global_step/sec: 4.55062 +INFO:tensorflow:step = 25901, loss = 0.64169, precision = 0.90625 (21.977 sec) +INFO:tensorflow:global_step/sec: 5.13765 +INFO:tensorflow:step = 26001, loss = 0.560083, precision = 0.929688 (19.460 sec) +INFO:tensorflow:global_step/sec: 5.13885 +INFO:tensorflow:step = 26101, loss = 0.530625, precision = 0.929688 (19.466 sec) +Saved checkpoint after 67 epoch(s) to data/resnet164_nb/checkpoints/00067... +INFO:tensorflow:global_step/sec: 4.58051 +INFO:tensorflow:step = 26201, loss = 0.628154, precision = 0.914062 (21.825 sec) +INFO:tensorflow:global_step/sec: 5.13445 +INFO:tensorflow:step = 26301, loss = 0.718855, precision = 0.867188 (19.476 sec) +INFO:tensorflow:global_step/sec: 5.14492 +INFO:tensorflow:step = 26401, loss = 0.544353, precision = 0.914062 (19.442 sec) +INFO:tensorflow:global_step/sec: 5.14662 +INFO:tensorflow:step = 26501, loss = 0.604969, precision = 0.875 (19.425 sec) +Saved checkpoint after 68 epoch(s) to data/resnet164_nb/checkpoints/00068... +INFO:tensorflow:global_step/sec: 4.57705 +INFO:tensorflow:step = 26601, loss = 0.619066, precision = 0.882812 (21.851 sec) +INFO:tensorflow:global_step/sec: 5.13729 +INFO:tensorflow:step = 26701, loss = 0.628164, precision = 0.890625 (19.463 sec) +INFO:tensorflow:global_step/sec: 5.14398 +INFO:tensorflow:step = 26801, loss = 0.526494, precision = 0.945312 (19.440 sec) +INFO:tensorflow:global_step/sec: 5.14412 +INFO:tensorflow:step = 26901, loss = 0.678756, precision = 0.859375 (19.442 sec) +Saved checkpoint after 69 epoch(s) to data/resnet164_nb/checkpoints/00069... +INFO:tensorflow:global_step/sec: 4.52777 +INFO:tensorflow:step = 27001, loss = 0.468671, precision = 0.945312 (22.086 sec) +INFO:tensorflow:global_step/sec: 5.14207 +INFO:tensorflow:step = 27101, loss = 0.595761, precision = 0.898438 (19.451 sec) +INFO:tensorflow:global_step/sec: 5.13917 +INFO:tensorflow:step = 27201, loss = 0.552874, precision = 0.914062 (19.452 sec) +INFO:tensorflow:global_step/sec: 5.14003 +INFO:tensorflow:step = 27301, loss = 0.610258, precision = 0.914062 (19.455 sec) +Saved checkpoint after 70 epoch(s) to data/resnet164_nb/checkpoints/00070... +INFO:tensorflow:global_step/sec: 4.4937 +INFO:tensorflow:step = 27401, loss = 0.571751, precision = 0.90625 (22.254 sec) +INFO:tensorflow:global_step/sec: 5.13992 +INFO:tensorflow:step = 27501, loss = 0.633182, precision = 0.867188 (19.455 sec) +INFO:tensorflow:global_step/sec: 5.1315 +INFO:tensorflow:step = 27601, loss = 0.616509, precision = 0.90625 (19.487 sec) +INFO:tensorflow:global_step/sec: 5.14802 +INFO:tensorflow:step = 27701, loss = 0.628271, precision = 0.875 (19.425 sec) +Saved checkpoint after 71 epoch(s) to data/resnet164_nb/checkpoints/00071... +INFO:tensorflow:global_step/sec: 4.58505 +INFO:tensorflow:step = 27801, loss = 0.71518, precision = 0.890625 (21.812 sec) +INFO:tensorflow:global_step/sec: 5.13741 +INFO:tensorflow:step = 27901, loss = 0.634384, precision = 0.867188 (19.466 sec) +INFO:tensorflow:global_step/sec: 5.14293 +INFO:tensorflow:step = 28001, loss = 0.689239, precision = 0.875 (19.442 sec) +INFO:tensorflow:global_step/sec: 5.13958 +INFO:tensorflow:step = 28101, loss = 0.649501, precision = 0.90625 (19.457 sec) +Saved checkpoint after 72 epoch(s) to data/resnet164_nb/checkpoints/00072... +INFO:tensorflow:global_step/sec: 4.602 +INFO:tensorflow:step = 28201, loss = 0.856322, precision = 0.835938 (21.731 sec) +INFO:tensorflow:global_step/sec: 5.13984 +INFO:tensorflow:step = 28301, loss = 0.633288, precision = 0.90625 (19.455 sec) +INFO:tensorflow:global_step/sec: 5.14577 +INFO:tensorflow:step = 28401, loss = 0.59221, precision = 0.929688 (19.435 sec) +INFO:tensorflow:global_step/sec: 5.13834 +INFO:tensorflow:step = 28501, loss = 0.771024, precision = 0.859375 (19.460 sec) +Saved checkpoint after 73 epoch(s) to data/resnet164_nb/checkpoints/00073... +INFO:tensorflow:global_step/sec: 4.4921 +INFO:tensorflow:step = 28601, loss = 0.636127, precision = 0.882812 (22.261 sec) +INFO:tensorflow:global_step/sec: 5.14225 +INFO:tensorflow:step = 28701, loss = 0.636063, precision = 0.867188 (19.448 sec) +INFO:tensorflow:global_step/sec: 5.14045 +INFO:tensorflow:step = 28801, loss = 0.739605, precision = 0.84375 (19.453 sec) +INFO:tensorflow:global_step/sec: 5.13957 +INFO:tensorflow:step = 28901, loss = 0.562614, precision = 0.921875 (19.457 sec) +Saved checkpoint after 74 epoch(s) to data/resnet164_nb/checkpoints/00074... +INFO:tensorflow:global_step/sec: 4.5877 +INFO:tensorflow:step = 29001, loss = 0.688546, precision = 0.90625 (21.798 sec) +INFO:tensorflow:global_step/sec: 5.14 +INFO:tensorflow:step = 29101, loss = 0.606661, precision = 0.90625 (19.457 sec) +INFO:tensorflow:global_step/sec: 5.13811 +INFO:tensorflow:step = 29201, loss = 0.661254, precision = 0.890625 (19.461 sec) +INFO:tensorflow:global_step/sec: 5.13282 +INFO:tensorflow:step = 29301, loss = 0.659574, precision = 0.898438 (19.486 sec) +Saved checkpoint after 75 epoch(s) to data/resnet164_nb/checkpoints/00075... +INFO:tensorflow:global_step/sec: 4.61642 +INFO:tensorflow:step = 29401, loss = 0.682205, precision = 0.851562 (21.662 sec) +INFO:tensorflow:global_step/sec: 5.1395 +INFO:tensorflow:step = 29501, loss = 0.665741, precision = 0.851562 (19.456 sec) +INFO:tensorflow:global_step/sec: 5.14104 +INFO:tensorflow:step = 29601, loss = 0.634875, precision = 0.898438 (19.451 sec) +INFO:tensorflow:global_step/sec: 5.1463 +INFO:tensorflow:step = 29701, loss = 0.592633, precision = 0.867188 (19.433 sec) +Saved checkpoint after 76 epoch(s) to data/resnet164_nb/checkpoints/00076... +INFO:tensorflow:global_step/sec: 4.52689 +INFO:tensorflow:step = 29801, loss = 0.669194, precision = 0.882812 (22.087 sec) +INFO:tensorflow:global_step/sec: 5.14176 +INFO:tensorflow:step = 29901, loss = 0.655538, precision = 0.890625 (19.448 sec) +INFO:tensorflow:global_step/sec: 5.1365 +INFO:tensorflow:step = 30001, loss = 0.592294, precision = 0.898438 (19.468 sec) +INFO:tensorflow:global_step/sec: 5.14461 +INFO:tensorflow:step = 30101, loss = 0.588378, precision = 0.90625 (19.443 sec) +Saved checkpoint after 77 epoch(s) to data/resnet164_nb/checkpoints/00077... +INFO:tensorflow:global_step/sec: 4.59487 +INFO:tensorflow:step = 30201, loss = 0.729245, precision = 0.890625 (21.761 sec) +INFO:tensorflow:global_step/sec: 5.1401 +INFO:tensorflow:step = 30301, loss = 0.520931, precision = 0.921875 (19.452 sec) +INFO:tensorflow:global_step/sec: 5.1403 +INFO:tensorflow:step = 30401, loss = 0.767407, precision = 0.828125 (19.459 sec) +Saved checkpoint after 78 epoch(s) to data/resnet164_nb/checkpoints/00078... +INFO:tensorflow:global_step/sec: 4.58554 +INFO:tensorflow:step = 30501, loss = 0.546076, precision = 0.90625 (21.804 sec) +INFO:tensorflow:global_step/sec: 5.12703 +INFO:tensorflow:step = 30601, loss = 0.678267, precision = 0.890625 (19.505 sec) +INFO:tensorflow:global_step/sec: 5.14283 +INFO:tensorflow:step = 30701, loss = 0.687247, precision = 0.875 (19.445 sec) +INFO:tensorflow:global_step/sec: 5.14259 +INFO:tensorflow:step = 30801, loss = 0.60873, precision = 0.890625 (19.444 sec) +Saved checkpoint after 79 epoch(s) to data/resnet164_nb/checkpoints/00079... +INFO:tensorflow:global_step/sec: 4.56706 +INFO:tensorflow:step = 30901, loss = 0.66072, precision = 0.882812 (21.897 sec) +INFO:tensorflow:global_step/sec: 5.13673 +INFO:tensorflow:step = 31001, loss = 0.671805, precision = 0.890625 (19.466 sec) +INFO:tensorflow:global_step/sec: 5.14163 +INFO:tensorflow:step = 31101, loss = 0.627488, precision = 0.898438 (19.453 sec) +INFO:tensorflow:global_step/sec: 5.14454 +INFO:tensorflow:step = 31201, loss = 0.678598, precision = 0.875 (19.435 sec) +Saved checkpoint after 80 epoch(s) to data/resnet164_nb/checkpoints/00080... +INFO:tensorflow:global_step/sec: 4.55318 +INFO:tensorflow:step = 31301, loss = 0.671521, precision = 0.890625 (21.963 sec) +INFO:tensorflow:global_step/sec: 5.13755 +INFO:tensorflow:step = 31401, loss = 0.553926, precision = 0.929688 (19.466 sec) +INFO:tensorflow:global_step/sec: 5.14391 +INFO:tensorflow:step = 31501, loss = 0.906099, precision = 0.773438 (19.439 sec) +INFO:tensorflow:global_step/sec: 5.13485 +INFO:tensorflow:step = 31601, loss = 0.583978, precision = 0.929688 (19.478 sec) +Saved checkpoint after 81 epoch(s) to data/resnet164_nb/checkpoints/00081... +INFO:tensorflow:global_step/sec: 4.58969 +INFO:tensorflow:step = 31701, loss = 0.652399, precision = 0.867188 (21.788 sec) +INFO:tensorflow:global_step/sec: 5.14453 +INFO:tensorflow:step = 31801, loss = 0.72385, precision = 0.828125 (19.436 sec) +INFO:tensorflow:global_step/sec: 5.15171 +INFO:tensorflow:step = 31901, loss = 0.65678, precision = 0.898438 (19.411 sec) +INFO:tensorflow:global_step/sec: 5.14889 +INFO:tensorflow:step = 32001, loss = 0.581689, precision = 0.921875 (19.420 sec) +Saved checkpoint after 82 epoch(s) to data/resnet164_nb/checkpoints/00082... +INFO:tensorflow:global_step/sec: 4.58352 +INFO:tensorflow:step = 32101, loss = 0.521254, precision = 0.945312 (21.818 sec) +INFO:tensorflow:global_step/sec: 5.14538 +INFO:tensorflow:step = 32201, loss = 0.618539, precision = 0.914062 (19.435 sec) +INFO:tensorflow:global_step/sec: 5.15158 +INFO:tensorflow:step = 32301, loss = 0.728502, precision = 0.867188 (19.417 sec) +INFO:tensorflow:global_step/sec: 5.1367 +INFO:tensorflow:step = 32401, loss = 0.629914, precision = 0.90625 (19.468 sec) +Saved checkpoint after 83 epoch(s) to data/resnet164_nb/checkpoints/00083... +INFO:tensorflow:global_step/sec: 4.52965 +INFO:tensorflow:step = 32501, loss = 0.623073, precision = 0.90625 (22.072 sec) +INFO:tensorflow:global_step/sec: 5.1412 +INFO:tensorflow:step = 32601, loss = 0.644676, precision = 0.875 (19.451 sec) +INFO:tensorflow:global_step/sec: 5.13519 +INFO:tensorflow:step = 32701, loss = 0.570638, precision = 0.914062 (19.474 sec) +INFO:tensorflow:global_step/sec: 5.13671 +INFO:tensorflow:step = 32801, loss = 0.784894, precision = 0.867188 (19.470 sec) +Saved checkpoint after 84 epoch(s) to data/resnet164_nb/checkpoints/00084... +INFO:tensorflow:global_step/sec: 4.55712 +INFO:tensorflow:step = 32901, loss = 0.654317, precision = 0.875 (21.942 sec) +INFO:tensorflow:global_step/sec: 5.13922 +INFO:tensorflow:step = 33001, loss = 0.751802, precision = 0.859375 (19.460 sec) +INFO:tensorflow:global_step/sec: 5.14731 +INFO:tensorflow:step = 33101, loss = 0.610558, precision = 0.890625 (19.426 sec) +INFO:tensorflow:global_step/sec: 5.14377 +INFO:tensorflow:step = 33201, loss = 0.660177, precision = 0.875 (19.441 sec) +Saved checkpoint after 85 epoch(s) to data/resnet164_nb/checkpoints/00085... +INFO:tensorflow:global_step/sec: 4.60022 +INFO:tensorflow:step = 33301, loss = 0.666092, precision = 0.890625 (21.738 sec) +INFO:tensorflow:global_step/sec: 5.13353 +INFO:tensorflow:step = 33401, loss = 0.694091, precision = 0.875 (19.482 sec) +INFO:tensorflow:global_step/sec: 5.13908 +INFO:tensorflow:step = 33501, loss = 0.675605, precision = 0.875 (19.456 sec) +INFO:tensorflow:global_step/sec: 5.14188 +INFO:tensorflow:step = 33601, loss = 0.601773, precision = 0.914062 (19.451 sec) +Saved checkpoint after 86 epoch(s) to data/resnet164_nb/checkpoints/00086... +INFO:tensorflow:global_step/sec: 4.55284 +INFO:tensorflow:step = 33701, loss = 0.717603, precision = 0.882812 (21.965 sec) +INFO:tensorflow:global_step/sec: 5.14153 +INFO:tensorflow:step = 33801, loss = 0.577381, precision = 0.898438 (19.446 sec) +INFO:tensorflow:global_step/sec: 5.14323 +INFO:tensorflow:step = 33901, loss = 0.652375, precision = 0.890625 (19.443 sec) +INFO:tensorflow:global_step/sec: 5.1552 +INFO:tensorflow:step = 34001, loss = 0.731492, precision = 0.875 (19.404 sec) +Saved checkpoint after 87 epoch(s) to data/resnet164_nb/checkpoints/00087... +INFO:tensorflow:global_step/sec: 4.52704 +INFO:tensorflow:step = 34101, loss = 0.687587, precision = 0.882812 (22.086 sec) +INFO:tensorflow:global_step/sec: 5.14766 +INFO:tensorflow:step = 34201, loss = 0.547817, precision = 0.9375 (19.426 sec) +INFO:tensorflow:global_step/sec: 5.15327 +INFO:tensorflow:step = 34301, loss = 0.512613, precision = 0.929688 (19.403 sec) +INFO:tensorflow:global_step/sec: 5.14568 +INFO:tensorflow:step = 34401, loss = 0.547154, precision = 0.914062 (19.434 sec) +Saved checkpoint after 88 epoch(s) to data/resnet164_nb/checkpoints/00088... +INFO:tensorflow:global_step/sec: 4.53159 +INFO:tensorflow:step = 34501, loss = 0.637635, precision = 0.890625 (22.068 sec) +INFO:tensorflow:global_step/sec: 5.14574 +INFO:tensorflow:step = 34601, loss = 0.599658, precision = 0.914062 (19.437 sec) +INFO:tensorflow:global_step/sec: 5.14925 +INFO:tensorflow:step = 34701, loss = 0.571993, precision = 0.929688 (19.417 sec) +Saved checkpoint after 89 epoch(s) to data/resnet164_nb/checkpoints/00089... +INFO:tensorflow:global_step/sec: 4.48962 +INFO:tensorflow:step = 34801, loss = 0.561183, precision = 0.90625 (22.274 sec) +INFO:tensorflow:global_step/sec: 5.14083 +INFO:tensorflow:step = 34901, loss = 0.646417, precision = 0.914062 (19.453 sec) +INFO:tensorflow:global_step/sec: 5.15645 +INFO:tensorflow:step = 35001, loss = 0.820465, precision = 0.84375 (19.398 sec) +INFO:tensorflow:global_step/sec: 5.14717 +INFO:tensorflow:step = 35101, loss = 0.494309, precision = 0.929688 (19.424 sec) +Saved checkpoint after 90 epoch(s) to data/resnet164_nb/checkpoints/00090... +INFO:tensorflow:global_step/sec: 4.58573 +INFO:tensorflow:step = 35201, loss = 0.673011, precision = 0.882812 (21.807 sec) +INFO:tensorflow:global_step/sec: 5.15035 +INFO:tensorflow:step = 35301, loss = 0.701973, precision = 0.875 (19.417 sec) +INFO:tensorflow:global_step/sec: 5.14534 +INFO:tensorflow:step = 35401, loss = 0.668115, precision = 0.859375 (19.434 sec) +INFO:tensorflow:global_step/sec: 5.14523 +INFO:tensorflow:step = 35501, loss = 0.593057, precision = 0.90625 (19.441 sec) +Saved checkpoint after 91 epoch(s) to data/resnet164_nb/checkpoints/00091... +INFO:tensorflow:global_step/sec: 4.57517 +INFO:tensorflow:step = 35601, loss = 0.656758, precision = 0.90625 (21.852 sec) +INFO:tensorflow:global_step/sec: 5.14854 +INFO:tensorflow:step = 35701, loss = 0.474812, precision = 0.945312 (19.423 sec) +INFO:tensorflow:global_step/sec: 5.138 +INFO:tensorflow:step = 35801, loss = 0.527894, precision = 0.945312 (19.463 sec) +INFO:tensorflow:global_step/sec: 5.13562 +INFO:tensorflow:step = 35901, loss = 0.524473, precision = 0.921875 (19.472 sec) +Saved checkpoint after 92 epoch(s) to data/resnet164_nb/checkpoints/00092... +INFO:tensorflow:global_step/sec: 4.58564 +INFO:tensorflow:step = 36001, loss = 0.483103, precision = 0.945312 (21.807 sec) +INFO:tensorflow:global_step/sec: 5.12839 +INFO:tensorflow:step = 36101, loss = 0.506885, precision = 0.945312 (19.504 sec) +INFO:tensorflow:global_step/sec: 5.14541 +INFO:tensorflow:step = 36201, loss = 0.405274, precision = 0.96875 (19.430 sec) +INFO:tensorflow:global_step/sec: 5.1403 +INFO:tensorflow:step = 36301, loss = 0.514382, precision = 0.914062 (19.454 sec) +Saved checkpoint after 93 epoch(s) to data/resnet164_nb/checkpoints/00093... +INFO:tensorflow:global_step/sec: 4.58757 +INFO:tensorflow:step = 36401, loss = 0.362444, precision = 0.984375 (21.799 sec) +INFO:tensorflow:global_step/sec: 5.14314 +INFO:tensorflow:step = 36501, loss = 0.378685, precision = 0.976562 (19.443 sec) +INFO:tensorflow:global_step/sec: 5.14147 +INFO:tensorflow:step = 36601, loss = 0.374572, precision = 0.96875 (19.449 sec) +INFO:tensorflow:global_step/sec: 5.14877 +INFO:tensorflow:step = 36701, loss = 0.461285, precision = 0.945312 (19.422 sec) +Saved checkpoint after 94 epoch(s) to data/resnet164_nb/checkpoints/00094... +INFO:tensorflow:global_step/sec: 4.5786 +INFO:tensorflow:step = 36801, loss = 0.403179, precision = 0.945312 (21.842 sec) +INFO:tensorflow:global_step/sec: 5.1485 +INFO:tensorflow:step = 36901, loss = 0.388456, precision = 0.976562 (19.423 sec) +INFO:tensorflow:global_step/sec: 5.13183 +INFO:tensorflow:step = 37001, loss = 0.38592, precision = 0.96875 (19.495 sec) +INFO:tensorflow:global_step/sec: 5.14146 +INFO:tensorflow:step = 37101, loss = 0.397135, precision = 0.953125 (19.441 sec) +Saved checkpoint after 95 epoch(s) to data/resnet164_nb/checkpoints/00095... +INFO:tensorflow:global_step/sec: 4.54497 +INFO:tensorflow:step = 37201, loss = 0.429952, precision = 0.953125 (22.004 sec) +INFO:tensorflow:global_step/sec: 5.14321 +INFO:tensorflow:step = 37301, loss = 0.370998, precision = 0.960938 (19.442 sec) +INFO:tensorflow:global_step/sec: 5.15447 +INFO:tensorflow:step = 37401, loss = 0.341337, precision = 0.976562 (19.401 sec) +INFO:tensorflow:global_step/sec: 5.14027 +INFO:tensorflow:step = 37501, loss = 0.425742, precision = 0.960938 (19.454 sec) +Saved checkpoint after 96 epoch(s) to data/resnet164_nb/checkpoints/00096... +INFO:tensorflow:global_step/sec: 4.58797 +INFO:tensorflow:step = 37601, loss = 0.34912, precision = 0.984375 (21.796 sec) +INFO:tensorflow:global_step/sec: 5.15641 +INFO:tensorflow:step = 37701, loss = 0.339571, precision = 0.984375 (19.393 sec) +INFO:tensorflow:global_step/sec: 5.14823 +INFO:tensorflow:step = 37801, loss = 0.418156, precision = 0.960938 (19.425 sec) +INFO:tensorflow:global_step/sec: 5.14042 +INFO:tensorflow:step = 37901, loss = 0.304691, precision = 1.0 (19.454 sec) +Saved checkpoint after 97 epoch(s) to data/resnet164_nb/checkpoints/00097... +INFO:tensorflow:global_step/sec: 4.58507 +INFO:tensorflow:step = 38001, loss = 0.329394, precision = 0.984375 (21.812 sec) +INFO:tensorflow:global_step/sec: 5.14797 +INFO:tensorflow:step = 38101, loss = 0.381786, precision = 0.960938 (19.423 sec) +INFO:tensorflow:global_step/sec: 5.13231 +INFO:tensorflow:step = 38201, loss = 0.326688, precision = 0.976562 (19.484 sec) +INFO:tensorflow:global_step/sec: 5.13398 +INFO:tensorflow:step = 38301, loss = 0.348839, precision = 0.96875 (19.486 sec) +Saved checkpoint after 98 epoch(s) to data/resnet164_nb/checkpoints/00098... +INFO:tensorflow:global_step/sec: 4.56459 +INFO:tensorflow:step = 38401, loss = 0.308306, precision = 0.976562 (21.900 sec) +INFO:tensorflow:global_step/sec: 5.13867 +INFO:tensorflow:step = 38501, loss = 0.297831, precision = 0.984375 (19.460 sec) +INFO:tensorflow:global_step/sec: 5.13281 +INFO:tensorflow:step = 38601, loss = 0.328927, precision = 0.976562 (19.485 sec) +INFO:tensorflow:global_step/sec: 5.13653 +INFO:tensorflow:step = 38701, loss = 0.339517, precision = 0.960938 (19.466 sec) +Saved checkpoint after 99 epoch(s) to data/resnet164_nb/checkpoints/00099... +INFO:tensorflow:global_step/sec: 4.54726 +INFO:tensorflow:step = 38801, loss = 0.288725, precision = 0.984375 (21.992 sec) +INFO:tensorflow:global_step/sec: 5.14258 +INFO:tensorflow:step = 38901, loss = 0.292482, precision = 0.992188 (19.446 sec) +INFO:tensorflow:global_step/sec: 5.14092 +INFO:tensorflow:step = 39001, loss = 0.32191, precision = 0.984375 (19.451 sec) +Saved checkpoint after 100 epoch(s) to data/resnet164_nb/checkpoints/00100... +INFO:tensorflow:global_step/sec: 4.59596 +INFO:tensorflow:step = 39101, loss = 0.274565, precision = 0.992188 (21.762 sec) +INFO:tensorflow:global_step/sec: 5.12335 +INFO:tensorflow:step = 39201, loss = 0.275659, precision = 0.992188 (19.515 sec) +INFO:tensorflow:global_step/sec: 5.14587 +INFO:tensorflow:step = 39301, loss = 0.295585, precision = 0.984375 (19.438 sec) +INFO:tensorflow:global_step/sec: 5.14439 +INFO:tensorflow:step = 39401, loss = 0.335045, precision = 0.96875 (19.433 sec) +Saved checkpoint after 101 epoch(s) to data/resnet164_nb/checkpoints/00101... +INFO:tensorflow:global_step/sec: 4.56535 +INFO:tensorflow:step = 39501, loss = 0.274651, precision = 0.984375 (21.905 sec) +INFO:tensorflow:global_step/sec: 5.14635 +INFO:tensorflow:step = 39601, loss = 0.320433, precision = 0.976562 (19.431 sec) +INFO:tensorflow:global_step/sec: 5.14796 +INFO:tensorflow:step = 39701, loss = 0.26672, precision = 0.992188 (19.428 sec) +INFO:tensorflow:global_step/sec: 5.15127 +INFO:tensorflow:step = 39801, loss = 0.287949, precision = 0.984375 (19.410 sec) +Saved checkpoint after 102 epoch(s) to data/resnet164_nb/checkpoints/00102... +INFO:tensorflow:global_step/sec: 4.5571 +INFO:tensorflow:step = 39901, loss = 0.284691, precision = 0.960938 (21.942 sec) +INFO:tensorflow:global_step/sec: 5.14211 +INFO:tensorflow:step = 40001, loss = 0.266894, precision = 0.984375 (19.452 sec) +INFO:tensorflow:global_step/sec: 5.14101 +INFO:tensorflow:step = 40101, loss = 0.262811, precision = 0.984375 (19.449 sec) +INFO:tensorflow:global_step/sec: 5.13869 +INFO:tensorflow:step = 40201, loss = 0.26692, precision = 0.992188 (19.459 sec) +Saved checkpoint after 103 epoch(s) to data/resnet164_nb/checkpoints/00103... +INFO:tensorflow:global_step/sec: 4.51063 +INFO:tensorflow:step = 40301, loss = 0.251352, precision = 0.992188 (22.171 sec) +INFO:tensorflow:global_step/sec: 5.13771 +INFO:tensorflow:step = 40401, loss = 0.269465, precision = 0.984375 (19.467 sec) +INFO:tensorflow:global_step/sec: 5.14261 +INFO:tensorflow:step = 40501, loss = 0.333277, precision = 0.96875 (19.446 sec) +INFO:tensorflow:global_step/sec: 5.13679 +INFO:tensorflow:step = 40601, loss = 0.270966, precision = 0.976562 (19.468 sec) +Saved checkpoint after 104 epoch(s) to data/resnet164_nb/checkpoints/00104... +INFO:tensorflow:global_step/sec: 4.59505 +INFO:tensorflow:step = 40701, loss = 0.278028, precision = 0.96875 (21.758 sec) +INFO:tensorflow:global_step/sec: 5.13722 +INFO:tensorflow:step = 40801, loss = 0.271843, precision = 0.984375 (19.466 sec) +INFO:tensorflow:global_step/sec: 5.15155 +INFO:tensorflow:step = 40901, loss = 0.271003, precision = 0.984375 (19.411 sec) +INFO:tensorflow:global_step/sec: 5.14985 +INFO:tensorflow:step = 41001, loss = 0.303139, precision = 0.96875 (19.419 sec) +Saved checkpoint after 105 epoch(s) to data/resnet164_nb/checkpoints/00105... +INFO:tensorflow:global_step/sec: 4.5837 +INFO:tensorflow:step = 41101, loss = 0.231775, precision = 0.992188 (21.818 sec) +INFO:tensorflow:global_step/sec: 5.14077 +INFO:tensorflow:step = 41201, loss = 0.233328, precision = 0.992188 (19.452 sec) +INFO:tensorflow:global_step/sec: 5.14489 +INFO:tensorflow:step = 41301, loss = 0.261662, precision = 0.984375 (19.435 sec) +INFO:tensorflow:global_step/sec: 5.14067 +INFO:tensorflow:step = 41401, loss = 0.23239, precision = 0.976562 (19.455 sec) +Saved checkpoint after 106 epoch(s) to data/resnet164_nb/checkpoints/00106... +INFO:tensorflow:global_step/sec: 4.57719 +INFO:tensorflow:step = 41501, loss = 0.28303, precision = 0.976562 (21.848 sec) +INFO:tensorflow:global_step/sec: 5.13931 +INFO:tensorflow:step = 41601, loss = 0.24078, precision = 0.984375 (19.462 sec) +INFO:tensorflow:global_step/sec: 5.15326 +INFO:tensorflow:step = 41701, loss = 0.243246, precision = 0.976562 (19.398 sec) +INFO:tensorflow:global_step/sec: 5.13672 +INFO:tensorflow:step = 41801, loss = 0.247553, precision = 0.992188 (19.468 sec) +Saved checkpoint after 107 epoch(s) to data/resnet164_nb/checkpoints/00107... +INFO:tensorflow:global_step/sec: 4.5591 +INFO:tensorflow:step = 41901, loss = 0.275477, precision = 0.960938 (21.934 sec) +INFO:tensorflow:global_step/sec: 5.15036 +INFO:tensorflow:step = 42001, loss = 0.282186, precision = 0.96875 (19.416 sec) +INFO:tensorflow:global_step/sec: 5.15647 +INFO:tensorflow:step = 42101, loss = 0.278312, precision = 0.976562 (19.393 sec) +INFO:tensorflow:global_step/sec: 5.14687 +INFO:tensorflow:step = 42201, loss = 0.238858, precision = 0.992188 (19.435 sec) +Saved checkpoint after 108 epoch(s) to data/resnet164_nb/checkpoints/00108... +INFO:tensorflow:global_step/sec: 4.57092 +INFO:tensorflow:step = 42301, loss = 0.293197, precision = 0.960938 (21.872 sec) +INFO:tensorflow:global_step/sec: 5.15014 +INFO:tensorflow:step = 42401, loss = 0.28315, precision = 0.960938 (19.421 sec) +INFO:tensorflow:global_step/sec: 5.14597 +INFO:tensorflow:step = 42501, loss = 0.236439, precision = 0.976562 (19.429 sec) +INFO:tensorflow:global_step/sec: 5.1443 +INFO:tensorflow:step = 42601, loss = 0.250384, precision = 0.976562 (19.439 sec) +Saved checkpoint after 109 epoch(s) to data/resnet164_nb/checkpoints/00109... +INFO:tensorflow:global_step/sec: 4.49853 +INFO:tensorflow:step = 42701, loss = 0.234677, precision = 0.976562 (22.229 sec) +INFO:tensorflow:global_step/sec: 5.16104 +INFO:tensorflow:step = 42801, loss = 0.247321, precision = 0.96875 (19.376 sec) +INFO:tensorflow:global_step/sec: 5.16095 +INFO:tensorflow:step = 42901, loss = 0.197983, precision = 1.0 (19.380 sec) +INFO:tensorflow:global_step/sec: 5.1432 +INFO:tensorflow:step = 43001, loss = 0.274906, precision = 0.976562 (19.440 sec) +Saved checkpoint after 110 epoch(s) to data/resnet164_nb/checkpoints/00110... +INFO:tensorflow:global_step/sec: 4.57056 +INFO:tensorflow:step = 43101, loss = 0.208093, precision = 1.0 (21.880 sec) +INFO:tensorflow:global_step/sec: 5.1413 +INFO:tensorflow:step = 43201, loss = 0.193576, precision = 1.0 (19.450 sec) +INFO:tensorflow:global_step/sec: 5.15073 +INFO:tensorflow:step = 43301, loss = 0.202001, precision = 1.0 (19.417 sec) +Saved checkpoint after 111 epoch(s) to data/resnet164_nb/checkpoints/00111... +INFO:tensorflow:global_step/sec: 4.5939 +INFO:tensorflow:step = 43401, loss = 0.194974, precision = 1.0 (21.765 sec) +INFO:tensorflow:global_step/sec: 5.13364 +INFO:tensorflow:step = 43501, loss = 0.218645, precision = 0.976562 (19.482 sec) +INFO:tensorflow:global_step/sec: 5.14494 +INFO:tensorflow:step = 43601, loss = 0.229014, precision = 0.992188 (19.434 sec) +INFO:tensorflow:global_step/sec: 5.14229 +INFO:tensorflow:step = 43701, loss = 0.209002, precision = 0.992188 (19.452 sec) +Saved checkpoint after 112 epoch(s) to data/resnet164_nb/checkpoints/00112... +INFO:tensorflow:global_step/sec: 4.51735 +INFO:tensorflow:step = 43801, loss = 0.242439, precision = 0.984375 (22.132 sec) +INFO:tensorflow:global_step/sec: 5.14486 +INFO:tensorflow:step = 43901, loss = 0.198011, precision = 1.0 (19.440 sec) +INFO:tensorflow:global_step/sec: 5.14447 +INFO:tensorflow:step = 44001, loss = 0.217598, precision = 0.984375 (19.436 sec) +INFO:tensorflow:global_step/sec: 5.14339 +INFO:tensorflow:step = 44101, loss = 0.253496, precision = 0.96875 (19.442 sec) +Saved checkpoint after 113 epoch(s) to data/resnet164_nb/checkpoints/00113... +INFO:tensorflow:global_step/sec: 4.56241 +INFO:tensorflow:step = 44201, loss = 0.185607, precision = 1.0 (21.921 sec) +INFO:tensorflow:global_step/sec: 5.14511 +INFO:tensorflow:step = 44301, loss = 0.234179, precision = 0.984375 (19.433 sec) +INFO:tensorflow:global_step/sec: 5.15106 +INFO:tensorflow:step = 44401, loss = 0.197475, precision = 0.992188 (19.414 sec) +INFO:tensorflow:global_step/sec: 5.14904 +INFO:tensorflow:step = 44501, loss = 0.186736, precision = 1.0 (19.421 sec) +Saved checkpoint after 114 epoch(s) to data/resnet164_nb/checkpoints/00114... +INFO:tensorflow:global_step/sec: 4.57577 +INFO:tensorflow:step = 44601, loss = 0.186398, precision = 1.0 (21.855 sec) +INFO:tensorflow:global_step/sec: 5.13441 +INFO:tensorflow:step = 44701, loss = 0.217787, precision = 0.984375 (19.476 sec) +INFO:tensorflow:global_step/sec: 5.14274 +INFO:tensorflow:step = 44801, loss = 0.260082, precision = 0.976562 (19.446 sec) +INFO:tensorflow:global_step/sec: 5.14561 +INFO:tensorflow:step = 44901, loss = 0.201498, precision = 0.984375 (19.432 sec) +Saved checkpoint after 115 epoch(s) to data/resnet164_nb/checkpoints/00115... +INFO:tensorflow:global_step/sec: 4.59448 +INFO:tensorflow:step = 45001, loss = 0.186809, precision = 1.0 (21.768 sec) +INFO:tensorflow:global_step/sec: 5.13051 +INFO:tensorflow:step = 45101, loss = 0.218373, precision = 0.976562 (19.489 sec) +INFO:tensorflow:global_step/sec: 5.1429 +INFO:tensorflow:step = 45201, loss = 0.222376, precision = 0.976562 (19.449 sec) +INFO:tensorflow:global_step/sec: 5.13864 +INFO:tensorflow:step = 45301, loss = 0.19432, precision = 0.992188 (19.455 sec) +Saved checkpoint after 116 epoch(s) to data/resnet164_nb/checkpoints/00116... +INFO:tensorflow:global_step/sec: 4.59952 +INFO:tensorflow:step = 45401, loss = 0.204518, precision = 0.992188 (21.742 sec) +INFO:tensorflow:global_step/sec: 5.1524 +INFO:tensorflow:step = 45501, loss = 0.198404, precision = 0.984375 (19.408 sec) +INFO:tensorflow:global_step/sec: 5.15349 +INFO:tensorflow:step = 45601, loss = 0.187509, precision = 0.984375 (19.404 sec) +INFO:tensorflow:global_step/sec: 5.15395 +INFO:tensorflow:step = 45701, loss = 0.188624, precision = 0.992188 (19.402 sec) +Saved checkpoint after 117 epoch(s) to data/resnet164_nb/checkpoints/00117... +INFO:tensorflow:global_step/sec: 4.57385 +INFO:tensorflow:step = 45801, loss = 0.234755, precision = 0.96875 (21.867 sec) +INFO:tensorflow:global_step/sec: 5.14841 +INFO:tensorflow:step = 45901, loss = 0.187794, precision = 1.0 (19.422 sec) +INFO:tensorflow:global_step/sec: 5.14767 +INFO:tensorflow:step = 46001, loss = 0.195697, precision = 0.984375 (19.424 sec) +INFO:tensorflow:global_step/sec: 5.1461 +INFO:tensorflow:step = 46101, loss = 0.192641, precision = 0.992188 (19.432 sec) +Saved checkpoint after 118 epoch(s) to data/resnet164_nb/checkpoints/00118... +INFO:tensorflow:global_step/sec: 4.57392 +INFO:tensorflow:step = 46201, loss = 0.218326, precision = 0.992188 (21.864 sec) +INFO:tensorflow:global_step/sec: 5.15348 +INFO:tensorflow:step = 46301, loss = 0.224297, precision = 0.96875 (19.407 sec) +INFO:tensorflow:global_step/sec: 5.14978 +INFO:tensorflow:step = 46401, loss = 0.210037, precision = 0.984375 (19.415 sec) +INFO:tensorflow:global_step/sec: 5.14744 +INFO:tensorflow:step = 46501, loss = 0.206236, precision = 0.976562 (19.432 sec) +Saved checkpoint after 119 epoch(s) to data/resnet164_nb/checkpoints/00119... +INFO:tensorflow:global_step/sec: 4.57698 +INFO:tensorflow:step = 46601, loss = 0.209489, precision = 0.984375 (21.844 sec) +INFO:tensorflow:global_step/sec: 5.1486 +INFO:tensorflow:step = 46701, loss = 0.178488, precision = 1.0 (19.430 sec) +INFO:tensorflow:global_step/sec: 5.13397 +INFO:tensorflow:step = 46801, loss = 0.197056, precision = 0.984375 (19.470 sec) +INFO:tensorflow:global_step/sec: 5.1479 +INFO:tensorflow:step = 46901, loss = 0.200331, precision = 0.976562 (19.425 sec) +Saved checkpoint after 120 epoch(s) to data/resnet164_nb/checkpoints/00120... +INFO:tensorflow:global_step/sec: 4.50031 +INFO:tensorflow:step = 47001, loss = 0.229565, precision = 0.976562 (22.220 sec) +INFO:tensorflow:global_step/sec: 5.12626 +INFO:tensorflow:step = 47101, loss = 0.197852, precision = 0.984375 (19.508 sec) +INFO:tensorflow:global_step/sec: 5.14189 +INFO:tensorflow:step = 47201, loss = 0.18219, precision = 0.992188 (19.448 sec) +INFO:tensorflow:global_step/sec: 5.14193 +INFO:tensorflow:step = 47301, loss = 0.160782, precision = 1.0 (19.448 sec) +Saved checkpoint after 121 epoch(s) to data/resnet164_nb/checkpoints/00121... +INFO:tensorflow:global_step/sec: 4.58193 +INFO:tensorflow:step = 47401, loss = 0.214824, precision = 0.96875 (21.829 sec) +INFO:tensorflow:global_step/sec: 5.13113 +INFO:tensorflow:step = 47501, loss = 0.176494, precision = 0.992188 (19.488 sec) +INFO:tensorflow:global_step/sec: 5.14415 +INFO:tensorflow:step = 47601, loss = 0.181496, precision = 0.992188 (19.438 sec) +INFO:tensorflow:global_step/sec: 5.13899 +INFO:tensorflow:step = 47701, loss = 0.187993, precision = 0.984375 (19.463 sec) +Saved checkpoint after 122 epoch(s) to data/resnet164_nb/checkpoints/00122... +INFO:tensorflow:global_step/sec: 4.61664 +INFO:tensorflow:step = 47801, loss = 0.173625, precision = 0.984375 (21.657 sec) +INFO:tensorflow:global_step/sec: 5.13986 +INFO:tensorflow:step = 47901, loss = 0.210989, precision = 0.976562 (19.455 sec) +INFO:tensorflow:global_step/sec: 5.14344 +INFO:tensorflow:step = 48001, loss = 0.219845, precision = 0.976562 (19.442 sec) +Saved checkpoint after 123 epoch(s) to data/resnet164_nb/checkpoints/00123... +INFO:tensorflow:global_step/sec: 4.59738 +INFO:tensorflow:step = 48101, loss = 0.22851, precision = 0.960938 (21.752 sec) +INFO:tensorflow:global_step/sec: 5.14205 +INFO:tensorflow:step = 48201, loss = 0.199462, precision = 0.976562 (19.447 sec) +INFO:tensorflow:global_step/sec: 5.14642 +INFO:tensorflow:step = 48301, loss = 0.212422, precision = 0.976562 (19.431 sec) +INFO:tensorflow:global_step/sec: 5.14917 +INFO:tensorflow:step = 48401, loss = 0.234469, precision = 0.953125 (19.420 sec) +Saved checkpoint after 124 epoch(s) to data/resnet164_nb/checkpoints/00124... +INFO:tensorflow:global_step/sec: 4.61275 +INFO:tensorflow:step = 48501, loss = 0.173736, precision = 0.992188 (21.680 sec) +INFO:tensorflow:global_step/sec: 5.13125 +INFO:tensorflow:step = 48601, loss = 0.210289, precision = 0.976562 (19.489 sec) +INFO:tensorflow:global_step/sec: 5.1501 +INFO:tensorflow:step = 48701, loss = 0.168138, precision = 0.992188 (19.416 sec) +INFO:tensorflow:global_step/sec: 5.13648 +INFO:tensorflow:step = 48801, loss = 0.175341, precision = 0.992188 (19.469 sec) +Saved checkpoint after 125 epoch(s) to data/resnet164_nb/checkpoints/00125... +INFO:tensorflow:global_step/sec: 4.59854 +INFO:tensorflow:step = 48901, loss = 0.18991, precision = 0.976562 (21.746 sec) +INFO:tensorflow:global_step/sec: 5.14743 +INFO:tensorflow:step = 49001, loss = 0.159968, precision = 1.0 (19.435 sec) +INFO:tensorflow:global_step/sec: 5.14194 +INFO:tensorflow:step = 49101, loss = 0.213026, precision = 0.976562 (19.440 sec) +INFO:tensorflow:global_step/sec: 5.14465 +INFO:tensorflow:step = 49201, loss = 0.19091, precision = 0.992188 (19.438 sec) +Saved checkpoint after 126 epoch(s) to data/resnet164_nb/checkpoints/00126... +INFO:tensorflow:global_step/sec: 4.56055 +INFO:tensorflow:step = 49301, loss = 0.164369, precision = 0.992188 (21.927 sec) +INFO:tensorflow:global_step/sec: 5.15459 +INFO:tensorflow:step = 49401, loss = 0.151459, precision = 1.0 (19.400 sec) +INFO:tensorflow:global_step/sec: 5.1499 +INFO:tensorflow:step = 49501, loss = 0.186873, precision = 0.984375 (19.418 sec) +INFO:tensorflow:global_step/sec: 5.14918 +INFO:tensorflow:step = 49601, loss = 0.17044, precision = 1.0 (19.421 sec) +Saved checkpoint after 127 epoch(s) to data/resnet164_nb/checkpoints/00127... +INFO:tensorflow:global_step/sec: 4.58589 +INFO:tensorflow:step = 49701, loss = 0.255609, precision = 0.960938 (21.806 sec) +INFO:tensorflow:global_step/sec: 5.15152 +INFO:tensorflow:step = 49801, loss = 0.210158, precision = 0.96875 (19.412 sec) +INFO:tensorflow:global_step/sec: 5.14994 +INFO:tensorflow:step = 49901, loss = 0.175369, precision = 1.0 (19.417 sec) +INFO:tensorflow:global_step/sec: 5.15354 +INFO:tensorflow:step = 50001, loss = 0.180837, precision = 0.992188 (19.404 sec) +Saved checkpoint after 128 epoch(s) to data/resnet164_nb/checkpoints/00128... +INFO:tensorflow:global_step/sec: 4.61219 +INFO:tensorflow:step = 50101, loss = 0.177282, precision = 0.992188 (21.682 sec) +INFO:tensorflow:global_step/sec: 5.14998 +INFO:tensorflow:step = 50201, loss = 0.216266, precision = 0.984375 (19.417 sec) +INFO:tensorflow:global_step/sec: 5.15363 +INFO:tensorflow:step = 50301, loss = 0.172924, precision = 0.992188 (19.404 sec) +INFO:tensorflow:global_step/sec: 5.14779 +INFO:tensorflow:step = 50401, loss = 0.190963, precision = 0.992188 (19.427 sec) +Saved checkpoint after 129 epoch(s) to data/resnet164_nb/checkpoints/00129... +INFO:tensorflow:global_step/sec: 4.4764 +INFO:tensorflow:step = 50501, loss = 0.186193, precision = 0.984375 (22.338 sec) +INFO:tensorflow:global_step/sec: 5.14388 +INFO:tensorflow:step = 50601, loss = 0.199669, precision = 0.976562 (19.448 sec) +INFO:tensorflow:global_step/sec: 5.13818 +INFO:tensorflow:step = 50701, loss = 0.188456, precision = 0.992188 (19.455 sec) +INFO:tensorflow:global_step/sec: 5.13407 +INFO:tensorflow:step = 50801, loss = 0.179332, precision = 0.984375 (19.477 sec) +Saved checkpoint after 130 epoch(s) to data/resnet164_nb/checkpoints/00130... +INFO:tensorflow:global_step/sec: 4.50656 +INFO:tensorflow:step = 50901, loss = 0.172016, precision = 0.992188 (22.193 sec) +INFO:tensorflow:global_step/sec: 5.14517 +INFO:tensorflow:step = 51001, loss = 0.171312, precision = 0.984375 (19.434 sec) +INFO:tensorflow:global_step/sec: 5.14474 +INFO:tensorflow:step = 51101, loss = 0.231606, precision = 0.96875 (19.438 sec) +INFO:tensorflow:global_step/sec: 5.14201 +INFO:tensorflow:step = 51201, loss = 0.278386, precision = 0.9375 (19.449 sec) +Saved checkpoint after 131 epoch(s) to data/resnet164_nb/checkpoints/00131... +INFO:tensorflow:global_step/sec: 4.53903 +INFO:tensorflow:step = 51301, loss = 0.211231, precision = 0.976562 (22.027 sec) +INFO:tensorflow:global_step/sec: 5.14796 +INFO:tensorflow:step = 51401, loss = 0.179106, precision = 0.984375 (19.426 sec) +INFO:tensorflow:global_step/sec: 5.15143 +INFO:tensorflow:step = 51501, loss = 0.187078, precision = 0.984375 (19.411 sec) +INFO:tensorflow:global_step/sec: 5.14385 +INFO:tensorflow:step = 51601, loss = 0.160599, precision = 0.992188 (19.441 sec) +Saved checkpoint after 132 epoch(s) to data/resnet164_nb/checkpoints/00132... +INFO:tensorflow:global_step/sec: 4.53849 +INFO:tensorflow:step = 51701, loss = 0.205985, precision = 0.96875 (22.034 sec) +INFO:tensorflow:global_step/sec: 5.15298 +INFO:tensorflow:step = 51801, loss = 0.143155, precision = 1.0 (19.406 sec) +INFO:tensorflow:global_step/sec: 5.15248 +INFO:tensorflow:step = 51901, loss = 0.162316, precision = 0.992188 (19.408 sec) +INFO:tensorflow:global_step/sec: 5.15144 +INFO:tensorflow:step = 52001, loss = 0.213088, precision = 0.984375 (19.412 sec) +Saved checkpoint after 133 epoch(s) to data/resnet164_nb/checkpoints/00133... +INFO:tensorflow:global_step/sec: 4.58431 +INFO:tensorflow:step = 52101, loss = 0.150499, precision = 1.0 (21.814 sec) +INFO:tensorflow:global_step/sec: 5.1471 +INFO:tensorflow:step = 52201, loss = 0.161449, precision = 0.992188 (19.428 sec) +INFO:tensorflow:global_step/sec: 5.13147 +INFO:tensorflow:step = 52301, loss = 0.178267, precision = 0.992188 (19.487 sec) +Saved checkpoint after 134 epoch(s) to data/resnet164_nb/checkpoints/00134... +INFO:tensorflow:global_step/sec: 4.54064 +INFO:tensorflow:step = 52401, loss = 0.162671, precision = 1.0 (22.024 sec) +INFO:tensorflow:global_step/sec: 5.14584 +INFO:tensorflow:step = 52501, loss = 0.230721, precision = 0.976562 (19.433 sec) +INFO:tensorflow:global_step/sec: 5.13962 +INFO:tensorflow:step = 52601, loss = 0.203327, precision = 0.96875 (19.459 sec) +INFO:tensorflow:global_step/sec: 5.13947 +INFO:tensorflow:step = 52701, loss = 0.169935, precision = 0.984375 (19.454 sec) +Saved checkpoint after 135 epoch(s) to data/resnet164_nb/checkpoints/00135... +INFO:tensorflow:global_step/sec: 4.60862 +INFO:tensorflow:step = 52801, loss = 0.154927, precision = 1.0 (21.699 sec) +INFO:tensorflow:global_step/sec: 5.13515 +INFO:tensorflow:step = 52901, loss = 0.302399, precision = 0.945312 (19.473 sec) +INFO:tensorflow:global_step/sec: 5.13689 +INFO:tensorflow:step = 53001, loss = 0.211592, precision = 0.96875 (19.469 sec) +INFO:tensorflow:global_step/sec: 5.14961 +INFO:tensorflow:step = 53101, loss = 0.152747, precision = 1.0 (19.417 sec) +Saved checkpoint after 136 epoch(s) to data/resnet164_nb/checkpoints/00136... +INFO:tensorflow:global_step/sec: 4.58985 +INFO:tensorflow:step = 53201, loss = 0.210721, precision = 0.96875 (21.791 sec) +INFO:tensorflow:global_step/sec: 5.13991 +INFO:tensorflow:step = 53301, loss = 0.170241, precision = 0.984375 (19.452 sec) +INFO:tensorflow:global_step/sec: 5.14039 +INFO:tensorflow:step = 53401, loss = 0.143954, precision = 1.0 (19.454 sec) +INFO:tensorflow:global_step/sec: 5.14972 +INFO:tensorflow:step = 53501, loss = 0.149609, precision = 1.0 (19.423 sec) +Saved checkpoint after 137 epoch(s) to data/resnet164_nb/checkpoints/00137... +INFO:tensorflow:global_step/sec: 4.49569 +INFO:tensorflow:step = 53601, loss = 0.152181, precision = 0.992188 (22.241 sec) +INFO:tensorflow:global_step/sec: 5.14248 +INFO:tensorflow:step = 53701, loss = 0.154057, precision = 0.992188 (19.449 sec) +INFO:tensorflow:global_step/sec: 5.15302 +INFO:tensorflow:step = 53801, loss = 0.156754, precision = 0.992188 (19.401 sec) +INFO:tensorflow:global_step/sec: 5.14509 +INFO:tensorflow:step = 53901, loss = 0.143018, precision = 1.0 (19.436 sec) +Saved checkpoint after 138 epoch(s) to data/resnet164_nb/checkpoints/00138... +INFO:tensorflow:global_step/sec: 4.54148 +INFO:tensorflow:step = 54001, loss = 0.141819, precision = 1.0 (22.023 sec) +INFO:tensorflow:global_step/sec: 5.14969 +INFO:tensorflow:step = 54101, loss = 0.168756, precision = 0.992188 (19.415 sec) +INFO:tensorflow:global_step/sec: 5.13955 +INFO:tensorflow:step = 54201, loss = 0.152279, precision = 0.992188 (19.460 sec) +INFO:tensorflow:global_step/sec: 5.13819 +INFO:tensorflow:step = 54301, loss = 0.139656, precision = 1.0 (19.459 sec) +Saved checkpoint after 139 epoch(s) to data/resnet164_nb/checkpoints/00139... +INFO:tensorflow:global_step/sec: 4.59505 +INFO:tensorflow:step = 54401, loss = 0.13927, precision = 1.0 (21.765 sec) +INFO:tensorflow:global_step/sec: 5.13938 +INFO:tensorflow:step = 54501, loss = 0.142045, precision = 1.0 (19.458 sec) +INFO:tensorflow:global_step/sec: 5.14924 +INFO:tensorflow:step = 54601, loss = 0.139246, precision = 1.0 (19.422 sec) +INFO:tensorflow:global_step/sec: 5.15211 +INFO:tensorflow:step = 54701, loss = 0.137965, precision = 1.0 (19.405 sec) +Saved checkpoint after 140 epoch(s) to data/resnet164_nb/checkpoints/00140... +INFO:tensorflow:global_step/sec: 4.55774 +INFO:tensorflow:step = 54801, loss = 0.142692, precision = 1.0 (21.941 sec) +INFO:tensorflow:global_step/sec: 5.13193 +INFO:tensorflow:step = 54901, loss = 0.137908, precision = 1.0 (19.486 sec) +INFO:tensorflow:global_step/sec: 5.14572 +INFO:tensorflow:step = 55001, loss = 0.138815, precision = 1.0 (19.434 sec) +INFO:tensorflow:global_step/sec: 5.13893 +INFO:tensorflow:step = 55101, loss = 0.140911, precision = 1.0 (19.459 sec) +Saved checkpoint after 141 epoch(s) to data/resnet164_nb/checkpoints/00141... +INFO:tensorflow:global_step/sec: 4.54382 +INFO:tensorflow:step = 55201, loss = 0.145668, precision = 1.0 (22.008 sec) +INFO:tensorflow:global_step/sec: 5.14179 +INFO:tensorflow:step = 55301, loss = 0.149873, precision = 1.0 (19.448 sec) +INFO:tensorflow:global_step/sec: 5.14517 +INFO:tensorflow:step = 55401, loss = 0.147709, precision = 1.0 (19.444 sec) +INFO:tensorflow:global_step/sec: 5.13618 +INFO:tensorflow:step = 55501, loss = 0.140688, precision = 1.0 (19.462 sec) +Saved checkpoint after 142 epoch(s) to data/resnet164_nb/checkpoints/00142... +INFO:tensorflow:global_step/sec: 4.46435 +INFO:tensorflow:step = 55601, loss = 0.17762, precision = 0.984375 (22.400 sec) +INFO:tensorflow:global_step/sec: 5.15014 +INFO:tensorflow:step = 55701, loss = 0.147747, precision = 0.992188 (19.416 sec) +INFO:tensorflow:global_step/sec: 5.15266 +INFO:tensorflow:step = 55801, loss = 0.142017, precision = 1.0 (19.409 sec) +INFO:tensorflow:global_step/sec: 5.15185 +INFO:tensorflow:step = 55901, loss = 0.150425, precision = 0.992188 (19.410 sec) +Saved checkpoint after 143 epoch(s) to data/resnet164_nb/checkpoints/00143... +INFO:tensorflow:global_step/sec: 4.55277 +INFO:tensorflow:step = 56001, loss = 0.153097, precision = 0.992188 (21.965 sec) +INFO:tensorflow:global_step/sec: 5.15421 +INFO:tensorflow:step = 56101, loss = 0.136368, precision = 1.0 (19.401 sec) +INFO:tensorflow:global_step/sec: 5.15304 +INFO:tensorflow:step = 56201, loss = 0.138153, precision = 1.0 (19.406 sec) +INFO:tensorflow:global_step/sec: 5.14974 +INFO:tensorflow:step = 56301, loss = 0.135868, precision = 1.0 (19.419 sec) +Saved checkpoint after 144 epoch(s) to data/resnet164_nb/checkpoints/00144... +INFO:tensorflow:global_step/sec: 4.56885 +INFO:tensorflow:step = 56401, loss = 0.137484, precision = 1.0 (21.889 sec) +INFO:tensorflow:global_step/sec: 5.13043 +INFO:tensorflow:step = 56501, loss = 0.151959, precision = 0.992188 (19.491 sec) +INFO:tensorflow:global_step/sec: 5.14678 +INFO:tensorflow:step = 56601, loss = 0.137425, precision = 1.0 (19.428 sec) +Saved checkpoint after 145 epoch(s) to data/resnet164_nb/checkpoints/00145... +INFO:tensorflow:global_step/sec: 4.59321 +INFO:tensorflow:step = 56701, loss = 0.139773, precision = 1.0 (21.773 sec) +INFO:tensorflow:global_step/sec: 5.13934 +INFO:tensorflow:step = 56801, loss = 0.135171, precision = 1.0 (19.456 sec) +INFO:tensorflow:global_step/sec: 5.13564 +INFO:tensorflow:step = 56901, loss = 0.135114, precision = 1.0 (19.475 sec) +INFO:tensorflow:global_step/sec: 5.15275 +INFO:tensorflow:step = 57001, loss = 0.136261, precision = 1.0 (19.404 sec) +Saved checkpoint after 146 epoch(s) to data/resnet164_nb/checkpoints/00146... +INFO:tensorflow:global_step/sec: 4.58093 +INFO:tensorflow:step = 57101, loss = 0.150979, precision = 0.992188 (21.831 sec) +INFO:tensorflow:global_step/sec: 5.14849 +INFO:tensorflow:step = 57201, loss = 0.142253, precision = 1.0 (19.427 sec) +INFO:tensorflow:global_step/sec: 5.1506 +INFO:tensorflow:step = 57301, loss = 0.137258, precision = 1.0 (19.411 sec) +INFO:tensorflow:global_step/sec: 5.1507 +INFO:tensorflow:step = 57401, loss = 0.137056, precision = 1.0 (19.415 sec) +Saved checkpoint after 147 epoch(s) to data/resnet164_nb/checkpoints/00147... +INFO:tensorflow:global_step/sec: 4.5654 +INFO:tensorflow:step = 57501, loss = 0.141229, precision = 0.992188 (21.906 sec) +INFO:tensorflow:global_step/sec: 5.14615 +INFO:tensorflow:step = 57601, loss = 0.133759, precision = 1.0 (19.431 sec) +INFO:tensorflow:global_step/sec: 5.14662 +INFO:tensorflow:step = 57701, loss = 0.135271, precision = 1.0 (19.430 sec) +INFO:tensorflow:global_step/sec: 5.14637 +INFO:tensorflow:step = 57801, loss = 0.135213, precision = 1.0 (19.434 sec) +Saved checkpoint after 148 epoch(s) to data/resnet164_nb/checkpoints/00148... +INFO:tensorflow:global_step/sec: 4.49154 +INFO:tensorflow:step = 57901, loss = 0.13742, precision = 1.0 (22.264 sec) +INFO:tensorflow:global_step/sec: 5.14779 +INFO:tensorflow:step = 58001, loss = 0.132733, precision = 1.0 (19.424 sec) +INFO:tensorflow:global_step/sec: 5.15106 +INFO:tensorflow:step = 58101, loss = 0.137183, precision = 1.0 (19.418 sec) +INFO:tensorflow:global_step/sec: 5.15144 +INFO:tensorflow:step = 58201, loss = 0.132372, precision = 1.0 (19.406 sec) +Saved checkpoint after 149 epoch(s) to data/resnet164_nb/checkpoints/00149... +INFO:tensorflow:global_step/sec: 4.5658 +INFO:tensorflow:step = 58301, loss = 0.132457, precision = 1.0 (21.902 sec) +INFO:tensorflow:global_step/sec: 5.16065 +INFO:tensorflow:step = 58401, loss = 0.131229, precision = 1.0 (19.378 sec) +INFO:tensorflow:global_step/sec: 5.15363 +INFO:tensorflow:step = 58501, loss = 0.133853, precision = 1.0 (19.403 sec) +INFO:tensorflow:global_step/sec: 5.14984 +INFO:tensorflow:step = 58601, loss = 0.141601, precision = 0.992188 (19.418 sec) +Saved checkpoint after 150 epoch(s) to data/resnet164_nb/checkpoints/00150... +INFO:tensorflow:global_step/sec: 4.57805 +INFO:tensorflow:step = 58701, loss = 0.135249, precision = 1.0 (21.844 sec) +INFO:tensorflow:global_step/sec: 5.15196 +INFO:tensorflow:step = 58801, loss = 0.137342, precision = 1.0 (19.409 sec) +INFO:tensorflow:global_step/sec: 5.1508 +INFO:tensorflow:step = 58901, loss = 0.134934, precision = 1.0 (19.415 sec) +INFO:tensorflow:global_step/sec: 5.14518 +INFO:tensorflow:step = 59001, loss = 0.131591, precision = 1.0 (19.440 sec) +Saved checkpoint after 151 epoch(s) to data/resnet164_nb/checkpoints/00151... +INFO:tensorflow:global_step/sec: 4.55838 +INFO:tensorflow:step = 59101, loss = 0.131498, precision = 1.0 (21.935 sec) +INFO:tensorflow:global_step/sec: 5.14658 +INFO:tensorflow:step = 59201, loss = 0.13108, precision = 1.0 (19.428 sec) +INFO:tensorflow:global_step/sec: 5.15159 +INFO:tensorflow:step = 59301, loss = 0.130996, precision = 1.0 (19.415 sec) +INFO:tensorflow:global_step/sec: 5.15291 +INFO:tensorflow:step = 59401, loss = 0.130358, precision = 1.0 (19.403 sec) +Saved checkpoint after 152 epoch(s) to data/resnet164_nb/checkpoints/00152... +INFO:tensorflow:global_step/sec: 4.55003 +INFO:tensorflow:step = 59501, loss = 0.130526, precision = 1.0 (21.977 sec) +INFO:tensorflow:global_step/sec: 5.14252 +INFO:tensorflow:step = 59601, loss = 0.138153, precision = 1.0 (19.448 sec) +INFO:tensorflow:global_step/sec: 5.13704 +INFO:tensorflow:step = 59701, loss = 0.130074, precision = 1.0 (19.464 sec) +INFO:tensorflow:global_step/sec: 5.14715 +INFO:tensorflow:step = 59801, loss = 0.131057, precision = 1.0 (19.428 sec) +Saved checkpoint after 153 epoch(s) to data/resnet164_nb/checkpoints/00153... +INFO:tensorflow:global_step/sec: 4.55183 +INFO:tensorflow:step = 59901, loss = 0.132079, precision = 1.0 (21.969 sec) +INFO:tensorflow:global_step/sec: 5.14098 +INFO:tensorflow:step = 60001, loss = 0.135162, precision = 1.0 (19.452 sec) +INFO:tensorflow:global_step/sec: 5.13656 +INFO:tensorflow:step = 60101, loss = 0.129928, precision = 1.0 (19.471 sec) +INFO:tensorflow:global_step/sec: 5.14546 +INFO:tensorflow:step = 60201, loss = 0.129247, precision = 1.0 (19.432 sec) +Saved checkpoint after 154 epoch(s) to data/resnet164_nb/checkpoints/00154... +INFO:tensorflow:global_step/sec: 4.63191 +INFO:tensorflow:step = 60301, loss = 0.130139, precision = 1.0 (21.594 sec) +INFO:tensorflow:global_step/sec: 5.1467 +INFO:tensorflow:step = 60401, loss = 0.129538, precision = 1.0 (19.426 sec) +INFO:tensorflow:global_step/sec: 5.14335 +INFO:tensorflow:step = 60501, loss = 0.132484, precision = 1.0 (19.446 sec) +INFO:tensorflow:global_step/sec: 5.14939 +INFO:tensorflow:step = 60601, loss = 0.129117, precision = 1.0 (19.415 sec) +Saved checkpoint after 155 epoch(s) to data/resnet164_nb/checkpoints/00155... +INFO:tensorflow:global_step/sec: 4.59721 +INFO:tensorflow:step = 60701, loss = 0.131652, precision = 1.0 (21.753 sec) +INFO:tensorflow:global_step/sec: 5.14888 +INFO:tensorflow:step = 60801, loss = 0.129092, precision = 1.0 (19.426 sec) +INFO:tensorflow:global_step/sec: 5.13604 +INFO:tensorflow:step = 60901, loss = 0.131596, precision = 1.0 (19.466 sec) +Saved checkpoint after 156 epoch(s) to data/resnet164_nb/checkpoints/00156... +INFO:tensorflow:global_step/sec: 4.60011 +INFO:tensorflow:step = 61001, loss = 0.138812, precision = 0.992188 (21.741 sec) +INFO:tensorflow:global_step/sec: 5.12993 +INFO:tensorflow:step = 61101, loss = 0.136344, precision = 0.992188 (19.491 sec) +INFO:tensorflow:global_step/sec: 5.13283 +INFO:tensorflow:step = 61201, loss = 0.148715, precision = 0.992188 (19.488 sec) +INFO:tensorflow:global_step/sec: 5.14131 +INFO:tensorflow:step = 61301, loss = 0.14642, precision = 0.992188 (19.445 sec) +Saved checkpoint after 157 epoch(s) to data/resnet164_nb/checkpoints/00157... +INFO:tensorflow:global_step/sec: 4.57896 +INFO:tensorflow:step = 61401, loss = 0.131841, precision = 1.0 (21.839 sec) +INFO:tensorflow:global_step/sec: 5.14564 +INFO:tensorflow:step = 61501, loss = 0.129693, precision = 1.0 (19.436 sec) +INFO:tensorflow:global_step/sec: 5.14712 +INFO:tensorflow:step = 61601, loss = 0.135256, precision = 1.0 (19.426 sec) +INFO:tensorflow:global_step/sec: 5.14718 +INFO:tensorflow:step = 61701, loss = 0.129135, precision = 1.0 (19.428 sec) +Saved checkpoint after 158 epoch(s) to data/resnet164_nb/checkpoints/00158... +INFO:tensorflow:global_step/sec: 4.59347 +INFO:tensorflow:step = 61801, loss = 0.127918, precision = 1.0 (21.774 sec) +INFO:tensorflow:global_step/sec: 5.14657 +INFO:tensorflow:step = 61901, loss = 0.128229, precision = 1.0 (19.428 sec) +INFO:tensorflow:global_step/sec: 5.14737 +INFO:tensorflow:step = 62001, loss = 0.128089, precision = 1.0 (19.430 sec) +INFO:tensorflow:global_step/sec: 5.15046 +INFO:tensorflow:step = 62101, loss = 0.128066, precision = 1.0 (19.412 sec) +Saved checkpoint after 159 epoch(s) to data/resnet164_nb/checkpoints/00159... +INFO:tensorflow:global_step/sec: 4.5619 +INFO:tensorflow:step = 62201, loss = 0.129534, precision = 1.0 (21.921 sec) +INFO:tensorflow:global_step/sec: 5.13903 +INFO:tensorflow:step = 62301, loss = 0.140981, precision = 0.992188 (19.459 sec) +INFO:tensorflow:global_step/sec: 5.15045 +INFO:tensorflow:step = 62401, loss = 0.127998, precision = 1.0 (19.415 sec) +INFO:tensorflow:global_step/sec: 5.15175 +INFO:tensorflow:step = 62501, loss = 0.127298, precision = 1.0 (19.411 sec) +Saved checkpoint after 160 epoch(s) to data/resnet164_nb/checkpoints/00160... +INFO:tensorflow:global_step/sec: 4.58075 +INFO:tensorflow:step = 62601, loss = 0.127331, precision = 1.0 (21.830 sec) +INFO:tensorflow:global_step/sec: 5.14427 +INFO:tensorflow:step = 62701, loss = 0.128603, precision = 1.0 (19.439 sec) +INFO:tensorflow:global_step/sec: 5.14428 +INFO:tensorflow:step = 62801, loss = 0.128584, precision = 1.0 (19.439 sec) +INFO:tensorflow:global_step/sec: 5.13888 +INFO:tensorflow:step = 62901, loss = 0.128555, precision = 1.0 (19.460 sec) +Saved checkpoint after 161 epoch(s) to data/resnet164_nb/checkpoints/00161... +INFO:tensorflow:global_step/sec: 4.52035 +INFO:tensorflow:step = 63001, loss = 0.126872, precision = 1.0 (22.122 sec) +INFO:tensorflow:global_step/sec: 5.14174 +INFO:tensorflow:step = 63101, loss = 0.135697, precision = 0.992188 (19.449 sec) +INFO:tensorflow:global_step/sec: 5.14296 +INFO:tensorflow:step = 63201, loss = 0.126068, precision = 1.0 (19.451 sec) +INFO:tensorflow:global_step/sec: 5.13198 +INFO:tensorflow:step = 63301, loss = 0.126469, precision = 1.0 (19.478 sec) +Saved checkpoint after 162 epoch(s) to data/resnet164_nb/checkpoints/00162... +INFO:tensorflow:global_step/sec: 4.549 +INFO:tensorflow:step = 63401, loss = 0.126045, precision = 1.0 (21.987 sec) +INFO:tensorflow:global_step/sec: 5.14343 +INFO:tensorflow:step = 63501, loss = 0.125549, precision = 1.0 (19.442 sec) +INFO:tensorflow:global_step/sec: 5.14089 +INFO:tensorflow:step = 63601, loss = 0.127586, precision = 1.0 (19.449 sec) +INFO:tensorflow:global_step/sec: 5.14159 +INFO:tensorflow:step = 63701, loss = 0.128913, precision = 1.0 (19.449 sec) +Saved checkpoint after 163 epoch(s) to data/resnet164_nb/checkpoints/00163... +INFO:tensorflow:global_step/sec: 4.58493 +INFO:tensorflow:step = 63801, loss = 0.125737, precision = 1.0 (21.811 sec) +INFO:tensorflow:global_step/sec: 5.14435 +INFO:tensorflow:step = 63901, loss = 0.127943, precision = 1.0 (19.438 sec) +INFO:tensorflow:global_step/sec: 5.14589 +INFO:tensorflow:step = 64001, loss = 0.140967, precision = 0.992188 (19.442 sec) +INFO:tensorflow:global_step/sec: 5.14584 +INFO:tensorflow:step = 64101, loss = 0.128733, precision = 1.0 (19.424 sec) +Saved checkpoint after 164 epoch(s) to data/resnet164_nb/checkpoints/00164... +INFO:tensorflow:global_step/sec: 4.56548 +INFO:tensorflow:step = 64201, loss = 0.125871, precision = 1.0 (21.903 sec) +INFO:tensorflow:global_step/sec: 5.14951 +INFO:tensorflow:step = 64301, loss = 0.124548, precision = 1.0 (19.419 sec) +INFO:tensorflow:global_step/sec: 5.1545 +INFO:tensorflow:step = 64401, loss = 0.125932, precision = 1.0 (19.401 sec) +INFO:tensorflow:global_step/sec: 5.14947 +INFO:tensorflow:step = 64501, loss = 0.12441, precision = 1.0 (19.419 sec) +Saved checkpoint after 165 epoch(s) to data/resnet164_nb/checkpoints/00165... +INFO:tensorflow:global_step/sec: 4.589 +INFO:tensorflow:step = 64601, loss = 0.12388, precision = 1.0 (21.791 sec) +INFO:tensorflow:global_step/sec: 5.15293 +INFO:tensorflow:step = 64701, loss = 0.124356, precision = 1.0 (19.406 sec) +INFO:tensorflow:global_step/sec: 5.14449 +INFO:tensorflow:step = 64801, loss = 0.125245, precision = 1.0 (19.438 sec) +INFO:tensorflow:global_step/sec: 5.14192 +INFO:tensorflow:step = 64901, loss = 0.12477, precision = 1.0 (19.448 sec) +Saved checkpoint after 166 epoch(s) to data/resnet164_nb/checkpoints/00166... +INFO:tensorflow:global_step/sec: 4.59451 +INFO:tensorflow:step = 65001, loss = 0.12392, precision = 1.0 (21.765 sec) +INFO:tensorflow:global_step/sec: 5.15287 +INFO:tensorflow:step = 65101, loss = 0.124413, precision = 1.0 (19.406 sec) +INFO:tensorflow:global_step/sec: 5.15113 +INFO:tensorflow:step = 65201, loss = 0.123074, precision = 1.0 (19.416 sec) +Saved checkpoint after 167 epoch(s) to data/resnet164_nb/checkpoints/00167... +INFO:tensorflow:global_step/sec: 4.55421 +INFO:tensorflow:step = 65301, loss = 0.124089, precision = 1.0 (21.955 sec) +INFO:tensorflow:global_step/sec: 5.14461 +INFO:tensorflow:step = 65401, loss = 0.123215, precision = 1.0 (19.438 sec) +INFO:tensorflow:global_step/sec: 5.1363 +INFO:tensorflow:step = 65501, loss = 0.123696, precision = 1.0 (19.471 sec) +INFO:tensorflow:global_step/sec: 5.14712 +INFO:tensorflow:step = 65601, loss = 0.123549, precision = 1.0 (19.426 sec) +Saved checkpoint after 168 epoch(s) to data/resnet164_nb/checkpoints/00168... +INFO:tensorflow:global_step/sec: 4.59007 +INFO:tensorflow:step = 65701, loss = 0.123474, precision = 1.0 (21.788 sec) +INFO:tensorflow:global_step/sec: 5.14118 +INFO:tensorflow:step = 65801, loss = 0.122476, precision = 1.0 (19.448 sec) +INFO:tensorflow:global_step/sec: 5.15162 +INFO:tensorflow:step = 65901, loss = 0.123227, precision = 1.0 (19.412 sec) +INFO:tensorflow:global_step/sec: 5.15322 +INFO:tensorflow:step = 66001, loss = 0.123056, precision = 1.0 (19.405 sec) +Saved checkpoint after 169 epoch(s) to data/resnet164_nb/checkpoints/00169... +INFO:tensorflow:global_step/sec: 4.54828 +INFO:tensorflow:step = 66101, loss = 0.123347, precision = 1.0 (21.987 sec) +INFO:tensorflow:global_step/sec: 5.13727 +INFO:tensorflow:step = 66201, loss = 0.121961, precision = 1.0 (19.465 sec) +INFO:tensorflow:global_step/sec: 5.1482 +INFO:tensorflow:step = 66301, loss = 0.122791, precision = 1.0 (19.424 sec) +INFO:tensorflow:global_step/sec: 5.14793 +INFO:tensorflow:step = 66401, loss = 0.12505, precision = 1.0 (19.425 sec) +Saved checkpoint after 170 epoch(s) to data/resnet164_nb/checkpoints/00170... +INFO:tensorflow:global_step/sec: 4.55221 +INFO:tensorflow:step = 66501, loss = 0.12475, precision = 1.0 (21.973 sec) +INFO:tensorflow:global_step/sec: 5.13032 +INFO:tensorflow:step = 66601, loss = 0.122434, precision = 1.0 (19.488 sec) +INFO:tensorflow:global_step/sec: 5.14198 +INFO:tensorflow:step = 66701, loss = 0.12135, precision = 1.0 (19.446 sec) +INFO:tensorflow:global_step/sec: 5.14613 +INFO:tensorflow:step = 66801, loss = 0.126786, precision = 1.0 (19.437 sec) +Saved checkpoint after 171 epoch(s) to data/resnet164_nb/checkpoints/00171... +INFO:tensorflow:global_step/sec: 4.59736 +INFO:tensorflow:step = 66901, loss = 0.12443, precision = 1.0 (21.749 sec) +INFO:tensorflow:global_step/sec: 5.1449 +INFO:tensorflow:step = 67001, loss = 0.121485, precision = 1.0 (19.436 sec) +INFO:tensorflow:global_step/sec: 5.14254 +INFO:tensorflow:step = 67101, loss = 0.121476, precision = 1.0 (19.446 sec) +INFO:tensorflow:global_step/sec: 5.13591 +INFO:tensorflow:step = 67201, loss = 0.124901, precision = 1.0 (19.469 sec) +Saved checkpoint after 172 epoch(s) to data/resnet164_nb/checkpoints/00172... +INFO:tensorflow:global_step/sec: 4.49312 +INFO:tensorflow:step = 67301, loss = 0.121417, precision = 1.0 (22.258 sec) +INFO:tensorflow:global_step/sec: 5.13877 +INFO:tensorflow:step = 67401, loss = 0.121062, precision = 1.0 (19.461 sec) +INFO:tensorflow:global_step/sec: 5.1459 +INFO:tensorflow:step = 67501, loss = 0.12191, precision = 1.0 (19.429 sec) +INFO:tensorflow:global_step/sec: 5.13976 +INFO:tensorflow:step = 67601, loss = 0.12179, precision = 1.0 (19.457 sec) +Saved checkpoint after 173 epoch(s) to data/resnet164_nb/checkpoints/00173... +INFO:tensorflow:global_step/sec: 4.5956 +INFO:tensorflow:step = 67701, loss = 0.120661, precision = 1.0 (21.759 sec) +INFO:tensorflow:global_step/sec: 5.15266 +INFO:tensorflow:step = 67801, loss = 0.119871, precision = 1.0 (19.408 sec) +INFO:tensorflow:global_step/sec: 5.15334 +INFO:tensorflow:step = 67901, loss = 0.124361, precision = 1.0 (19.404 sec) +INFO:tensorflow:global_step/sec: 5.15002 +INFO:tensorflow:step = 68001, loss = 0.12156, precision = 1.0 (19.418 sec) +Saved checkpoint after 174 epoch(s) to data/resnet164_nb/checkpoints/00174... +INFO:tensorflow:global_step/sec: 4.5395 +INFO:tensorflow:step = 68101, loss = 0.12167, precision = 1.0 (22.032 sec) +INFO:tensorflow:global_step/sec: 5.15015 +INFO:tensorflow:step = 68201, loss = 0.1225, precision = 1.0 (19.415 sec) +INFO:tensorflow:global_step/sec: 5.15009 +INFO:tensorflow:step = 68301, loss = 0.121883, precision = 1.0 (19.415 sec) +INFO:tensorflow:global_step/sec: 5.1511 +INFO:tensorflow:step = 68401, loss = 0.120996, precision = 1.0 (19.414 sec) +Saved checkpoint after 175 epoch(s) to data/resnet164_nb/checkpoints/00175... +INFO:tensorflow:global_step/sec: 4.59199 +INFO:tensorflow:step = 68501, loss = 0.119815, precision = 1.0 (21.777 sec) +INFO:tensorflow:global_step/sec: 5.15295 +INFO:tensorflow:step = 68601, loss = 0.119106, precision = 1.0 (19.409 sec) +INFO:tensorflow:global_step/sec: 5.14745 +INFO:tensorflow:step = 68701, loss = 0.119442, precision = 1.0 (19.424 sec) +INFO:tensorflow:global_step/sec: 5.14872 +INFO:tensorflow:step = 68801, loss = 0.119833, precision = 1.0 (19.422 sec) +Saved checkpoint after 176 epoch(s) to data/resnet164_nb/checkpoints/00176... +INFO:tensorflow:global_step/sec: 4.58037 +INFO:tensorflow:step = 68901, loss = 0.119, precision = 1.0 (21.833 sec) +INFO:tensorflow:global_step/sec: 5.15379 +INFO:tensorflow:step = 69001, loss = 0.120977, precision = 1.0 (19.406 sec) +INFO:tensorflow:global_step/sec: 5.15409 +INFO:tensorflow:step = 69101, loss = 0.120075, precision = 1.0 (19.399 sec) +INFO:tensorflow:global_step/sec: 5.15002 +INFO:tensorflow:step = 69201, loss = 0.123374, precision = 1.0 (19.417 sec) +Saved checkpoint after 177 epoch(s) to data/resnet164_nb/checkpoints/00177... +INFO:tensorflow:global_step/sec: 4.59373 +INFO:tensorflow:step = 69301, loss = 0.119011, precision = 1.0 (21.769 sec) +INFO:tensorflow:global_step/sec: 5.14453 +INFO:tensorflow:step = 69401, loss = 0.120969, precision = 1.0 (19.438 sec) +INFO:tensorflow:global_step/sec: 5.13698 +INFO:tensorflow:step = 69501, loss = 0.120648, precision = 1.0 (19.468 sec) +Saved checkpoint after 178 epoch(s) to data/resnet164_nb/checkpoints/00178... +INFO:tensorflow:global_step/sec: 4.50831 +INFO:tensorflow:step = 69601, loss = 0.118242, precision = 1.0 (22.179 sec) +INFO:tensorflow:global_step/sec: 5.13945 +INFO:tensorflow:step = 69701, loss = 0.117839, precision = 1.0 (19.462 sec) +INFO:tensorflow:global_step/sec: 5.14344 +INFO:tensorflow:step = 69801, loss = 0.123822, precision = 0.992188 (19.440 sec) +INFO:tensorflow:global_step/sec: 5.14085 +INFO:tensorflow:step = 69901, loss = 0.123647, precision = 1.0 (19.450 sec) +Saved checkpoint after 179 epoch(s) to data/resnet164_nb/checkpoints/00179... +INFO:tensorflow:global_step/sec: 4.54621 +INFO:tensorflow:step = 70001, loss = 0.119829, precision = 1.0 (21.996 sec) +INFO:tensorflow:global_step/sec: 5.14906 +INFO:tensorflow:step = 70101, loss = 0.118004, precision = 1.0 (19.421 sec) +INFO:tensorflow:global_step/sec: 5.15393 +INFO:tensorflow:step = 70201, loss = 0.118002, precision = 1.0 (19.402 sec) +INFO:tensorflow:global_step/sec: 5.1557 +INFO:tensorflow:step = 70301, loss = 0.117543, precision = 1.0 (19.396 sec) +Saved checkpoint after 180 epoch(s) to data/resnet164_nb/checkpoints/00180... +INFO:tensorflow:global_step/sec: 4.53676 +INFO:tensorflow:step = 70401, loss = 0.117537, precision = 1.0 (22.042 sec) +INFO:tensorflow:global_step/sec: 5.13809 +INFO:tensorflow:step = 70501, loss = 0.118386, precision = 1.0 (19.463 sec) +INFO:tensorflow:global_step/sec: 5.14901 +INFO:tensorflow:step = 70601, loss = 0.117629, precision = 1.0 (19.421 sec) +INFO:tensorflow:global_step/sec: 5.14448 +INFO:tensorflow:step = 70701, loss = 0.117888, precision = 1.0 (19.439 sec) +Saved checkpoint after 181 epoch(s) to data/resnet164_nb/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1p100_dawn/resnet20_train.log b/tensorflow/CIFAR10/logs/1p100_dawn/resnet20_train.log new file mode 100644 index 0000000..f7870bd --- /dev/null +++ b/tensorflow/CIFAR10/logs/1p100_dawn/resnet20_train.log @@ -0,0 +1,1727 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/269.03k params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/10.38b flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + +======================End of Report========================== +2017-08-03 00:02:32.693331: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla P100-PCIE-16GB +major: 6 minor: 0 memoryClockRate (GHz) 1.3285 +pciBusID 0000:05:00.0 +Total memory: 15.89GiB +Free memory: 15.61GiB +2017-08-03 00:02:32.693378: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-08-03 00:02:32.693385: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-08-03 00:02:32.693394: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:05:00.0) +2017-08-03 00:02:33.216116: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-08-03 00:02:33.216204: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 56 visible devices +2017-08-03 00:02:33.240129: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x40192c0 executing computations on platform Host. Devices: +2017-08-03 00:02:33.240215: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-08-03 00:02:33.240746: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-08-03 00:02:33.240769: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 56 visible devices +2017-08-03 00:02:33.260758: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x3fc7940 executing computations on platform CUDA. Devices: +2017-08-03 00:02:33.260801: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla P100-PCIE-16GB, Compute Capability 6.0 +INFO:tensorflow:step = 1, loss = 2.68061, precision = 0.09375 +INFO:tensorflow:global_step/sec: 33.9868 +INFO:tensorflow:step = 101, loss = 2.00101, precision = 0.367188 (2.943 sec) +INFO:tensorflow:global_step/sec: 37.1699 +INFO:tensorflow:step = 201, loss = 1.99259, precision = 0.429688 (2.691 sec) +INFO:tensorflow:global_step/sec: 37.2534 +INFO:tensorflow:step = 301, loss = 1.91987, precision = 0.351562 (2.685 sec) +total_params: 269034 +Saved checkpoint after 1 epoch(s) to data/resnet20/checkpoints/00001... +INFO:tensorflow:global_step/sec: 31.2056 +INFO:tensorflow:step = 401, loss = 1.95698, precision = 0.367188 (3.204 sec) +INFO:tensorflow:global_step/sec: 36.7379 +INFO:tensorflow:step = 501, loss = 1.88766, precision = 0.40625 (2.722 sec) +INFO:tensorflow:global_step/sec: 36.6457 +INFO:tensorflow:step = 601, loss = 1.52655, precision = 0.53125 (2.729 sec) +INFO:tensorflow:global_step/sec: 36.5714 +INFO:tensorflow:step = 701, loss = 1.61445, precision = 0.492188 (2.734 sec) +Saved checkpoint after 2 epoch(s) to data/resnet20/checkpoints/00002... +INFO:tensorflow:global_step/sec: 31.3376 +INFO:tensorflow:step = 801, loss = 1.31947, precision = 0.609375 (3.191 sec) +INFO:tensorflow:global_step/sec: 36.2171 +INFO:tensorflow:step = 901, loss = 1.23731, precision = 0.664062 (2.761 sec) +INFO:tensorflow:global_step/sec: 36.507 +INFO:tensorflow:step = 1001, loss = 1.03648, precision = 0.742188 (2.739 sec) +INFO:tensorflow:global_step/sec: 36.5595 +INFO:tensorflow:step = 1101, loss = 1.13547, precision = 0.6875 (2.735 sec) +Saved checkpoint after 3 epoch(s) to data/resnet20/checkpoints/00003... +INFO:tensorflow:global_step/sec: 30.9625 +INFO:tensorflow:step = 1201, loss = 1.04044, precision = 0.71875 (3.230 sec) +INFO:tensorflow:global_step/sec: 36.1378 +INFO:tensorflow:step = 1301, loss = 0.988597, precision = 0.734375 (2.768 sec) +INFO:tensorflow:global_step/sec: 36.2954 +INFO:tensorflow:step = 1401, loss = 1.0626, precision = 0.757812 (2.755 sec) +INFO:tensorflow:global_step/sec: 36.3509 +INFO:tensorflow:step = 1501, loss = 1.05861, precision = 0.71875 (2.751 sec) +Saved checkpoint after 4 epoch(s) to data/resnet20/checkpoints/00004... +INFO:tensorflow:global_step/sec: 30.4896 +INFO:tensorflow:step = 1601, loss = 0.909604, precision = 0.773438 (3.280 sec) +INFO:tensorflow:global_step/sec: 36.3416 +INFO:tensorflow:step = 1701, loss = 0.796503, precision = 0.796875 (2.752 sec) +INFO:tensorflow:global_step/sec: 36.3363 +INFO:tensorflow:step = 1801, loss = 0.842827, precision = 0.78125 (2.752 sec) +INFO:tensorflow:global_step/sec: 36.5074 +INFO:tensorflow:step = 1901, loss = 0.917468, precision = 0.75 (2.739 sec) +Saved checkpoint after 5 epoch(s) to data/resnet20/checkpoints/00005... +INFO:tensorflow:global_step/sec: 30.923 +INFO:tensorflow:step = 2001, loss = 0.888184, precision = 0.757812 (3.234 sec) +INFO:tensorflow:global_step/sec: 36.4091 +INFO:tensorflow:step = 2101, loss = 0.790183, precision = 0.804688 (2.746 sec) +INFO:tensorflow:global_step/sec: 36.5253 +INFO:tensorflow:step = 2201, loss = 0.869925, precision = 0.757812 (2.738 sec) +INFO:tensorflow:global_step/sec: 36.6024 +INFO:tensorflow:step = 2301, loss = 0.962855, precision = 0.75 (2.732 sec) +Saved checkpoint after 6 epoch(s) to data/resnet20/checkpoints/00006... +INFO:tensorflow:global_step/sec: 31.2621 +INFO:tensorflow:step = 2401, loss = 0.796132, precision = 0.789062 (3.199 sec) +INFO:tensorflow:global_step/sec: 36.7405 +INFO:tensorflow:step = 2501, loss = 0.786942, precision = 0.8125 (2.722 sec) +INFO:tensorflow:global_step/sec: 36.3167 +INFO:tensorflow:step = 2601, loss = 0.912961, precision = 0.742188 (2.753 sec) +INFO:tensorflow:global_step/sec: 36.5684 +INFO:tensorflow:step = 2701, loss = 0.719123, precision = 0.835938 (2.735 sec) +Saved checkpoint after 7 epoch(s) to data/resnet20/checkpoints/00007... +INFO:tensorflow:global_step/sec: 31.3954 +INFO:tensorflow:step = 2801, loss = 0.869238, precision = 0.75 (3.186 sec) +INFO:tensorflow:global_step/sec: 36.7444 +INFO:tensorflow:step = 2901, loss = 1.01009, precision = 0.703125 (2.721 sec) +INFO:tensorflow:global_step/sec: 36.4738 +INFO:tensorflow:step = 3001, loss = 0.845218, precision = 0.789062 (2.742 sec) +INFO:tensorflow:global_step/sec: 36.2063 +INFO:tensorflow:step = 3101, loss = 0.797981, precision = 0.796875 (2.762 sec) +Saved checkpoint after 8 epoch(s) to data/resnet20/checkpoints/00008... +INFO:tensorflow:global_step/sec: 30.7306 +INFO:tensorflow:step = 3201, loss = 0.730514, precision = 0.835938 (3.254 sec) +INFO:tensorflow:global_step/sec: 36.6362 +INFO:tensorflow:step = 3301, loss = 0.849458, precision = 0.757812 (2.730 sec) +INFO:tensorflow:global_step/sec: 36.5824 +INFO:tensorflow:step = 3401, loss = 0.798813, precision = 0.796875 (2.733 sec) +INFO:tensorflow:global_step/sec: 36.7809 +INFO:tensorflow:step = 3501, loss = 0.791147, precision = 0.796875 (2.719 sec) +Saved checkpoint after 9 epoch(s) to data/resnet20/checkpoints/00009... +INFO:tensorflow:global_step/sec: 31.548 +INFO:tensorflow:step = 3601, loss = 0.698246, precision = 0.875 (3.170 sec) +INFO:tensorflow:global_step/sec: 36.8433 +INFO:tensorflow:step = 3701, loss = 0.743351, precision = 0.820312 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.7011 +INFO:tensorflow:step = 3801, loss = 0.847044, precision = 0.804688 (2.725 sec) +INFO:tensorflow:global_step/sec: 36.5004 +INFO:tensorflow:step = 3901, loss = 0.733214, precision = 0.8125 (2.740 sec) +Saved checkpoint after 10 epoch(s) to data/resnet20/checkpoints/00010... +INFO:tensorflow:global_step/sec: 31.2314 +INFO:tensorflow:step = 4001, loss = 0.747592, precision = 0.804688 (3.202 sec) +INFO:tensorflow:global_step/sec: 36.714 +INFO:tensorflow:step = 4101, loss = 0.725167, precision = 0.859375 (2.723 sec) +INFO:tensorflow:global_step/sec: 36.594 +INFO:tensorflow:step = 4201, loss = 0.734989, precision = 0.835938 (2.733 sec) +Saved checkpoint after 11 epoch(s) to data/resnet20/checkpoints/00011... +INFO:tensorflow:global_step/sec: 31.8402 +INFO:tensorflow:step = 4301, loss = 0.626409, precision = 0.859375 (3.140 sec) +INFO:tensorflow:global_step/sec: 36.3565 +INFO:tensorflow:step = 4401, loss = 0.647716, precision = 0.84375 (2.751 sec) +INFO:tensorflow:global_step/sec: 36.6307 +INFO:tensorflow:step = 4501, loss = 0.720233, precision = 0.859375 (2.730 sec) +INFO:tensorflow:global_step/sec: 36.7153 +INFO:tensorflow:step = 4601, loss = 0.862528, precision = 0.765625 (2.724 sec) +Saved checkpoint after 12 epoch(s) to data/resnet20/checkpoints/00012... +INFO:tensorflow:global_step/sec: 31.3233 +INFO:tensorflow:step = 4701, loss = 0.816589, precision = 0.773438 (3.193 sec) +INFO:tensorflow:global_step/sec: 36.7496 +INFO:tensorflow:step = 4801, loss = 0.71257, precision = 0.867188 (2.721 sec) +INFO:tensorflow:global_step/sec: 36.8338 +INFO:tensorflow:step = 4901, loss = 0.722708, precision = 0.828125 (2.715 sec) +INFO:tensorflow:global_step/sec: 36.7084 +INFO:tensorflow:step = 5001, loss = 0.75995, precision = 0.8125 (2.724 sec) +Saved checkpoint after 13 epoch(s) to data/resnet20/checkpoints/00013... +INFO:tensorflow:global_step/sec: 31.1766 +INFO:tensorflow:step = 5101, loss = 0.810508, precision = 0.804688 (3.207 sec) +INFO:tensorflow:global_step/sec: 36.6017 +INFO:tensorflow:step = 5201, loss = 0.7905, precision = 0.835938 (2.732 sec) +INFO:tensorflow:global_step/sec: 36.4243 +INFO:tensorflow:step = 5301, loss = 0.667147, precision = 0.828125 (2.745 sec) +INFO:tensorflow:global_step/sec: 36.7766 +INFO:tensorflow:step = 5401, loss = 0.720591, precision = 0.820312 (2.719 sec) +Saved checkpoint after 14 epoch(s) to data/resnet20/checkpoints/00014... +INFO:tensorflow:global_step/sec: 31.2719 +INFO:tensorflow:step = 5501, loss = 0.6348, precision = 0.875 (3.198 sec) +INFO:tensorflow:global_step/sec: 36.5938 +INFO:tensorflow:step = 5601, loss = 0.697133, precision = 0.851562 (2.733 sec) +INFO:tensorflow:global_step/sec: 36.6123 +INFO:tensorflow:step = 5701, loss = 0.726616, precision = 0.8125 (2.731 sec) +INFO:tensorflow:global_step/sec: 36.6884 +INFO:tensorflow:step = 5801, loss = 0.667568, precision = 0.828125 (2.726 sec) +Saved checkpoint after 15 epoch(s) to data/resnet20/checkpoints/00015... +INFO:tensorflow:global_step/sec: 31.4989 +INFO:tensorflow:step = 5901, loss = 0.634159, precision = 0.859375 (3.175 sec) +INFO:tensorflow:global_step/sec: 36.6315 +INFO:tensorflow:step = 6001, loss = 0.677902, precision = 0.8125 (2.730 sec) +INFO:tensorflow:global_step/sec: 36.5564 +INFO:tensorflow:step = 6101, loss = 0.549132, precision = 0.890625 (2.735 sec) +INFO:tensorflow:global_step/sec: 36.4447 +INFO:tensorflow:step = 6201, loss = 0.682414, precision = 0.84375 (2.744 sec) +Saved checkpoint after 16 epoch(s) to data/resnet20/checkpoints/00016... +INFO:tensorflow:global_step/sec: 31.2156 +INFO:tensorflow:step = 6301, loss = 0.789985, precision = 0.8125 (3.204 sec) +INFO:tensorflow:global_step/sec: 36.6698 +INFO:tensorflow:step = 6401, loss = 0.721547, precision = 0.84375 (2.727 sec) +INFO:tensorflow:global_step/sec: 36.9489 +INFO:tensorflow:step = 6501, loss = 0.667445, precision = 0.859375 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.4158 +INFO:tensorflow:step = 6601, loss = 0.773853, precision = 0.8125 (2.746 sec) +Saved checkpoint after 17 epoch(s) to data/resnet20/checkpoints/00017... +INFO:tensorflow:global_step/sec: 30.7667 +INFO:tensorflow:step = 6701, loss = 0.740369, precision = 0.8125 (3.250 sec) +INFO:tensorflow:global_step/sec: 36.3825 +INFO:tensorflow:step = 6801, loss = 0.76212, precision = 0.8125 (2.748 sec) +INFO:tensorflow:global_step/sec: 36.7444 +INFO:tensorflow:step = 6901, loss = 0.839481, precision = 0.796875 (2.721 sec) +INFO:tensorflow:global_step/sec: 36.8777 +INFO:tensorflow:step = 7001, loss = 0.698426, precision = 0.804688 (2.712 sec) +Saved checkpoint after 18 epoch(s) to data/resnet20/checkpoints/00018... +INFO:tensorflow:global_step/sec: 30.7202 +INFO:tensorflow:step = 7101, loss = 0.825533, precision = 0.796875 (3.255 sec) +INFO:tensorflow:global_step/sec: 36.3624 +INFO:tensorflow:step = 7201, loss = 0.676471, precision = 0.820312 (2.750 sec) +INFO:tensorflow:global_step/sec: 36.9389 +INFO:tensorflow:step = 7301, loss = 0.649459, precision = 0.882812 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.4279 +INFO:tensorflow:step = 7401, loss = 0.734944, precision = 0.835938 (2.745 sec) +Saved checkpoint after 19 epoch(s) to data/resnet20/checkpoints/00019... +INFO:tensorflow:global_step/sec: 30.8197 +INFO:tensorflow:step = 7501, loss = 0.662699, precision = 0.875 (3.245 sec) +INFO:tensorflow:global_step/sec: 36.5571 +INFO:tensorflow:step = 7601, loss = 0.700586, precision = 0.820312 (2.735 sec) +INFO:tensorflow:global_step/sec: 36.8056 +INFO:tensorflow:step = 7701, loss = 0.706559, precision = 0.84375 (2.717 sec) +INFO:tensorflow:global_step/sec: 36.5638 +INFO:tensorflow:step = 7801, loss = 0.576616, precision = 0.914062 (2.735 sec) +Saved checkpoint after 20 epoch(s) to data/resnet20/checkpoints/00020... +INFO:tensorflow:global_step/sec: 30.4308 +INFO:tensorflow:step = 7901, loss = 0.872471, precision = 0.820312 (3.286 sec) +INFO:tensorflow:global_step/sec: 36.5192 +INFO:tensorflow:step = 8001, loss = 0.597721, precision = 0.875 (2.738 sec) +INFO:tensorflow:global_step/sec: 36.6844 +INFO:tensorflow:step = 8101, loss = 0.661379, precision = 0.867188 (2.726 sec) +INFO:tensorflow:global_step/sec: 36.7378 +INFO:tensorflow:step = 8201, loss = 0.803072, precision = 0.804688 (2.722 sec) +Saved checkpoint after 21 epoch(s) to data/resnet20/checkpoints/00021... +INFO:tensorflow:global_step/sec: 31.6265 +INFO:tensorflow:step = 8301, loss = 0.730877, precision = 0.828125 (3.162 sec) +INFO:tensorflow:global_step/sec: 36.5383 +INFO:tensorflow:step = 8401, loss = 1.00679, precision = 0.75 (2.737 sec) +INFO:tensorflow:global_step/sec: 36.7523 +INFO:tensorflow:step = 8501, loss = 0.704638, precision = 0.804688 (2.721 sec) +INFO:tensorflow:global_step/sec: 36.7394 +INFO:tensorflow:step = 8601, loss = 0.589934, precision = 0.84375 (2.722 sec) +Saved checkpoint after 22 epoch(s) to data/resnet20/checkpoints/00022... +INFO:tensorflow:global_step/sec: 30.9159 +INFO:tensorflow:step = 8701, loss = 0.570267, precision = 0.90625 (3.235 sec) +INFO:tensorflow:global_step/sec: 36.8638 +INFO:tensorflow:step = 8801, loss = 0.791721, precision = 0.84375 (2.713 sec) +INFO:tensorflow:global_step/sec: 36.495 +INFO:tensorflow:step = 8901, loss = 0.694525, precision = 0.835938 (2.740 sec) +Saved checkpoint after 23 epoch(s) to data/resnet20/checkpoints/00023... +INFO:tensorflow:global_step/sec: 31.4015 +INFO:tensorflow:step = 9001, loss = 0.637058, precision = 0.851562 (3.185 sec) +INFO:tensorflow:global_step/sec: 36.5714 +INFO:tensorflow:step = 9101, loss = 0.673616, precision = 0.84375 (2.734 sec) +INFO:tensorflow:global_step/sec: 36.925 +INFO:tensorflow:step = 9201, loss = 0.733546, precision = 0.804688 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.5401 +INFO:tensorflow:step = 9301, loss = 0.682018, precision = 0.835938 (2.737 sec) +Saved checkpoint after 24 epoch(s) to data/resnet20/checkpoints/00024... +INFO:tensorflow:global_step/sec: 30.9373 +INFO:tensorflow:step = 9401, loss = 0.734275, precision = 0.859375 (3.232 sec) +INFO:tensorflow:global_step/sec: 36.3707 +INFO:tensorflow:step = 9501, loss = 0.613025, precision = 0.859375 (2.750 sec) +INFO:tensorflow:global_step/sec: 36.7105 +INFO:tensorflow:step = 9601, loss = 0.599912, precision = 0.859375 (2.724 sec) +INFO:tensorflow:global_step/sec: 36.5887 +INFO:tensorflow:step = 9701, loss = 0.707193, precision = 0.820312 (2.733 sec) +Saved checkpoint after 25 epoch(s) to data/resnet20/checkpoints/00025... +INFO:tensorflow:global_step/sec: 31.8239 +INFO:tensorflow:step = 9801, loss = 0.645722, precision = 0.875 (3.142 sec) +INFO:tensorflow:global_step/sec: 36.854 +INFO:tensorflow:step = 9901, loss = 0.601023, precision = 0.90625 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.6422 +INFO:tensorflow:step = 10001, loss = 0.821797, precision = 0.820312 (2.729 sec) +INFO:tensorflow:global_step/sec: 36.6226 +INFO:tensorflow:step = 10101, loss = 0.662084, precision = 0.898438 (2.730 sec) +Saved checkpoint after 26 epoch(s) to data/resnet20/checkpoints/00026... +INFO:tensorflow:global_step/sec: 31.08 +INFO:tensorflow:step = 10201, loss = 0.616583, precision = 0.882812 (3.218 sec) +INFO:tensorflow:global_step/sec: 36.5885 +INFO:tensorflow:step = 10301, loss = 0.624081, precision = 0.90625 (2.733 sec) +INFO:tensorflow:global_step/sec: 36.6699 +INFO:tensorflow:step = 10401, loss = 0.652148, precision = 0.875 (2.727 sec) +INFO:tensorflow:global_step/sec: 36.632 +INFO:tensorflow:step = 10501, loss = 0.626866, precision = 0.859375 (2.730 sec) +Saved checkpoint after 27 epoch(s) to data/resnet20/checkpoints/00027... +INFO:tensorflow:global_step/sec: 31.1474 +INFO:tensorflow:step = 10601, loss = 0.700255, precision = 0.84375 (3.211 sec) +INFO:tensorflow:global_step/sec: 36.5403 +INFO:tensorflow:step = 10701, loss = 0.639249, precision = 0.835938 (2.737 sec) +INFO:tensorflow:global_step/sec: 36.6111 +INFO:tensorflow:step = 10801, loss = 0.746165, precision = 0.8125 (2.731 sec) +INFO:tensorflow:global_step/sec: 36.9092 +INFO:tensorflow:step = 10901, loss = 0.614565, precision = 0.890625 (2.709 sec) +Saved checkpoint after 28 epoch(s) to data/resnet20/checkpoints/00028... +INFO:tensorflow:global_step/sec: 31.013 +INFO:tensorflow:step = 11001, loss = 0.579167, precision = 0.882812 (3.225 sec) +INFO:tensorflow:global_step/sec: 36.8457 +INFO:tensorflow:step = 11101, loss = 0.682114, precision = 0.828125 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.7836 +INFO:tensorflow:step = 11201, loss = 0.663038, precision = 0.875 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.8809 +INFO:tensorflow:step = 11301, loss = 0.665302, precision = 0.828125 (2.711 sec) +Saved checkpoint after 29 epoch(s) to data/resnet20/checkpoints/00029... +INFO:tensorflow:global_step/sec: 31.102 +INFO:tensorflow:step = 11401, loss = 0.60645, precision = 0.898438 (3.215 sec) +INFO:tensorflow:global_step/sec: 36.4086 +INFO:tensorflow:step = 11501, loss = 0.694972, precision = 0.828125 (2.747 sec) +INFO:tensorflow:global_step/sec: 36.6199 +INFO:tensorflow:step = 11601, loss = 0.608904, precision = 0.875 (2.731 sec) +INFO:tensorflow:global_step/sec: 36.5583 +INFO:tensorflow:step = 11701, loss = 0.681695, precision = 0.828125 (2.735 sec) +Saved checkpoint after 30 epoch(s) to data/resnet20/checkpoints/00030... +INFO:tensorflow:global_step/sec: 31.3405 +INFO:tensorflow:step = 11801, loss = 0.651063, precision = 0.859375 (3.191 sec) +INFO:tensorflow:global_step/sec: 36.6295 +INFO:tensorflow:step = 11901, loss = 0.631688, precision = 0.890625 (2.730 sec) +INFO:tensorflow:global_step/sec: 36.7037 +INFO:tensorflow:step = 12001, loss = 0.600401, precision = 0.890625 (2.724 sec) +INFO:tensorflow:global_step/sec: 36.81 +INFO:tensorflow:step = 12101, loss = 0.635845, precision = 0.890625 (2.717 sec) +Saved checkpoint after 31 epoch(s) to data/resnet20/checkpoints/00031... +INFO:tensorflow:global_step/sec: 31.137 +INFO:tensorflow:step = 12201, loss = 0.659406, precision = 0.882812 (3.212 sec) +INFO:tensorflow:global_step/sec: 37.1682 +INFO:tensorflow:step = 12301, loss = 0.764168, precision = 0.851562 (2.691 sec) +INFO:tensorflow:global_step/sec: 36.6911 +INFO:tensorflow:step = 12401, loss = 0.648244, precision = 0.867188 (2.725 sec) +INFO:tensorflow:global_step/sec: 37.0434 +INFO:tensorflow:step = 12501, loss = 0.736476, precision = 0.835938 (2.700 sec) +Saved checkpoint after 32 epoch(s) to data/resnet20/checkpoints/00032... +INFO:tensorflow:global_step/sec: 31.404 +INFO:tensorflow:step = 12601, loss = 0.714813, precision = 0.851562 (3.184 sec) +INFO:tensorflow:global_step/sec: 37.3697 +INFO:tensorflow:step = 12701, loss = 0.657914, precision = 0.859375 (2.676 sec) +INFO:tensorflow:global_step/sec: 37.664 +INFO:tensorflow:step = 12801, loss = 0.657276, precision = 0.882812 (2.655 sec) +INFO:tensorflow:global_step/sec: 37.3367 +INFO:tensorflow:step = 12901, loss = 0.601974, precision = 0.882812 (2.678 sec) +Saved checkpoint after 33 epoch(s) to data/resnet20/checkpoints/00033... +INFO:tensorflow:global_step/sec: 32.1034 +INFO:tensorflow:step = 13001, loss = 0.624358, precision = 0.851562 (3.115 sec) +INFO:tensorflow:global_step/sec: 36.9624 +INFO:tensorflow:step = 13101, loss = 0.735229, precision = 0.804688 (2.705 sec) +INFO:tensorflow:global_step/sec: 36.5375 +INFO:tensorflow:step = 13201, loss = 0.689314, precision = 0.859375 (2.737 sec) +Saved checkpoint after 34 epoch(s) to data/resnet20/checkpoints/00034... +INFO:tensorflow:global_step/sec: 31.0807 +INFO:tensorflow:step = 13301, loss = 0.658033, precision = 0.851562 (3.218 sec) +INFO:tensorflow:global_step/sec: 36.5216 +INFO:tensorflow:step = 13401, loss = 0.632308, precision = 0.890625 (2.738 sec) +INFO:tensorflow:global_step/sec: 36.5414 +INFO:tensorflow:step = 13501, loss = 0.594866, precision = 0.882812 (2.737 sec) +INFO:tensorflow:global_step/sec: 36.4299 +INFO:tensorflow:step = 13601, loss = 0.862075, precision = 0.757812 (2.745 sec) +Saved checkpoint after 35 epoch(s) to data/resnet20/checkpoints/00035... +INFO:tensorflow:global_step/sec: 30.895 +INFO:tensorflow:step = 13701, loss = 0.885717, precision = 0.789062 (3.237 sec) +INFO:tensorflow:global_step/sec: 36.5304 +INFO:tensorflow:step = 13801, loss = 0.512465, precision = 0.921875 (2.737 sec) +INFO:tensorflow:global_step/sec: 36.7729 +INFO:tensorflow:step = 13901, loss = 0.717732, precision = 0.851562 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.7351 +INFO:tensorflow:step = 14001, loss = 0.736639, precision = 0.828125 (2.722 sec) +Saved checkpoint after 36 epoch(s) to data/resnet20/checkpoints/00036... +INFO:tensorflow:global_step/sec: 31.5312 +INFO:tensorflow:step = 14101, loss = 0.608539, precision = 0.882812 (3.172 sec) +INFO:tensorflow:global_step/sec: 36.9264 +INFO:tensorflow:step = 14201, loss = 0.602021, precision = 0.875 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.8261 +INFO:tensorflow:step = 14301, loss = 0.903494, precision = 0.78125 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.8611 +INFO:tensorflow:step = 14401, loss = 0.789973, precision = 0.820312 (2.713 sec) +Saved checkpoint after 37 epoch(s) to data/resnet20/checkpoints/00037... +INFO:tensorflow:global_step/sec: 31.1343 +INFO:tensorflow:step = 14501, loss = 0.73986, precision = 0.828125 (3.212 sec) +INFO:tensorflow:global_step/sec: 37.1752 +INFO:tensorflow:step = 14601, loss = 0.624382, precision = 0.867188 (2.690 sec) +INFO:tensorflow:global_step/sec: 36.9439 +INFO:tensorflow:step = 14701, loss = 0.649032, precision = 0.882812 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.5813 +INFO:tensorflow:step = 14801, loss = 0.680569, precision = 0.875 (2.734 sec) +Saved checkpoint after 38 epoch(s) to data/resnet20/checkpoints/00038... +INFO:tensorflow:global_step/sec: 30.8168 +INFO:tensorflow:step = 14901, loss = 0.663856, precision = 0.851562 (3.245 sec) +INFO:tensorflow:global_step/sec: 37.15 +INFO:tensorflow:step = 15001, loss = 0.639683, precision = 0.851562 (2.692 sec) +INFO:tensorflow:global_step/sec: 37.1623 +INFO:tensorflow:step = 15101, loss = 0.620906, precision = 0.835938 (2.691 sec) +INFO:tensorflow:global_step/sec: 36.8249 +INFO:tensorflow:step = 15201, loss = 0.651277, precision = 0.867188 (2.716 sec) +Saved checkpoint after 39 epoch(s) to data/resnet20/checkpoints/00039... +INFO:tensorflow:global_step/sec: 31.1933 +INFO:tensorflow:step = 15301, loss = 0.657676, precision = 0.828125 (3.206 sec) +INFO:tensorflow:global_step/sec: 36.8145 +INFO:tensorflow:step = 15401, loss = 0.568478, precision = 0.90625 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.815 +INFO:tensorflow:step = 15501, loss = 0.565685, precision = 0.90625 (2.717 sec) +INFO:tensorflow:global_step/sec: 37.1239 +INFO:tensorflow:step = 15601, loss = 0.640833, precision = 0.90625 (2.694 sec) +Saved checkpoint after 40 epoch(s) to data/resnet20/checkpoints/00040... +INFO:tensorflow:global_step/sec: 30.6065 +INFO:tensorflow:step = 15701, loss = 0.611595, precision = 0.851562 (3.267 sec) +INFO:tensorflow:global_step/sec: 37.0707 +INFO:tensorflow:step = 15801, loss = 0.724388, precision = 0.851562 (2.698 sec) +INFO:tensorflow:global_step/sec: 36.8464 +INFO:tensorflow:step = 15901, loss = 0.590611, precision = 0.90625 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.5696 +INFO:tensorflow:step = 16001, loss = 0.761906, precision = 0.851562 (2.735 sec) +Saved checkpoint after 41 epoch(s) to data/resnet20/checkpoints/00041... +INFO:tensorflow:global_step/sec: 31.515 +INFO:tensorflow:step = 16101, loss = 0.584455, precision = 0.875 (3.173 sec) +INFO:tensorflow:global_step/sec: 36.83 +INFO:tensorflow:step = 16201, loss = 0.708605, precision = 0.867188 (2.715 sec) +INFO:tensorflow:global_step/sec: 37.0939 +INFO:tensorflow:step = 16301, loss = 0.715807, precision = 0.851562 (2.696 sec) +INFO:tensorflow:global_step/sec: 37.0506 +INFO:tensorflow:step = 16401, loss = 0.493945, precision = 0.921875 (2.699 sec) +Saved checkpoint after 42 epoch(s) to data/resnet20/checkpoints/00042... +INFO:tensorflow:global_step/sec: 31.5431 +INFO:tensorflow:step = 16501, loss = 0.717815, precision = 0.851562 (3.170 sec) +INFO:tensorflow:global_step/sec: 37.0497 +INFO:tensorflow:step = 16601, loss = 0.595899, precision = 0.875 (2.699 sec) +INFO:tensorflow:global_step/sec: 36.5936 +INFO:tensorflow:step = 16701, loss = 0.680476, precision = 0.875 (2.733 sec) +INFO:tensorflow:global_step/sec: 37.0139 +INFO:tensorflow:step = 16801, loss = 0.907738, precision = 0.78125 (2.702 sec) +Saved checkpoint after 43 epoch(s) to data/resnet20/checkpoints/00043... +INFO:tensorflow:global_step/sec: 31.3871 +INFO:tensorflow:step = 16901, loss = 0.620794, precision = 0.882812 (3.186 sec) +INFO:tensorflow:global_step/sec: 37.3531 +INFO:tensorflow:step = 17001, loss = 0.7248, precision = 0.851562 (2.677 sec) +INFO:tensorflow:global_step/sec: 37.2006 +INFO:tensorflow:step = 17101, loss = 0.667465, precision = 0.859375 (2.688 sec) +INFO:tensorflow:global_step/sec: 36.9815 +INFO:tensorflow:step = 17201, loss = 0.709587, precision = 0.867188 (2.705 sec) +Saved checkpoint after 44 epoch(s) to data/resnet20/checkpoints/00044... +INFO:tensorflow:global_step/sec: 31.2474 +INFO:tensorflow:step = 17301, loss = 0.721344, precision = 0.820312 (3.200 sec) +INFO:tensorflow:global_step/sec: 36.9529 +INFO:tensorflow:step = 17401, loss = 0.580729, precision = 0.890625 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.9075 +INFO:tensorflow:step = 17501, loss = 0.72536, precision = 0.835938 (2.709 sec) +Saved checkpoint after 45 epoch(s) to data/resnet20/checkpoints/00045... +INFO:tensorflow:global_step/sec: 31.4619 +INFO:tensorflow:step = 17601, loss = 0.709579, precision = 0.851562 (3.179 sec) +INFO:tensorflow:global_step/sec: 36.8216 +INFO:tensorflow:step = 17701, loss = 0.755051, precision = 0.851562 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.7754 +INFO:tensorflow:step = 17801, loss = 0.685333, precision = 0.898438 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.732 +INFO:tensorflow:step = 17901, loss = 0.664626, precision = 0.867188 (2.723 sec) +Saved checkpoint after 46 epoch(s) to data/resnet20/checkpoints/00046... +INFO:tensorflow:global_step/sec: 31.092 +INFO:tensorflow:step = 18001, loss = 0.540409, precision = 0.921875 (3.216 sec) +INFO:tensorflow:global_step/sec: 36.9357 +INFO:tensorflow:step = 18101, loss = 0.600214, precision = 0.898438 (2.707 sec) +INFO:tensorflow:global_step/sec: 37.0418 +INFO:tensorflow:step = 18201, loss = 0.765904, precision = 0.804688 (2.700 sec) +INFO:tensorflow:global_step/sec: 36.924 +INFO:tensorflow:step = 18301, loss = 0.72374, precision = 0.859375 (2.708 sec) +Saved checkpoint after 47 epoch(s) to data/resnet20/checkpoints/00047... +INFO:tensorflow:global_step/sec: 31.3776 +INFO:tensorflow:step = 18401, loss = 0.680403, precision = 0.851562 (3.187 sec) +INFO:tensorflow:global_step/sec: 36.5568 +INFO:tensorflow:step = 18501, loss = 0.621526, precision = 0.875 (2.735 sec) +INFO:tensorflow:global_step/sec: 36.7206 +INFO:tensorflow:step = 18601, loss = 0.636099, precision = 0.890625 (2.724 sec) +INFO:tensorflow:global_step/sec: 36.9259 +INFO:tensorflow:step = 18701, loss = 0.71095, precision = 0.859375 (2.708 sec) +Saved checkpoint after 48 epoch(s) to data/resnet20/checkpoints/00048... +INFO:tensorflow:global_step/sec: 31.2373 +INFO:tensorflow:step = 18801, loss = 0.612405, precision = 0.9375 (3.201 sec) +INFO:tensorflow:global_step/sec: 36.7245 +INFO:tensorflow:step = 18901, loss = 0.71039, precision = 0.835938 (2.723 sec) +INFO:tensorflow:global_step/sec: 36.6958 +INFO:tensorflow:step = 19001, loss = 0.617696, precision = 0.882812 (2.725 sec) +INFO:tensorflow:global_step/sec: 37.068 +INFO:tensorflow:step = 19101, loss = 0.662404, precision = 0.84375 (2.698 sec) +Saved checkpoint after 49 epoch(s) to data/resnet20/checkpoints/00049... +INFO:tensorflow:global_step/sec: 31.1417 +INFO:tensorflow:step = 19201, loss = 0.567496, precision = 0.898438 (3.211 sec) +INFO:tensorflow:global_step/sec: 36.6711 +INFO:tensorflow:step = 19301, loss = 0.699809, precision = 0.867188 (2.727 sec) +INFO:tensorflow:global_step/sec: 36.8198 +INFO:tensorflow:step = 19401, loss = 0.599155, precision = 0.90625 (2.716 sec) +INFO:tensorflow:global_step/sec: 37.0502 +INFO:tensorflow:step = 19501, loss = 0.600716, precision = 0.890625 (2.699 sec) +Saved checkpoint after 50 epoch(s) to data/resnet20/checkpoints/00050... +INFO:tensorflow:global_step/sec: 31.899 +INFO:tensorflow:step = 19601, loss = 0.736475, precision = 0.867188 (3.135 sec) +INFO:tensorflow:global_step/sec: 37.0328 +INFO:tensorflow:step = 19701, loss = 0.563628, precision = 0.90625 (2.700 sec) +INFO:tensorflow:global_step/sec: 36.953 +INFO:tensorflow:step = 19801, loss = 0.688484, precision = 0.867188 (2.706 sec) +INFO:tensorflow:global_step/sec: 36.7701 +INFO:tensorflow:step = 19901, loss = 0.647948, precision = 0.859375 (2.719 sec) +Saved checkpoint after 51 epoch(s) to data/resnet20/checkpoints/00051... +INFO:tensorflow:global_step/sec: 31.3915 +INFO:tensorflow:step = 20001, loss = 0.660744, precision = 0.867188 (3.186 sec) +INFO:tensorflow:global_step/sec: 36.6768 +INFO:tensorflow:step = 20101, loss = 0.682581, precision = 0.820312 (2.726 sec) +INFO:tensorflow:global_step/sec: 37.0056 +INFO:tensorflow:step = 20201, loss = 0.803198, precision = 0.835938 (2.702 sec) +INFO:tensorflow:global_step/sec: 37.199 +INFO:tensorflow:step = 20301, loss = 0.711824, precision = 0.875 (2.688 sec) +Saved checkpoint after 52 epoch(s) to data/resnet20/checkpoints/00052... +INFO:tensorflow:global_step/sec: 31.4941 +INFO:tensorflow:step = 20401, loss = 0.684902, precision = 0.828125 (3.176 sec) +INFO:tensorflow:global_step/sec: 36.8447 +INFO:tensorflow:step = 20501, loss = 0.597211, precision = 0.914062 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.9467 +INFO:tensorflow:step = 20601, loss = 0.740443, precision = 0.851562 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.7889 +INFO:tensorflow:step = 20701, loss = 0.595744, precision = 0.898438 (2.718 sec) +Saved checkpoint after 53 epoch(s) to data/resnet20/checkpoints/00053... +INFO:tensorflow:global_step/sec: 31.3861 +INFO:tensorflow:step = 20801, loss = 0.758135, precision = 0.8125 (3.186 sec) +INFO:tensorflow:global_step/sec: 36.5855 +INFO:tensorflow:step = 20901, loss = 0.736187, precision = 0.882812 (2.733 sec) +INFO:tensorflow:global_step/sec: 36.7868 +INFO:tensorflow:step = 21001, loss = 0.603109, precision = 0.882812 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.7078 +INFO:tensorflow:step = 21101, loss = 0.620879, precision = 0.890625 (2.724 sec) +Saved checkpoint after 54 epoch(s) to data/resnet20/checkpoints/00054... +INFO:tensorflow:global_step/sec: 31.3125 +INFO:tensorflow:step = 21201, loss = 0.76653, precision = 0.820312 (3.194 sec) +INFO:tensorflow:global_step/sec: 36.7979 +INFO:tensorflow:step = 21301, loss = 0.703813, precision = 0.882812 (2.717 sec) +INFO:tensorflow:global_step/sec: 37.0652 +INFO:tensorflow:step = 21401, loss = 0.581393, precision = 0.882812 (2.698 sec) +INFO:tensorflow:global_step/sec: 36.7573 +INFO:tensorflow:step = 21501, loss = 0.702136, precision = 0.875 (2.721 sec) +Saved checkpoint after 55 epoch(s) to data/resnet20/checkpoints/00055... +INFO:tensorflow:global_step/sec: 31.4805 +INFO:tensorflow:step = 21601, loss = 0.597173, precision = 0.890625 (3.177 sec) +INFO:tensorflow:global_step/sec: 36.7826 +INFO:tensorflow:step = 21701, loss = 0.713952, precision = 0.875 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.9114 +INFO:tensorflow:step = 21801, loss = 0.701855, precision = 0.859375 (2.709 sec) +Saved checkpoint after 56 epoch(s) to data/resnet20/checkpoints/00056... +INFO:tensorflow:global_step/sec: 31.9847 +INFO:tensorflow:step = 21901, loss = 0.62086, precision = 0.882812 (3.127 sec) +INFO:tensorflow:global_step/sec: 36.892 +INFO:tensorflow:step = 22001, loss = 0.467132, precision = 0.921875 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.4816 +INFO:tensorflow:step = 22101, loss = 0.602439, precision = 0.898438 (2.741 sec) +INFO:tensorflow:global_step/sec: 36.7123 +INFO:tensorflow:step = 22201, loss = 0.554188, precision = 0.90625 (2.724 sec) +Saved checkpoint after 57 epoch(s) to data/resnet20/checkpoints/00057... +INFO:tensorflow:global_step/sec: 31.4272 +INFO:tensorflow:step = 22301, loss = 0.646833, precision = 0.867188 (3.182 sec) +INFO:tensorflow:global_step/sec: 36.7437 +INFO:tensorflow:step = 22401, loss = 0.685159, precision = 0.84375 (2.722 sec) +INFO:tensorflow:global_step/sec: 36.8093 +INFO:tensorflow:step = 22501, loss = 0.632148, precision = 0.859375 (2.717 sec) +INFO:tensorflow:global_step/sec: 37.1797 +INFO:tensorflow:step = 22601, loss = 0.672925, precision = 0.875 (2.690 sec) +Saved checkpoint after 58 epoch(s) to data/resnet20/checkpoints/00058... +INFO:tensorflow:global_step/sec: 31.0494 +INFO:tensorflow:step = 22701, loss = 0.601497, precision = 0.898438 (3.220 sec) +INFO:tensorflow:global_step/sec: 36.8675 +INFO:tensorflow:step = 22801, loss = 0.809993, precision = 0.828125 (2.712 sec) +INFO:tensorflow:global_step/sec: 36.8782 +INFO:tensorflow:step = 22901, loss = 0.6207, precision = 0.882812 (2.712 sec) +INFO:tensorflow:global_step/sec: 36.8635 +INFO:tensorflow:step = 23001, loss = 0.674742, precision = 0.875 (2.713 sec) +Saved checkpoint after 59 epoch(s) to data/resnet20/checkpoints/00059... +INFO:tensorflow:global_step/sec: 31.29 +INFO:tensorflow:step = 23101, loss = 0.571147, precision = 0.921875 (3.196 sec) +INFO:tensorflow:global_step/sec: 36.8652 +INFO:tensorflow:step = 23201, loss = 0.63097, precision = 0.859375 (2.713 sec) +INFO:tensorflow:global_step/sec: 36.8778 +INFO:tensorflow:step = 23301, loss = 0.656497, precision = 0.867188 (2.712 sec) +INFO:tensorflow:global_step/sec: 36.9192 +INFO:tensorflow:step = 23401, loss = 0.658761, precision = 0.890625 (2.708 sec) +Saved checkpoint after 60 epoch(s) to data/resnet20/checkpoints/00060... +INFO:tensorflow:global_step/sec: 31.1178 +INFO:tensorflow:step = 23501, loss = 0.570772, precision = 0.882812 (3.214 sec) +INFO:tensorflow:global_step/sec: 36.5392 +INFO:tensorflow:step = 23601, loss = 0.646856, precision = 0.851562 (2.737 sec) +INFO:tensorflow:global_step/sec: 36.8489 +INFO:tensorflow:step = 23701, loss = 0.602991, precision = 0.875 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.5147 +INFO:tensorflow:step = 23801, loss = 0.63796, precision = 0.882812 (2.739 sec) +Saved checkpoint after 61 epoch(s) to data/resnet20/checkpoints/00061... +INFO:tensorflow:global_step/sec: 31.0598 +INFO:tensorflow:step = 23901, loss = 0.622749, precision = 0.859375 (3.220 sec) +INFO:tensorflow:global_step/sec: 36.7219 +INFO:tensorflow:step = 24001, loss = 0.832498, precision = 0.804688 (2.723 sec) +INFO:tensorflow:global_step/sec: 36.4837 +INFO:tensorflow:step = 24101, loss = 0.587171, precision = 0.875 (2.741 sec) +INFO:tensorflow:global_step/sec: 36.3925 +INFO:tensorflow:step = 24201, loss = 0.728096, precision = 0.882812 (2.748 sec) +Saved checkpoint after 62 epoch(s) to data/resnet20/checkpoints/00062... +INFO:tensorflow:global_step/sec: 31.3469 +INFO:tensorflow:step = 24301, loss = 0.648914, precision = 0.828125 (3.190 sec) +INFO:tensorflow:global_step/sec: 36.7404 +INFO:tensorflow:step = 24401, loss = 0.597155, precision = 0.875 (2.722 sec) +INFO:tensorflow:global_step/sec: 36.8479 +INFO:tensorflow:step = 24501, loss = 0.683201, precision = 0.820312 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.8239 +INFO:tensorflow:step = 24601, loss = 0.674858, precision = 0.851562 (2.716 sec) +Saved checkpoint after 63 epoch(s) to data/resnet20/checkpoints/00063... +INFO:tensorflow:global_step/sec: 31.0665 +INFO:tensorflow:step = 24701, loss = 0.62014, precision = 0.859375 (3.219 sec) +INFO:tensorflow:global_step/sec: 37.0432 +INFO:tensorflow:step = 24801, loss = 0.777603, precision = 0.84375 (2.699 sec) +INFO:tensorflow:global_step/sec: 36.7406 +INFO:tensorflow:step = 24901, loss = 0.620829, precision = 0.875 (2.722 sec) +INFO:tensorflow:global_step/sec: 36.9052 +INFO:tensorflow:step = 25001, loss = 0.666708, precision = 0.875 (2.709 sec) +Saved checkpoint after 64 epoch(s) to data/resnet20/checkpoints/00064... +INFO:tensorflow:global_step/sec: 31.3029 +INFO:tensorflow:step = 25101, loss = 0.522227, precision = 0.929688 (3.195 sec) +INFO:tensorflow:global_step/sec: 36.8324 +INFO:tensorflow:step = 25201, loss = 0.612325, precision = 0.859375 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.7662 +INFO:tensorflow:step = 25301, loss = 0.593641, precision = 0.84375 (2.720 sec) +INFO:tensorflow:global_step/sec: 36.5938 +INFO:tensorflow:step = 25401, loss = 0.72501, precision = 0.828125 (2.733 sec) +Saved checkpoint after 65 epoch(s) to data/resnet20/checkpoints/00065... +INFO:tensorflow:global_step/sec: 31.085 +INFO:tensorflow:step = 25501, loss = 0.555607, precision = 0.921875 (3.217 sec) +INFO:tensorflow:global_step/sec: 36.9606 +INFO:tensorflow:step = 25601, loss = 0.805237, precision = 0.8125 (2.705 sec) +INFO:tensorflow:global_step/sec: 36.8717 +INFO:tensorflow:step = 25701, loss = 0.573198, precision = 0.890625 (2.712 sec) +INFO:tensorflow:global_step/sec: 36.9087 +INFO:tensorflow:step = 25801, loss = 0.651045, precision = 0.84375 (2.710 sec) +Saved checkpoint after 66 epoch(s) to data/resnet20/checkpoints/00066... +INFO:tensorflow:global_step/sec: 31.3963 +INFO:tensorflow:step = 25901, loss = 0.517518, precision = 0.914062 (3.185 sec) +INFO:tensorflow:global_step/sec: 36.9915 +INFO:tensorflow:step = 26001, loss = 0.544592, precision = 0.90625 (2.703 sec) +INFO:tensorflow:global_step/sec: 36.8538 +INFO:tensorflow:step = 26101, loss = 0.59364, precision = 0.882812 (2.713 sec) +Saved checkpoint after 67 epoch(s) to data/resnet20/checkpoints/00067... +INFO:tensorflow:global_step/sec: 31.6763 +INFO:tensorflow:step = 26201, loss = 0.695491, precision = 0.851562 (3.157 sec) +INFO:tensorflow:global_step/sec: 36.7882 +INFO:tensorflow:step = 26301, loss = 0.729084, precision = 0.890625 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.6398 +INFO:tensorflow:step = 26401, loss = 0.69452, precision = 0.859375 (2.729 sec) +INFO:tensorflow:global_step/sec: 36.6089 +INFO:tensorflow:step = 26501, loss = 0.735517, precision = 0.835938 (2.732 sec) +Saved checkpoint after 68 epoch(s) to data/resnet20/checkpoints/00068... +INFO:tensorflow:global_step/sec: 31.7243 +INFO:tensorflow:step = 26601, loss = 0.798288, precision = 0.789062 (3.152 sec) +INFO:tensorflow:global_step/sec: 36.679 +INFO:tensorflow:step = 26701, loss = 0.701408, precision = 0.84375 (2.728 sec) +INFO:tensorflow:global_step/sec: 36.5967 +INFO:tensorflow:step = 26801, loss = 0.676335, precision = 0.859375 (2.731 sec) +INFO:tensorflow:global_step/sec: 36.6353 +INFO:tensorflow:step = 26901, loss = 0.595611, precision = 0.859375 (2.730 sec) +Saved checkpoint after 69 epoch(s) to data/resnet20/checkpoints/00069... +INFO:tensorflow:global_step/sec: 31.649 +INFO:tensorflow:step = 27001, loss = 0.682547, precision = 0.84375 (3.159 sec) +INFO:tensorflow:global_step/sec: 36.9166 +INFO:tensorflow:step = 27101, loss = 0.683892, precision = 0.898438 (2.709 sec) +INFO:tensorflow:global_step/sec: 36.7964 +INFO:tensorflow:step = 27201, loss = 0.62492, precision = 0.867188 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.7488 +INFO:tensorflow:step = 27301, loss = 0.669969, precision = 0.882812 (2.721 sec) +Saved checkpoint after 70 epoch(s) to data/resnet20/checkpoints/00070... +INFO:tensorflow:global_step/sec: 31.284 +INFO:tensorflow:step = 27401, loss = 0.680698, precision = 0.8125 (3.196 sec) +INFO:tensorflow:global_step/sec: 36.8011 +INFO:tensorflow:step = 27501, loss = 0.610269, precision = 0.914062 (2.717 sec) +INFO:tensorflow:global_step/sec: 36.6732 +INFO:tensorflow:step = 27601, loss = 0.670278, precision = 0.882812 (2.727 sec) +INFO:tensorflow:global_step/sec: 36.9653 +INFO:tensorflow:step = 27701, loss = 0.607711, precision = 0.875 (2.705 sec) +Saved checkpoint after 71 epoch(s) to data/resnet20/checkpoints/00071... +INFO:tensorflow:global_step/sec: 31.1922 +INFO:tensorflow:step = 27801, loss = 0.705552, precision = 0.859375 (3.206 sec) +INFO:tensorflow:global_step/sec: 37.0854 +INFO:tensorflow:step = 27901, loss = 0.716737, precision = 0.859375 (2.696 sec) +INFO:tensorflow:global_step/sec: 37.0395 +INFO:tensorflow:step = 28001, loss = 0.701087, precision = 0.835938 (2.700 sec) +INFO:tensorflow:global_step/sec: 37.1684 +INFO:tensorflow:step = 28101, loss = 0.592297, precision = 0.914062 (2.690 sec) +Saved checkpoint after 72 epoch(s) to data/resnet20/checkpoints/00072... +INFO:tensorflow:global_step/sec: 31.3795 +INFO:tensorflow:step = 28201, loss = 0.800866, precision = 0.804688 (3.187 sec) +INFO:tensorflow:global_step/sec: 37.0911 +INFO:tensorflow:step = 28301, loss = 0.68073, precision = 0.867188 (2.696 sec) +INFO:tensorflow:global_step/sec: 37.1527 +INFO:tensorflow:step = 28401, loss = 0.802916, precision = 0.820312 (2.691 sec) +INFO:tensorflow:global_step/sec: 37.1296 +INFO:tensorflow:step = 28501, loss = 0.60761, precision = 0.90625 (2.694 sec) +Saved checkpoint after 73 epoch(s) to data/resnet20/checkpoints/00073... +INFO:tensorflow:global_step/sec: 31.1691 +INFO:tensorflow:step = 28601, loss = 0.69894, precision = 0.84375 (3.208 sec) +INFO:tensorflow:global_step/sec: 36.5085 +INFO:tensorflow:step = 28701, loss = 0.626737, precision = 0.867188 (2.739 sec) +INFO:tensorflow:global_step/sec: 36.9363 +INFO:tensorflow:step = 28801, loss = 0.676653, precision = 0.84375 (2.707 sec) +INFO:tensorflow:global_step/sec: 37.2319 +INFO:tensorflow:step = 28901, loss = 0.77938, precision = 0.890625 (2.686 sec) +Saved checkpoint after 74 epoch(s) to data/resnet20/checkpoints/00074... +INFO:tensorflow:global_step/sec: 31.6306 +INFO:tensorflow:step = 29001, loss = 0.838053, precision = 0.828125 (3.162 sec) +INFO:tensorflow:global_step/sec: 36.7678 +INFO:tensorflow:step = 29101, loss = 0.668686, precision = 0.875 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.9459 +INFO:tensorflow:step = 29201, loss = 0.71341, precision = 0.859375 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.8762 +INFO:tensorflow:step = 29301, loss = 0.64552, precision = 0.875 (2.711 sec) +Saved checkpoint after 75 epoch(s) to data/resnet20/checkpoints/00075... +INFO:tensorflow:global_step/sec: 31.4876 +INFO:tensorflow:step = 29401, loss = 0.640728, precision = 0.898438 (3.176 sec) +INFO:tensorflow:global_step/sec: 36.8167 +INFO:tensorflow:step = 29501, loss = 0.727708, precision = 0.851562 (2.716 sec) +INFO:tensorflow:global_step/sec: 37.0671 +INFO:tensorflow:step = 29601, loss = 0.613064, precision = 0.875 (2.698 sec) +INFO:tensorflow:global_step/sec: 36.9756 +INFO:tensorflow:step = 29701, loss = 0.628492, precision = 0.882812 (2.704 sec) +Saved checkpoint after 76 epoch(s) to data/resnet20/checkpoints/00076... +INFO:tensorflow:global_step/sec: 31.2696 +INFO:tensorflow:step = 29801, loss = 0.695478, precision = 0.859375 (3.198 sec) +INFO:tensorflow:global_step/sec: 36.6435 +INFO:tensorflow:step = 29901, loss = 0.613503, precision = 0.882812 (2.729 sec) +INFO:tensorflow:global_step/sec: 36.9439 +INFO:tensorflow:step = 30001, loss = 0.758841, precision = 0.820312 (2.706 sec) +INFO:tensorflow:global_step/sec: 36.7161 +INFO:tensorflow:step = 30101, loss = 0.597345, precision = 0.890625 (2.724 sec) +Saved checkpoint after 77 epoch(s) to data/resnet20/checkpoints/00077... +INFO:tensorflow:global_step/sec: 31.1072 +INFO:tensorflow:step = 30201, loss = 0.733651, precision = 0.867188 (3.215 sec) +INFO:tensorflow:global_step/sec: 37.0289 +INFO:tensorflow:step = 30301, loss = 0.657394, precision = 0.890625 (2.700 sec) +INFO:tensorflow:global_step/sec: 36.8444 +INFO:tensorflow:step = 30401, loss = 0.61299, precision = 0.898438 (2.715 sec) +Saved checkpoint after 78 epoch(s) to data/resnet20/checkpoints/00078... +INFO:tensorflow:global_step/sec: 31.0346 +INFO:tensorflow:step = 30501, loss = 0.736815, precision = 0.84375 (3.222 sec) +INFO:tensorflow:global_step/sec: 36.6546 +INFO:tensorflow:step = 30601, loss = 0.674018, precision = 0.867188 (2.728 sec) +INFO:tensorflow:global_step/sec: 36.9384 +INFO:tensorflow:step = 30701, loss = 0.528972, precision = 0.90625 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.6324 +INFO:tensorflow:step = 30801, loss = 0.574519, precision = 0.914062 (2.730 sec) +Saved checkpoint after 79 epoch(s) to data/resnet20/checkpoints/00079... +INFO:tensorflow:global_step/sec: 29.5932 +INFO:tensorflow:step = 30901, loss = 0.713179, precision = 0.882812 (3.378 sec) +INFO:tensorflow:global_step/sec: 36.3749 +INFO:tensorflow:step = 31001, loss = 0.626368, precision = 0.882812 (2.749 sec) +INFO:tensorflow:global_step/sec: 36.7343 +INFO:tensorflow:step = 31101, loss = 0.65159, precision = 0.851562 (2.722 sec) +INFO:tensorflow:global_step/sec: 36.7426 +INFO:tensorflow:step = 31201, loss = 0.632867, precision = 0.875 (2.721 sec) +Saved checkpoint after 80 epoch(s) to data/resnet20/checkpoints/00080... +INFO:tensorflow:global_step/sec: 30.9747 +INFO:tensorflow:step = 31301, loss = 0.628437, precision = 0.898438 (3.229 sec) +INFO:tensorflow:global_step/sec: 36.4842 +INFO:tensorflow:step = 31401, loss = 0.709023, precision = 0.828125 (2.741 sec) +INFO:tensorflow:global_step/sec: 36.9645 +INFO:tensorflow:step = 31501, loss = 0.59266, precision = 0.890625 (2.705 sec) +INFO:tensorflow:global_step/sec: 36.7641 +INFO:tensorflow:step = 31601, loss = 0.727019, precision = 0.875 (2.720 sec) +Saved checkpoint after 81 epoch(s) to data/resnet20/checkpoints/00081... +INFO:tensorflow:global_step/sec: 31.0254 +INFO:tensorflow:step = 31701, loss = 0.60836, precision = 0.882812 (3.223 sec) +INFO:tensorflow:global_step/sec: 36.9531 +INFO:tensorflow:step = 31801, loss = 0.520888, precision = 0.90625 (2.706 sec) +INFO:tensorflow:global_step/sec: 37.0407 +INFO:tensorflow:step = 31901, loss = 0.541345, precision = 0.90625 (2.700 sec) +INFO:tensorflow:global_step/sec: 36.6729 +INFO:tensorflow:step = 32001, loss = 0.676791, precision = 0.851562 (2.726 sec) +Saved checkpoint after 82 epoch(s) to data/resnet20/checkpoints/00082... +INFO:tensorflow:global_step/sec: 31.1833 +INFO:tensorflow:step = 32101, loss = 0.700257, precision = 0.84375 (3.207 sec) +INFO:tensorflow:global_step/sec: 36.7619 +INFO:tensorflow:step = 32201, loss = 0.617776, precision = 0.875 (2.720 sec) +INFO:tensorflow:global_step/sec: 36.8374 +INFO:tensorflow:step = 32301, loss = 0.783643, precision = 0.835938 (2.715 sec) +INFO:tensorflow:global_step/sec: 36.5786 +INFO:tensorflow:step = 32401, loss = 0.687524, precision = 0.867188 (2.734 sec) +Saved checkpoint after 83 epoch(s) to data/resnet20/checkpoints/00083... +INFO:tensorflow:global_step/sec: 30.8083 +INFO:tensorflow:step = 32501, loss = 0.577449, precision = 0.890625 (3.246 sec) +INFO:tensorflow:global_step/sec: 36.861 +INFO:tensorflow:step = 32601, loss = 0.811041, precision = 0.828125 (2.713 sec) +INFO:tensorflow:global_step/sec: 36.8291 +INFO:tensorflow:step = 32701, loss = 0.593049, precision = 0.890625 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.3508 +INFO:tensorflow:step = 32801, loss = 0.682498, precision = 0.867188 (2.751 sec) +Saved checkpoint after 84 epoch(s) to data/resnet20/checkpoints/00084... +INFO:tensorflow:global_step/sec: 30.9613 +INFO:tensorflow:step = 32901, loss = 0.645048, precision = 0.898438 (3.229 sec) +INFO:tensorflow:global_step/sec: 37.1704 +INFO:tensorflow:step = 33001, loss = 0.725157, precision = 0.8125 (2.690 sec) +INFO:tensorflow:global_step/sec: 37.2232 +INFO:tensorflow:step = 33101, loss = 0.688686, precision = 0.8125 (2.686 sec) +INFO:tensorflow:global_step/sec: 36.9631 +INFO:tensorflow:step = 33201, loss = 0.558968, precision = 0.898438 (2.706 sec) +Saved checkpoint after 85 epoch(s) to data/resnet20/checkpoints/00085... +INFO:tensorflow:global_step/sec: 30.9318 +INFO:tensorflow:step = 33301, loss = 0.693714, precision = 0.859375 (3.233 sec) +INFO:tensorflow:global_step/sec: 36.9506 +INFO:tensorflow:step = 33401, loss = 0.44752, precision = 0.96875 (2.706 sec) +INFO:tensorflow:global_step/sec: 36.9394 +INFO:tensorflow:step = 33501, loss = 0.613104, precision = 0.867188 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.7976 +INFO:tensorflow:step = 33601, loss = 0.555017, precision = 0.890625 (2.717 sec) +Saved checkpoint after 86 epoch(s) to data/resnet20/checkpoints/00086... +INFO:tensorflow:global_step/sec: 31.3327 +INFO:tensorflow:step = 33701, loss = 0.66595, precision = 0.859375 (3.192 sec) +INFO:tensorflow:global_step/sec: 36.9902 +INFO:tensorflow:step = 33801, loss = 0.606915, precision = 0.914062 (2.703 sec) +INFO:tensorflow:global_step/sec: 37.0883 +INFO:tensorflow:step = 33901, loss = 0.679267, precision = 0.851562 (2.696 sec) +INFO:tensorflow:global_step/sec: 36.5577 +INFO:tensorflow:step = 34001, loss = 0.807418, precision = 0.8125 (2.736 sec) +Saved checkpoint after 87 epoch(s) to data/resnet20/checkpoints/00087... +INFO:tensorflow:global_step/sec: 31.558 +INFO:tensorflow:step = 34101, loss = 0.675401, precision = 0.859375 (3.169 sec) +INFO:tensorflow:global_step/sec: 36.9239 +INFO:tensorflow:step = 34201, loss = 0.616933, precision = 0.875 (2.709 sec) +INFO:tensorflow:global_step/sec: 36.7822 +INFO:tensorflow:step = 34301, loss = 0.561826, precision = 0.90625 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.8301 +INFO:tensorflow:step = 34401, loss = 0.731936, precision = 0.859375 (2.715 sec) +Saved checkpoint after 88 epoch(s) to data/resnet20/checkpoints/00088... +INFO:tensorflow:global_step/sec: 30.9146 +INFO:tensorflow:step = 34501, loss = 0.726222, precision = 0.835938 (3.235 sec) +INFO:tensorflow:global_step/sec: 37.1503 +INFO:tensorflow:step = 34601, loss = 0.750386, precision = 0.820312 (2.692 sec) +INFO:tensorflow:global_step/sec: 37.0838 +INFO:tensorflow:step = 34701, loss = 0.672868, precision = 0.851562 (2.697 sec) +Saved checkpoint after 89 epoch(s) to data/resnet20/checkpoints/00089... +INFO:tensorflow:global_step/sec: 31.6232 +INFO:tensorflow:step = 34801, loss = 0.566628, precision = 0.867188 (3.162 sec) +INFO:tensorflow:global_step/sec: 36.6973 +INFO:tensorflow:step = 34901, loss = 0.697829, precision = 0.828125 (2.725 sec) +INFO:tensorflow:global_step/sec: 37.034 +INFO:tensorflow:step = 35001, loss = 0.584156, precision = 0.890625 (2.700 sec) +INFO:tensorflow:global_step/sec: 36.6049 +INFO:tensorflow:step = 35101, loss = 0.69058, precision = 0.867188 (2.732 sec) +Saved checkpoint after 90 epoch(s) to data/resnet20/checkpoints/00090... +INFO:tensorflow:global_step/sec: 31.2261 +INFO:tensorflow:step = 35201, loss = 0.561034, precision = 0.898438 (3.202 sec) +INFO:tensorflow:global_step/sec: 36.9333 +INFO:tensorflow:step = 35301, loss = 0.770816, precision = 0.804688 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.7845 +INFO:tensorflow:step = 35401, loss = 0.598582, precision = 0.890625 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.9476 +INFO:tensorflow:step = 35501, loss = 0.635792, precision = 0.859375 (2.707 sec) +Saved checkpoint after 91 epoch(s) to data/resnet20/checkpoints/00091... +INFO:tensorflow:global_step/sec: 31.2059 +INFO:tensorflow:step = 35601, loss = 0.59827, precision = 0.90625 (3.204 sec) +INFO:tensorflow:global_step/sec: 37.131 +INFO:tensorflow:step = 35701, loss = 0.54265, precision = 0.914062 (2.694 sec) +INFO:tensorflow:global_step/sec: 36.4726 +INFO:tensorflow:step = 35801, loss = 0.511978, precision = 0.9375 (2.742 sec) +INFO:tensorflow:global_step/sec: 37.1032 +INFO:tensorflow:step = 35901, loss = 0.606709, precision = 0.882812 (2.695 sec) +Saved checkpoint after 92 epoch(s) to data/resnet20/checkpoints/00092... +INFO:tensorflow:global_step/sec: 31.4309 +INFO:tensorflow:step = 36001, loss = 0.513661, precision = 0.9375 (3.182 sec) +INFO:tensorflow:global_step/sec: 37.0025 +INFO:tensorflow:step = 36101, loss = 0.414992, precision = 0.960938 (2.703 sec) +INFO:tensorflow:global_step/sec: 36.9145 +INFO:tensorflow:step = 36201, loss = 0.562372, precision = 0.890625 (2.709 sec) +INFO:tensorflow:global_step/sec: 36.8489 +INFO:tensorflow:step = 36301, loss = 0.517166, precision = 0.882812 (2.714 sec) +Saved checkpoint after 93 epoch(s) to data/resnet20/checkpoints/00093... +INFO:tensorflow:global_step/sec: 31.5473 +INFO:tensorflow:step = 36401, loss = 0.49792, precision = 0.90625 (3.170 sec) +INFO:tensorflow:global_step/sec: 36.6929 +INFO:tensorflow:step = 36501, loss = 0.442701, precision = 0.921875 (2.726 sec) +INFO:tensorflow:global_step/sec: 36.8346 +INFO:tensorflow:step = 36601, loss = 0.436786, precision = 0.9375 (2.715 sec) +INFO:tensorflow:global_step/sec: 36.9712 +INFO:tensorflow:step = 36701, loss = 0.386868, precision = 0.96875 (2.705 sec) +Saved checkpoint after 94 epoch(s) to data/resnet20/checkpoints/00094... +INFO:tensorflow:global_step/sec: 31.2223 +INFO:tensorflow:step = 36801, loss = 0.413793, precision = 0.9375 (3.203 sec) +INFO:tensorflow:global_step/sec: 37.1596 +INFO:tensorflow:step = 36901, loss = 0.439765, precision = 0.9375 (2.691 sec) +INFO:tensorflow:global_step/sec: 37.0156 +INFO:tensorflow:step = 37001, loss = 0.403567, precision = 0.96875 (2.701 sec) +INFO:tensorflow:global_step/sec: 37.1834 +INFO:tensorflow:step = 37101, loss = 0.343192, precision = 0.976562 (2.690 sec) +Saved checkpoint after 95 epoch(s) to data/resnet20/checkpoints/00095... +INFO:tensorflow:global_step/sec: 31.2355 +INFO:tensorflow:step = 37201, loss = 0.35095, precision = 0.953125 (3.201 sec) +INFO:tensorflow:global_step/sec: 36.7034 +INFO:tensorflow:step = 37301, loss = 0.345201, precision = 0.960938 (2.725 sec) +INFO:tensorflow:global_step/sec: 36.6846 +INFO:tensorflow:step = 37401, loss = 0.422802, precision = 0.945312 (2.726 sec) +INFO:tensorflow:global_step/sec: 36.8257 +INFO:tensorflow:step = 37501, loss = 0.366715, precision = 0.9375 (2.716 sec) +Saved checkpoint after 96 epoch(s) to data/resnet20/checkpoints/00096... +INFO:tensorflow:global_step/sec: 31.5683 +INFO:tensorflow:step = 37601, loss = 0.462881, precision = 0.929688 (3.168 sec) +INFO:tensorflow:global_step/sec: 36.8183 +INFO:tensorflow:step = 37701, loss = 0.458898, precision = 0.929688 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.9283 +INFO:tensorflow:step = 37801, loss = 0.384685, precision = 0.953125 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.7844 +INFO:tensorflow:step = 37901, loss = 0.376337, precision = 0.960938 (2.718 sec) +Saved checkpoint after 97 epoch(s) to data/resnet20/checkpoints/00097... +INFO:tensorflow:global_step/sec: 31.0382 +INFO:tensorflow:step = 38001, loss = 0.416526, precision = 0.945312 (3.222 sec) +INFO:tensorflow:global_step/sec: 36.9726 +INFO:tensorflow:step = 38101, loss = 0.337416, precision = 0.960938 (2.705 sec) +INFO:tensorflow:global_step/sec: 36.8533 +INFO:tensorflow:step = 38201, loss = 0.379174, precision = 0.953125 (2.713 sec) +INFO:tensorflow:global_step/sec: 37.0322 +INFO:tensorflow:step = 38301, loss = 0.316853, precision = 0.96875 (2.700 sec) +Saved checkpoint after 98 epoch(s) to data/resnet20/checkpoints/00098... +INFO:tensorflow:global_step/sec: 31.4081 +INFO:tensorflow:step = 38401, loss = 0.420928, precision = 0.921875 (3.184 sec) +INFO:tensorflow:global_step/sec: 36.8792 +INFO:tensorflow:step = 38501, loss = 0.404361, precision = 0.929688 (2.711 sec) +INFO:tensorflow:global_step/sec: 37.1335 +INFO:tensorflow:step = 38601, loss = 0.342872, precision = 0.945312 (2.693 sec) +INFO:tensorflow:global_step/sec: 36.5063 +INFO:tensorflow:step = 38701, loss = 0.304442, precision = 0.976562 (2.739 sec) +Saved checkpoint after 99 epoch(s) to data/resnet20/checkpoints/00099... +INFO:tensorflow:global_step/sec: 30.6273 +INFO:tensorflow:step = 38801, loss = 0.327438, precision = 0.945312 (3.265 sec) +INFO:tensorflow:global_step/sec: 36.9717 +INFO:tensorflow:step = 38901, loss = 0.347762, precision = 0.953125 (2.704 sec) +INFO:tensorflow:global_step/sec: 36.9118 +INFO:tensorflow:step = 39001, loss = 0.357917, precision = 0.953125 (2.710 sec) +Saved checkpoint after 100 epoch(s) to data/resnet20/checkpoints/00100... +INFO:tensorflow:global_step/sec: 31.3977 +INFO:tensorflow:step = 39101, loss = 0.419093, precision = 0.929688 (3.185 sec) +INFO:tensorflow:global_step/sec: 36.7287 +INFO:tensorflow:step = 39201, loss = 0.271341, precision = 0.984375 (2.723 sec) +INFO:tensorflow:global_step/sec: 37.0142 +INFO:tensorflow:step = 39301, loss = 0.301802, precision = 0.976562 (2.702 sec) +INFO:tensorflow:global_step/sec: 36.9857 +INFO:tensorflow:step = 39401, loss = 0.322542, precision = 0.953125 (2.703 sec) +Saved checkpoint after 101 epoch(s) to data/resnet20/checkpoints/00101... +INFO:tensorflow:global_step/sec: 31.2699 +INFO:tensorflow:step = 39501, loss = 0.345319, precision = 0.953125 (3.198 sec) +INFO:tensorflow:global_step/sec: 36.9965 +INFO:tensorflow:step = 39601, loss = 0.328086, precision = 0.953125 (2.703 sec) +INFO:tensorflow:global_step/sec: 36.9814 +INFO:tensorflow:step = 39701, loss = 0.406049, precision = 0.929688 (2.704 sec) +INFO:tensorflow:global_step/sec: 37.1213 +INFO:tensorflow:step = 39801, loss = 0.324695, precision = 0.96875 (2.694 sec) +Saved checkpoint after 102 epoch(s) to data/resnet20/checkpoints/00102... +INFO:tensorflow:global_step/sec: 31.3828 +INFO:tensorflow:step = 39901, loss = 0.305224, precision = 0.976562 (3.186 sec) +INFO:tensorflow:global_step/sec: 36.7605 +INFO:tensorflow:step = 40001, loss = 0.262923, precision = 0.976562 (2.721 sec) +INFO:tensorflow:global_step/sec: 36.9703 +INFO:tensorflow:step = 40101, loss = 0.275686, precision = 0.960938 (2.705 sec) +INFO:tensorflow:global_step/sec: 36.8725 +INFO:tensorflow:step = 40201, loss = 0.271022, precision = 0.976562 (2.712 sec) +Saved checkpoint after 103 epoch(s) to data/resnet20/checkpoints/00103... +INFO:tensorflow:global_step/sec: 31.2823 +INFO:tensorflow:step = 40301, loss = 0.267126, precision = 0.976562 (3.196 sec) +INFO:tensorflow:global_step/sec: 37.0214 +INFO:tensorflow:step = 40401, loss = 0.280038, precision = 0.96875 (2.701 sec) +INFO:tensorflow:global_step/sec: 36.7799 +INFO:tensorflow:step = 40501, loss = 0.335676, precision = 0.945312 (2.719 sec) +INFO:tensorflow:global_step/sec: 37.0614 +INFO:tensorflow:step = 40601, loss = 0.347046, precision = 0.960938 (2.698 sec) +Saved checkpoint after 104 epoch(s) to data/resnet20/checkpoints/00104... +INFO:tensorflow:global_step/sec: 31.2441 +INFO:tensorflow:step = 40701, loss = 0.294548, precision = 0.976562 (3.201 sec) +INFO:tensorflow:global_step/sec: 37.2249 +INFO:tensorflow:step = 40801, loss = 0.333558, precision = 0.96875 (2.686 sec) +INFO:tensorflow:global_step/sec: 36.9643 +INFO:tensorflow:step = 40901, loss = 0.312363, precision = 0.96875 (2.706 sec) +INFO:tensorflow:global_step/sec: 36.6529 +INFO:tensorflow:step = 41001, loss = 0.334994, precision = 0.945312 (2.728 sec) +Saved checkpoint after 105 epoch(s) to data/resnet20/checkpoints/00105... +INFO:tensorflow:global_step/sec: 31.3044 +INFO:tensorflow:step = 41101, loss = 0.384776, precision = 0.921875 (3.195 sec) +INFO:tensorflow:global_step/sec: 36.6972 +INFO:tensorflow:step = 41201, loss = 0.326737, precision = 0.96875 (2.724 sec) +INFO:tensorflow:global_step/sec: 36.8988 +INFO:tensorflow:step = 41301, loss = 0.300618, precision = 0.96875 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.8789 +INFO:tensorflow:step = 41401, loss = 0.265084, precision = 0.976562 (2.712 sec) +Saved checkpoint after 106 epoch(s) to data/resnet20/checkpoints/00106... +INFO:tensorflow:global_step/sec: 31.0524 +INFO:tensorflow:step = 41501, loss = 0.289211, precision = 0.96875 (3.220 sec) +INFO:tensorflow:global_step/sec: 36.9153 +INFO:tensorflow:step = 41601, loss = 0.326824, precision = 0.953125 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.8394 +INFO:tensorflow:step = 41701, loss = 0.239347, precision = 0.984375 (2.715 sec) +INFO:tensorflow:global_step/sec: 36.8682 +INFO:tensorflow:step = 41801, loss = 0.291423, precision = 0.96875 (2.712 sec) +Saved checkpoint after 107 epoch(s) to data/resnet20/checkpoints/00107... +INFO:tensorflow:global_step/sec: 31.3859 +INFO:tensorflow:step = 41901, loss = 0.275046, precision = 0.984375 (3.186 sec) +INFO:tensorflow:global_step/sec: 36.7811 +INFO:tensorflow:step = 42001, loss = 0.269973, precision = 0.96875 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.7764 +INFO:tensorflow:step = 42101, loss = 0.283336, precision = 0.945312 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.7815 +INFO:tensorflow:step = 42201, loss = 0.27514, precision = 0.945312 (2.719 sec) +Saved checkpoint after 108 epoch(s) to data/resnet20/checkpoints/00108... +INFO:tensorflow:global_step/sec: 31.362 +INFO:tensorflow:step = 42301, loss = 0.364667, precision = 0.929688 (3.189 sec) +INFO:tensorflow:global_step/sec: 36.945 +INFO:tensorflow:step = 42401, loss = 0.271091, precision = 0.96875 (2.706 sec) +INFO:tensorflow:global_step/sec: 36.9074 +INFO:tensorflow:step = 42501, loss = 0.290038, precision = 0.953125 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.8266 +INFO:tensorflow:step = 42601, loss = 0.240734, precision = 0.96875 (2.716 sec) +Saved checkpoint after 109 epoch(s) to data/resnet20/checkpoints/00109... +INFO:tensorflow:global_step/sec: 31.3518 +INFO:tensorflow:step = 42701, loss = 0.263061, precision = 0.984375 (3.189 sec) +INFO:tensorflow:global_step/sec: 37.1253 +INFO:tensorflow:step = 42801, loss = 0.288583, precision = 0.960938 (2.694 sec) +INFO:tensorflow:global_step/sec: 36.8694 +INFO:tensorflow:step = 42901, loss = 0.257009, precision = 0.953125 (2.712 sec) +INFO:tensorflow:global_step/sec: 36.8091 +INFO:tensorflow:step = 43001, loss = 0.209638, precision = 0.992188 (2.717 sec) +Saved checkpoint after 110 epoch(s) to data/resnet20/checkpoints/00110... +INFO:tensorflow:global_step/sec: 31.3033 +INFO:tensorflow:step = 43101, loss = 0.224229, precision = 0.984375 (3.195 sec) +INFO:tensorflow:global_step/sec: 36.8696 +INFO:tensorflow:step = 43201, loss = 0.330391, precision = 0.929688 (2.712 sec) +INFO:tensorflow:global_step/sec: 36.6798 +INFO:tensorflow:step = 43301, loss = 0.298487, precision = 0.96875 (2.726 sec) +Saved checkpoint after 111 epoch(s) to data/resnet20/checkpoints/00111... +INFO:tensorflow:global_step/sec: 31.5271 +INFO:tensorflow:step = 43401, loss = 0.284991, precision = 0.960938 (3.172 sec) +INFO:tensorflow:global_step/sec: 36.5845 +INFO:tensorflow:step = 43501, loss = 0.219251, precision = 0.984375 (2.733 sec) +INFO:tensorflow:global_step/sec: 36.8727 +INFO:tensorflow:step = 43601, loss = 0.271116, precision = 0.96875 (2.712 sec) +INFO:tensorflow:global_step/sec: 36.9733 +INFO:tensorflow:step = 43701, loss = 0.292971, precision = 0.960938 (2.705 sec) +Saved checkpoint after 112 epoch(s) to data/resnet20/checkpoints/00112... +INFO:tensorflow:global_step/sec: 31.5521 +INFO:tensorflow:step = 43801, loss = 0.286619, precision = 0.953125 (3.170 sec) +INFO:tensorflow:global_step/sec: 37.0313 +INFO:tensorflow:step = 43901, loss = 0.253975, precision = 0.953125 (2.700 sec) +INFO:tensorflow:global_step/sec: 37.0302 +INFO:tensorflow:step = 44001, loss = 0.329034, precision = 0.9375 (2.701 sec) +INFO:tensorflow:global_step/sec: 36.8875 +INFO:tensorflow:step = 44101, loss = 0.243183, precision = 0.96875 (2.711 sec) +Saved checkpoint after 113 epoch(s) to data/resnet20/checkpoints/00113... +INFO:tensorflow:global_step/sec: 31.4577 +INFO:tensorflow:step = 44201, loss = 0.243855, precision = 0.96875 (3.179 sec) +INFO:tensorflow:global_step/sec: 36.9249 +INFO:tensorflow:step = 44301, loss = 0.245649, precision = 0.976562 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.9983 +INFO:tensorflow:step = 44401, loss = 0.285695, precision = 0.96875 (2.703 sec) +INFO:tensorflow:global_step/sec: 37.0467 +INFO:tensorflow:step = 44501, loss = 0.292585, precision = 0.960938 (2.699 sec) +Saved checkpoint after 114 epoch(s) to data/resnet20/checkpoints/00114... +INFO:tensorflow:global_step/sec: 31.4473 +INFO:tensorflow:step = 44601, loss = 0.270987, precision = 0.960938 (3.180 sec) +INFO:tensorflow:global_step/sec: 36.9052 +INFO:tensorflow:step = 44701, loss = 0.3067, precision = 0.9375 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.9157 +INFO:tensorflow:step = 44801, loss = 0.268978, precision = 0.953125 (2.709 sec) +INFO:tensorflow:global_step/sec: 36.6316 +INFO:tensorflow:step = 44901, loss = 0.24786, precision = 0.953125 (2.730 sec) +Saved checkpoint after 115 epoch(s) to data/resnet20/checkpoints/00115... +INFO:tensorflow:global_step/sec: 32.035 +INFO:tensorflow:step = 45001, loss = 0.199062, precision = 0.992188 (3.122 sec) +INFO:tensorflow:global_step/sec: 36.9838 +INFO:tensorflow:step = 45101, loss = 0.211359, precision = 0.984375 (2.704 sec) +INFO:tensorflow:global_step/sec: 36.9727 +INFO:tensorflow:step = 45201, loss = 0.243219, precision = 0.96875 (2.705 sec) +INFO:tensorflow:global_step/sec: 37.0347 +INFO:tensorflow:step = 45301, loss = 0.218188, precision = 0.976562 (2.700 sec) +Saved checkpoint after 116 epoch(s) to data/resnet20/checkpoints/00116... +INFO:tensorflow:global_step/sec: 31.1801 +INFO:tensorflow:step = 45401, loss = 0.224084, precision = 0.976562 (3.208 sec) +INFO:tensorflow:global_step/sec: 36.8273 +INFO:tensorflow:step = 45501, loss = 0.207619, precision = 0.992188 (2.715 sec) +INFO:tensorflow:global_step/sec: 37.0329 +INFO:tensorflow:step = 45601, loss = 0.252193, precision = 0.96875 (2.700 sec) +INFO:tensorflow:global_step/sec: 37.0005 +INFO:tensorflow:step = 45701, loss = 0.25098, precision = 0.953125 (2.703 sec) +Saved checkpoint after 117 epoch(s) to data/resnet20/checkpoints/00117... +INFO:tensorflow:global_step/sec: 31.7821 +INFO:tensorflow:step = 45801, loss = 0.222372, precision = 0.96875 (3.147 sec) +INFO:tensorflow:global_step/sec: 36.8781 +INFO:tensorflow:step = 45901, loss = 0.195242, precision = 0.984375 (2.711 sec) +INFO:tensorflow:global_step/sec: 36.9566 +INFO:tensorflow:step = 46001, loss = 0.244559, precision = 0.960938 (2.706 sec) +INFO:tensorflow:global_step/sec: 37.1073 +INFO:tensorflow:step = 46101, loss = 0.232293, precision = 0.96875 (2.695 sec) +Saved checkpoint after 118 epoch(s) to data/resnet20/checkpoints/00118... +INFO:tensorflow:global_step/sec: 31.5879 +INFO:tensorflow:step = 46201, loss = 0.245145, precision = 0.96875 (3.166 sec) +INFO:tensorflow:global_step/sec: 36.7579 +INFO:tensorflow:step = 46301, loss = 0.291109, precision = 0.953125 (2.721 sec) +INFO:tensorflow:global_step/sec: 36.5946 +INFO:tensorflow:step = 46401, loss = 0.278449, precision = 0.945312 (2.733 sec) +INFO:tensorflow:global_step/sec: 36.9789 +INFO:tensorflow:step = 46501, loss = 0.284411, precision = 0.945312 (2.704 sec) +Saved checkpoint after 119 epoch(s) to data/resnet20/checkpoints/00119... +INFO:tensorflow:global_step/sec: 30.9015 +INFO:tensorflow:step = 46601, loss = 0.233784, precision = 0.984375 (3.237 sec) +INFO:tensorflow:global_step/sec: 37.0746 +INFO:tensorflow:step = 46701, loss = 0.306007, precision = 0.914062 (2.697 sec) +INFO:tensorflow:global_step/sec: 36.9272 +INFO:tensorflow:step = 46801, loss = 0.27014, precision = 0.945312 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.8803 +INFO:tensorflow:step = 46901, loss = 0.242079, precision = 0.960938 (2.712 sec) +Saved checkpoint after 120 epoch(s) to data/resnet20/checkpoints/00120... +INFO:tensorflow:global_step/sec: 31.0018 +INFO:tensorflow:step = 47001, loss = 0.204658, precision = 0.984375 (3.226 sec) +INFO:tensorflow:global_step/sec: 36.7403 +INFO:tensorflow:step = 47101, loss = 0.284226, precision = 0.960938 (2.721 sec) +INFO:tensorflow:global_step/sec: 37.0526 +INFO:tensorflow:step = 47201, loss = 0.241929, precision = 0.96875 (2.699 sec) +INFO:tensorflow:global_step/sec: 36.8529 +INFO:tensorflow:step = 47301, loss = 0.336192, precision = 0.9375 (2.713 sec) +Saved checkpoint after 121 epoch(s) to data/resnet20/checkpoints/00121... +INFO:tensorflow:global_step/sec: 31.7091 +INFO:tensorflow:step = 47401, loss = 0.248927, precision = 0.976562 (3.153 sec) +INFO:tensorflow:global_step/sec: 36.6762 +INFO:tensorflow:step = 47501, loss = 0.208962, precision = 0.984375 (2.727 sec) +INFO:tensorflow:global_step/sec: 36.9749 +INFO:tensorflow:step = 47601, loss = 0.280658, precision = 0.953125 (2.704 sec) +INFO:tensorflow:global_step/sec: 36.8177 +INFO:tensorflow:step = 47701, loss = 0.178619, precision = 0.992188 (2.716 sec) +Saved checkpoint after 122 epoch(s) to data/resnet20/checkpoints/00122... +INFO:tensorflow:global_step/sec: 31.4014 +INFO:tensorflow:step = 47801, loss = 0.238432, precision = 0.976562 (3.185 sec) +INFO:tensorflow:global_step/sec: 37.0089 +INFO:tensorflow:step = 47901, loss = 0.272315, precision = 0.960938 (2.702 sec) +INFO:tensorflow:global_step/sec: 37.0369 +INFO:tensorflow:step = 48001, loss = 0.290416, precision = 0.960938 (2.700 sec) +Saved checkpoint after 123 epoch(s) to data/resnet20/checkpoints/00123... +INFO:tensorflow:global_step/sec: 31.134 +INFO:tensorflow:step = 48101, loss = 0.219676, precision = 0.96875 (3.212 sec) +INFO:tensorflow:global_step/sec: 36.7794 +INFO:tensorflow:step = 48201, loss = 0.201908, precision = 0.984375 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.9889 +INFO:tensorflow:step = 48301, loss = 0.243259, precision = 0.96875 (2.704 sec) +INFO:tensorflow:global_step/sec: 36.813 +INFO:tensorflow:step = 48401, loss = 0.278864, precision = 0.960938 (2.716 sec) +Saved checkpoint after 124 epoch(s) to data/resnet20/checkpoints/00124... +INFO:tensorflow:global_step/sec: 31.2577 +INFO:tensorflow:step = 48501, loss = 0.205734, precision = 0.992188 (3.199 sec) +INFO:tensorflow:global_step/sec: 36.7794 +INFO:tensorflow:step = 48601, loss = 0.242061, precision = 0.976562 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.9583 +INFO:tensorflow:step = 48701, loss = 0.212079, precision = 0.976562 (2.705 sec) +INFO:tensorflow:global_step/sec: 37.0908 +INFO:tensorflow:step = 48801, loss = 0.214931, precision = 0.96875 (2.696 sec) +Saved checkpoint after 125 epoch(s) to data/resnet20/checkpoints/00125... +INFO:tensorflow:global_step/sec: 31.2824 +INFO:tensorflow:step = 48901, loss = 0.225347, precision = 0.96875 (3.196 sec) +INFO:tensorflow:global_step/sec: 36.9462 +INFO:tensorflow:step = 49001, loss = 0.208851, precision = 0.976562 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.8442 +INFO:tensorflow:step = 49101, loss = 0.279505, precision = 0.953125 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.6758 +INFO:tensorflow:step = 49201, loss = 0.213665, precision = 0.96875 (2.726 sec) +Saved checkpoint after 126 epoch(s) to data/resnet20/checkpoints/00126... +INFO:tensorflow:global_step/sec: 31.7597 +INFO:tensorflow:step = 49301, loss = 0.229086, precision = 0.960938 (3.149 sec) +INFO:tensorflow:global_step/sec: 36.9528 +INFO:tensorflow:step = 49401, loss = 0.272321, precision = 0.945312 (2.706 sec) +INFO:tensorflow:global_step/sec: 37.0025 +INFO:tensorflow:step = 49501, loss = 0.290302, precision = 0.945312 (2.703 sec) +INFO:tensorflow:global_step/sec: 36.9874 +INFO:tensorflow:step = 49601, loss = 0.236693, precision = 0.953125 (2.704 sec) +Saved checkpoint after 127 epoch(s) to data/resnet20/checkpoints/00127... +INFO:tensorflow:global_step/sec: 31.2795 +INFO:tensorflow:step = 49701, loss = 0.249367, precision = 0.976562 (3.197 sec) +INFO:tensorflow:global_step/sec: 36.8433 +INFO:tensorflow:step = 49801, loss = 0.25743, precision = 0.960938 (2.713 sec) +INFO:tensorflow:global_step/sec: 36.9768 +INFO:tensorflow:step = 49901, loss = 0.281806, precision = 0.953125 (2.705 sec) +INFO:tensorflow:global_step/sec: 36.9627 +INFO:tensorflow:step = 50001, loss = 0.277999, precision = 0.953125 (2.705 sec) +Saved checkpoint after 128 epoch(s) to data/resnet20/checkpoints/00128... +INFO:tensorflow:global_step/sec: 31.4105 +INFO:tensorflow:step = 50101, loss = 0.213722, precision = 0.976562 (3.184 sec) +INFO:tensorflow:global_step/sec: 36.897 +INFO:tensorflow:step = 50201, loss = 0.246598, precision = 0.953125 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.816 +INFO:tensorflow:step = 50301, loss = 0.291204, precision = 0.953125 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.9463 +INFO:tensorflow:step = 50401, loss = 0.216856, precision = 0.984375 (2.706 sec) +Saved checkpoint after 129 epoch(s) to data/resnet20/checkpoints/00129... +INFO:tensorflow:global_step/sec: 31.347 +INFO:tensorflow:step = 50501, loss = 0.244497, precision = 0.953125 (3.190 sec) +INFO:tensorflow:global_step/sec: 37.1466 +INFO:tensorflow:step = 50601, loss = 0.215796, precision = 0.976562 (2.693 sec) +INFO:tensorflow:global_step/sec: 37.014 +INFO:tensorflow:step = 50701, loss = 0.254267, precision = 0.945312 (2.701 sec) +INFO:tensorflow:global_step/sec: 37.0189 +INFO:tensorflow:step = 50801, loss = 0.336952, precision = 0.914062 (2.701 sec) +Saved checkpoint after 130 epoch(s) to data/resnet20/checkpoints/00130... +INFO:tensorflow:global_step/sec: 31.2133 +INFO:tensorflow:step = 50901, loss = 0.220866, precision = 0.96875 (3.204 sec) +INFO:tensorflow:global_step/sec: 36.8227 +INFO:tensorflow:step = 51001, loss = 0.25844, precision = 0.953125 (2.715 sec) +INFO:tensorflow:global_step/sec: 36.8217 +INFO:tensorflow:step = 51101, loss = 0.240979, precision = 0.960938 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.8672 +INFO:tensorflow:step = 51201, loss = 0.213336, precision = 0.96875 (2.712 sec) +Saved checkpoint after 131 epoch(s) to data/resnet20/checkpoints/00131... +INFO:tensorflow:global_step/sec: 31.4172 +INFO:tensorflow:step = 51301, loss = 0.222119, precision = 0.96875 (3.183 sec) +INFO:tensorflow:global_step/sec: 36.9911 +INFO:tensorflow:step = 51401, loss = 0.197003, precision = 0.984375 (2.703 sec) +INFO:tensorflow:global_step/sec: 36.9403 +INFO:tensorflow:step = 51501, loss = 0.24554, precision = 0.960938 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.8426 +INFO:tensorflow:step = 51601, loss = 0.257634, precision = 0.945312 (2.714 sec) +Saved checkpoint after 132 epoch(s) to data/resnet20/checkpoints/00132... +INFO:tensorflow:global_step/sec: 31.4857 +INFO:tensorflow:step = 51701, loss = 0.328382, precision = 0.929688 (3.176 sec) +INFO:tensorflow:global_step/sec: 36.5386 +INFO:tensorflow:step = 51801, loss = 0.202523, precision = 0.992188 (2.737 sec) +INFO:tensorflow:global_step/sec: 36.8102 +INFO:tensorflow:step = 51901, loss = 0.256058, precision = 0.953125 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.8603 +INFO:tensorflow:step = 52001, loss = 0.212155, precision = 0.984375 (2.713 sec) +Saved checkpoint after 133 epoch(s) to data/resnet20/checkpoints/00133... +INFO:tensorflow:global_step/sec: 30.7405 +INFO:tensorflow:step = 52101, loss = 0.191367, precision = 0.992188 (3.253 sec) +INFO:tensorflow:global_step/sec: 36.9331 +INFO:tensorflow:step = 52201, loss = 0.257525, precision = 0.953125 (2.708 sec) +INFO:tensorflow:global_step/sec: 37.1067 +INFO:tensorflow:step = 52301, loss = 0.359191, precision = 0.914062 (2.694 sec) +Saved checkpoint after 134 epoch(s) to data/resnet20/checkpoints/00134... +INFO:tensorflow:global_step/sec: 30.9817 +INFO:tensorflow:step = 52401, loss = 0.33814, precision = 0.9375 (3.229 sec) +INFO:tensorflow:global_step/sec: 36.7565 +INFO:tensorflow:step = 52501, loss = 0.205714, precision = 0.984375 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.9589 +INFO:tensorflow:step = 52601, loss = 0.222252, precision = 0.976562 (2.706 sec) +INFO:tensorflow:global_step/sec: 36.8724 +INFO:tensorflow:step = 52701, loss = 0.20525, precision = 0.96875 (2.712 sec) +Saved checkpoint after 135 epoch(s) to data/resnet20/checkpoints/00135... +INFO:tensorflow:global_step/sec: 30.981 +INFO:tensorflow:step = 52801, loss = 0.263434, precision = 0.953125 (3.228 sec) +INFO:tensorflow:global_step/sec: 36.9847 +INFO:tensorflow:step = 52901, loss = 0.278864, precision = 0.9375 (2.704 sec) +INFO:tensorflow:global_step/sec: 36.9385 +INFO:tensorflow:step = 53001, loss = 0.255519, precision = 0.976562 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.9375 +INFO:tensorflow:step = 53101, loss = 0.195565, precision = 0.976562 (2.708 sec) +Saved checkpoint after 136 epoch(s) to data/resnet20/checkpoints/00136... +INFO:tensorflow:global_step/sec: 31.2796 +INFO:tensorflow:step = 53201, loss = 0.214501, precision = 0.96875 (3.197 sec) +INFO:tensorflow:global_step/sec: 36.9939 +INFO:tensorflow:step = 53301, loss = 0.248778, precision = 0.960938 (2.703 sec) +INFO:tensorflow:global_step/sec: 36.8966 +INFO:tensorflow:step = 53401, loss = 0.211632, precision = 0.976562 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.9045 +INFO:tensorflow:step = 53501, loss = 0.183102, precision = 0.984375 (2.710 sec) +Saved checkpoint after 137 epoch(s) to data/resnet20/checkpoints/00137... +INFO:tensorflow:global_step/sec: 31.275 +INFO:tensorflow:step = 53601, loss = 0.188328, precision = 0.984375 (3.197 sec) +INFO:tensorflow:global_step/sec: 36.862 +INFO:tensorflow:step = 53701, loss = 0.184193, precision = 0.984375 (2.713 sec) +INFO:tensorflow:global_step/sec: 36.7944 +INFO:tensorflow:step = 53801, loss = 0.229776, precision = 0.953125 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.7788 +INFO:tensorflow:step = 53901, loss = 0.181482, precision = 0.992188 (2.718 sec) +Saved checkpoint after 138 epoch(s) to data/resnet20/checkpoints/00138... +INFO:tensorflow:global_step/sec: 30.2333 +INFO:tensorflow:step = 54001, loss = 0.231449, precision = 0.96875 (3.308 sec) +INFO:tensorflow:global_step/sec: 36.773 +INFO:tensorflow:step = 54101, loss = 0.217466, precision = 0.960938 (2.720 sec) +INFO:tensorflow:global_step/sec: 36.7581 +INFO:tensorflow:step = 54201, loss = 0.174513, precision = 0.984375 (2.720 sec) +INFO:tensorflow:global_step/sec: 36.7263 +INFO:tensorflow:step = 54301, loss = 0.181335, precision = 0.992188 (2.723 sec) +Saved checkpoint after 139 epoch(s) to data/resnet20/checkpoints/00139... +INFO:tensorflow:global_step/sec: 31.1983 +INFO:tensorflow:step = 54401, loss = 0.242516, precision = 0.96875 (3.205 sec) +INFO:tensorflow:global_step/sec: 36.6974 +INFO:tensorflow:step = 54501, loss = 0.178643, precision = 0.992188 (2.725 sec) +INFO:tensorflow:global_step/sec: 36.9489 +INFO:tensorflow:step = 54601, loss = 0.19085, precision = 0.984375 (2.706 sec) +INFO:tensorflow:global_step/sec: 37.049 +INFO:tensorflow:step = 54701, loss = 0.158843, precision = 1.0 (2.699 sec) +Saved checkpoint after 140 epoch(s) to data/resnet20/checkpoints/00140... +INFO:tensorflow:global_step/sec: 30.8603 +INFO:tensorflow:step = 54801, loss = 0.213294, precision = 0.96875 (3.240 sec) +INFO:tensorflow:global_step/sec: 36.578 +INFO:tensorflow:step = 54901, loss = 0.190153, precision = 0.984375 (2.734 sec) +INFO:tensorflow:global_step/sec: 36.9225 +INFO:tensorflow:step = 55001, loss = 0.188678, precision = 0.984375 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.9322 +INFO:tensorflow:step = 55101, loss = 0.169594, precision = 0.992188 (2.707 sec) +Saved checkpoint after 141 epoch(s) to data/resnet20/checkpoints/00141... +INFO:tensorflow:global_step/sec: 31.1221 +INFO:tensorflow:step = 55201, loss = 0.192435, precision = 0.984375 (3.213 sec) +INFO:tensorflow:global_step/sec: 36.753 +INFO:tensorflow:step = 55301, loss = 0.157594, precision = 1.0 (2.721 sec) +INFO:tensorflow:global_step/sec: 36.821 +INFO:tensorflow:step = 55401, loss = 0.174782, precision = 0.992188 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.5088 +INFO:tensorflow:step = 55501, loss = 0.193744, precision = 0.976562 (2.739 sec) +Saved checkpoint after 142 epoch(s) to data/resnet20/checkpoints/00142... +INFO:tensorflow:global_step/sec: 30.8529 +INFO:tensorflow:step = 55601, loss = 0.189681, precision = 0.976562 (3.241 sec) +INFO:tensorflow:global_step/sec: 36.8559 +INFO:tensorflow:step = 55701, loss = 0.198446, precision = 0.976562 (2.713 sec) +INFO:tensorflow:global_step/sec: 36.5762 +INFO:tensorflow:step = 55801, loss = 0.179215, precision = 0.992188 (2.734 sec) +INFO:tensorflow:global_step/sec: 36.9721 +INFO:tensorflow:step = 55901, loss = 0.172929, precision = 0.984375 (2.705 sec) +Saved checkpoint after 143 epoch(s) to data/resnet20/checkpoints/00143... +INFO:tensorflow:global_step/sec: 31.8138 +INFO:tensorflow:step = 56001, loss = 0.154535, precision = 0.992188 (3.143 sec) +INFO:tensorflow:global_step/sec: 36.6292 +INFO:tensorflow:step = 56101, loss = 0.167652, precision = 0.992188 (2.730 sec) +INFO:tensorflow:global_step/sec: 36.7471 +INFO:tensorflow:step = 56201, loss = 0.168188, precision = 0.992188 (2.721 sec) +INFO:tensorflow:global_step/sec: 36.8886 +INFO:tensorflow:step = 56301, loss = 0.173102, precision = 1.0 (2.711 sec) +Saved checkpoint after 144 epoch(s) to data/resnet20/checkpoints/00144... +INFO:tensorflow:global_step/sec: 31.0474 +INFO:tensorflow:step = 56401, loss = 0.174679, precision = 0.976562 (3.221 sec) +INFO:tensorflow:global_step/sec: 36.649 +INFO:tensorflow:step = 56501, loss = 0.167132, precision = 0.992188 (2.729 sec) +INFO:tensorflow:global_step/sec: 36.7389 +INFO:tensorflow:step = 56601, loss = 0.165803, precision = 0.992188 (2.722 sec) +Saved checkpoint after 145 epoch(s) to data/resnet20/checkpoints/00145... +INFO:tensorflow:global_step/sec: 30.8262 +INFO:tensorflow:step = 56701, loss = 0.155081, precision = 1.0 (3.244 sec) +INFO:tensorflow:global_step/sec: 36.7817 +INFO:tensorflow:step = 56801, loss = 0.176344, precision = 0.992188 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.8228 +INFO:tensorflow:step = 56901, loss = 0.158264, precision = 1.0 (2.716 sec) +INFO:tensorflow:global_step/sec: 37.0252 +INFO:tensorflow:step = 57001, loss = 0.169096, precision = 0.992188 (2.700 sec) +Saved checkpoint after 146 epoch(s) to data/resnet20/checkpoints/00146... +INFO:tensorflow:global_step/sec: 31.317 +INFO:tensorflow:step = 57101, loss = 0.15416, precision = 0.992188 (3.194 sec) +INFO:tensorflow:global_step/sec: 36.6095 +INFO:tensorflow:step = 57201, loss = 0.171888, precision = 0.976562 (2.731 sec) +INFO:tensorflow:global_step/sec: 36.9081 +INFO:tensorflow:step = 57301, loss = 0.223513, precision = 0.976562 (2.709 sec) +INFO:tensorflow:global_step/sec: 37.2465 +INFO:tensorflow:step = 57401, loss = 0.173455, precision = 0.984375 (2.685 sec) +Saved checkpoint after 147 epoch(s) to data/resnet20/checkpoints/00147... +INFO:tensorflow:global_step/sec: 31.9174 +INFO:tensorflow:step = 57501, loss = 0.186576, precision = 0.976562 (3.133 sec) +INFO:tensorflow:global_step/sec: 36.8636 +INFO:tensorflow:step = 57601, loss = 0.167696, precision = 0.984375 (2.712 sec) +INFO:tensorflow:global_step/sec: 36.991 +INFO:tensorflow:step = 57701, loss = 0.154017, precision = 0.992188 (2.703 sec) +INFO:tensorflow:global_step/sec: 36.7938 +INFO:tensorflow:step = 57801, loss = 0.172213, precision = 0.984375 (2.718 sec) +Saved checkpoint after 148 epoch(s) to data/resnet20/checkpoints/00148... +INFO:tensorflow:global_step/sec: 31.3033 +INFO:tensorflow:step = 57901, loss = 0.179394, precision = 0.984375 (3.195 sec) +INFO:tensorflow:global_step/sec: 36.3428 +INFO:tensorflow:step = 58001, loss = 0.158747, precision = 0.992188 (2.751 sec) +INFO:tensorflow:global_step/sec: 37.0096 +INFO:tensorflow:step = 58101, loss = 0.158177, precision = 1.0 (2.702 sec) +INFO:tensorflow:global_step/sec: 36.9222 +INFO:tensorflow:step = 58201, loss = 0.193593, precision = 0.976562 (2.708 sec) +Saved checkpoint after 149 epoch(s) to data/resnet20/checkpoints/00149... +INFO:tensorflow:global_step/sec: 31.0227 +INFO:tensorflow:step = 58301, loss = 0.166657, precision = 0.984375 (3.224 sec) +INFO:tensorflow:global_step/sec: 37.184 +INFO:tensorflow:step = 58401, loss = 0.143539, precision = 0.992188 (2.689 sec) +INFO:tensorflow:global_step/sec: 36.9354 +INFO:tensorflow:step = 58501, loss = 0.172238, precision = 0.976562 (2.707 sec) +INFO:tensorflow:global_step/sec: 37.0145 +INFO:tensorflow:step = 58601, loss = 0.143683, precision = 1.0 (2.702 sec) +Saved checkpoint after 150 epoch(s) to data/resnet20/checkpoints/00150... +INFO:tensorflow:global_step/sec: 31.6345 +INFO:tensorflow:step = 58701, loss = 0.178152, precision = 0.984375 (3.161 sec) +INFO:tensorflow:global_step/sec: 36.941 +INFO:tensorflow:step = 58801, loss = 0.151961, precision = 1.0 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.8261 +INFO:tensorflow:step = 58901, loss = 0.153354, precision = 0.992188 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.9061 +INFO:tensorflow:step = 59001, loss = 0.183845, precision = 0.984375 (2.710 sec) +Saved checkpoint after 151 epoch(s) to data/resnet20/checkpoints/00151... +INFO:tensorflow:global_step/sec: 31.2819 +INFO:tensorflow:step = 59101, loss = 0.173406, precision = 0.992188 (3.197 sec) +INFO:tensorflow:global_step/sec: 36.9844 +INFO:tensorflow:step = 59201, loss = 0.165577, precision = 0.992188 (2.704 sec) +INFO:tensorflow:global_step/sec: 37.1254 +INFO:tensorflow:step = 59301, loss = 0.199242, precision = 0.984375 (2.693 sec) +INFO:tensorflow:global_step/sec: 36.9452 +INFO:tensorflow:step = 59401, loss = 0.167567, precision = 0.992188 (2.707 sec) +Saved checkpoint after 152 epoch(s) to data/resnet20/checkpoints/00152... +INFO:tensorflow:global_step/sec: 31.1313 +INFO:tensorflow:step = 59501, loss = 0.160926, precision = 0.984375 (3.212 sec) +INFO:tensorflow:global_step/sec: 36.6439 +INFO:tensorflow:step = 59601, loss = 0.17254, precision = 0.984375 (2.729 sec) +INFO:tensorflow:global_step/sec: 36.9319 +INFO:tensorflow:step = 59701, loss = 0.166392, precision = 0.976562 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.8698 +INFO:tensorflow:step = 59801, loss = 0.140044, precision = 1.0 (2.712 sec) +Saved checkpoint after 153 epoch(s) to data/resnet20/checkpoints/00153... +INFO:tensorflow:global_step/sec: 31.1059 +INFO:tensorflow:step = 59901, loss = 0.159772, precision = 0.992188 (3.215 sec) +INFO:tensorflow:global_step/sec: 36.9475 +INFO:tensorflow:step = 60001, loss = 0.153492, precision = 0.992188 (2.706 sec) +INFO:tensorflow:global_step/sec: 37.0002 +INFO:tensorflow:step = 60101, loss = 0.192872, precision = 0.96875 (2.703 sec) +INFO:tensorflow:global_step/sec: 36.7269 +INFO:tensorflow:step = 60201, loss = 0.139455, precision = 1.0 (2.723 sec) +Saved checkpoint after 154 epoch(s) to data/resnet20/checkpoints/00154... +INFO:tensorflow:global_step/sec: 31.5265 +INFO:tensorflow:step = 60301, loss = 0.157112, precision = 0.984375 (3.172 sec) +INFO:tensorflow:global_step/sec: 36.8183 +INFO:tensorflow:step = 60401, loss = 0.17821, precision = 0.976562 (2.716 sec) +INFO:tensorflow:global_step/sec: 36.7986 +INFO:tensorflow:step = 60501, loss = 0.175759, precision = 0.976562 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.8388 +INFO:tensorflow:step = 60601, loss = 0.153689, precision = 0.992188 (2.715 sec) +Saved checkpoint after 155 epoch(s) to data/resnet20/checkpoints/00155... +INFO:tensorflow:global_step/sec: 31.3816 +INFO:tensorflow:step = 60701, loss = 0.162065, precision = 0.992188 (3.187 sec) +INFO:tensorflow:global_step/sec: 36.9318 +INFO:tensorflow:step = 60801, loss = 0.1697, precision = 0.992188 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.572 +INFO:tensorflow:step = 60901, loss = 0.192017, precision = 0.976562 (2.735 sec) +Saved checkpoint after 156 epoch(s) to data/resnet20/checkpoints/00156... +INFO:tensorflow:global_step/sec: 31.3735 +INFO:tensorflow:step = 61001, loss = 0.142642, precision = 1.0 (3.187 sec) +INFO:tensorflow:global_step/sec: 36.7801 +INFO:tensorflow:step = 61101, loss = 0.209056, precision = 0.976562 (2.719 sec) +INFO:tensorflow:global_step/sec: 36.9989 +INFO:tensorflow:step = 61201, loss = 0.186398, precision = 0.976562 (2.702 sec) +INFO:tensorflow:global_step/sec: 36.9355 +INFO:tensorflow:step = 61301, loss = 0.193452, precision = 0.992188 (2.708 sec) +Saved checkpoint after 157 epoch(s) to data/resnet20/checkpoints/00157... +INFO:tensorflow:global_step/sec: 31.143 +INFO:tensorflow:step = 61401, loss = 0.166453, precision = 0.984375 (3.211 sec) +INFO:tensorflow:global_step/sec: 36.525 +INFO:tensorflow:step = 61501, loss = 0.151864, precision = 0.992188 (2.737 sec) +INFO:tensorflow:global_step/sec: 37.2228 +INFO:tensorflow:step = 61601, loss = 0.153019, precision = 0.992188 (2.687 sec) +INFO:tensorflow:global_step/sec: 36.7995 +INFO:tensorflow:step = 61701, loss = 0.170912, precision = 0.984375 (2.717 sec) +Saved checkpoint after 158 epoch(s) to data/resnet20/checkpoints/00158... +INFO:tensorflow:global_step/sec: 30.9121 +INFO:tensorflow:step = 61801, loss = 0.149323, precision = 1.0 (3.236 sec) +INFO:tensorflow:global_step/sec: 36.4048 +INFO:tensorflow:step = 61901, loss = 0.154913, precision = 1.0 (2.747 sec) +INFO:tensorflow:global_step/sec: 37.0715 +INFO:tensorflow:step = 62001, loss = 0.159872, precision = 1.0 (2.697 sec) +INFO:tensorflow:global_step/sec: 36.9813 +INFO:tensorflow:step = 62101, loss = 0.160106, precision = 0.992188 (2.705 sec) +Saved checkpoint after 159 epoch(s) to data/resnet20/checkpoints/00159... +INFO:tensorflow:global_step/sec: 31.614 +INFO:tensorflow:step = 62201, loss = 0.161367, precision = 0.992188 (3.163 sec) +INFO:tensorflow:global_step/sec: 37.0079 +INFO:tensorflow:step = 62301, loss = 0.150774, precision = 1.0 (2.701 sec) +INFO:tensorflow:global_step/sec: 36.8571 +INFO:tensorflow:step = 62401, loss = 0.158603, precision = 0.984375 (2.713 sec) +INFO:tensorflow:global_step/sec: 36.9513 +INFO:tensorflow:step = 62501, loss = 0.170235, precision = 0.984375 (2.707 sec) +Saved checkpoint after 160 epoch(s) to data/resnet20/checkpoints/00160... +INFO:tensorflow:global_step/sec: 31.3185 +INFO:tensorflow:step = 62601, loss = 0.158299, precision = 0.992188 (3.193 sec) +INFO:tensorflow:global_step/sec: 36.7389 +INFO:tensorflow:step = 62701, loss = 0.147869, precision = 0.984375 (2.722 sec) +INFO:tensorflow:global_step/sec: 36.8318 +INFO:tensorflow:step = 62801, loss = 0.19054, precision = 0.96875 (2.715 sec) +INFO:tensorflow:global_step/sec: 37.1001 +INFO:tensorflow:step = 62901, loss = 0.145628, precision = 1.0 (2.695 sec) +Saved checkpoint after 161 epoch(s) to data/resnet20/checkpoints/00161... +INFO:tensorflow:global_step/sec: 31.7565 +INFO:tensorflow:step = 63001, loss = 0.186825, precision = 0.976562 (3.149 sec) +INFO:tensorflow:global_step/sec: 36.8884 +INFO:tensorflow:step = 63101, loss = 0.158597, precision = 0.992188 (2.711 sec) +INFO:tensorflow:global_step/sec: 36.9072 +INFO:tensorflow:step = 63201, loss = 0.142202, precision = 1.0 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.7642 +INFO:tensorflow:step = 63301, loss = 0.150852, precision = 1.0 (2.719 sec) +Saved checkpoint after 162 epoch(s) to data/resnet20/checkpoints/00162... +INFO:tensorflow:global_step/sec: 31.573 +INFO:tensorflow:step = 63401, loss = 0.145316, precision = 1.0 (3.168 sec) +INFO:tensorflow:global_step/sec: 37.0302 +INFO:tensorflow:step = 63501, loss = 0.141173, precision = 1.0 (2.701 sec) +INFO:tensorflow:global_step/sec: 36.5001 +INFO:tensorflow:step = 63601, loss = 0.128776, precision = 1.0 (2.739 sec) +INFO:tensorflow:global_step/sec: 36.6878 +INFO:tensorflow:step = 63701, loss = 0.13839, precision = 1.0 (2.726 sec) +Saved checkpoint after 163 epoch(s) to data/resnet20/checkpoints/00163... +INFO:tensorflow:global_step/sec: 31.2892 +INFO:tensorflow:step = 63801, loss = 0.149395, precision = 0.992188 (3.197 sec) +INFO:tensorflow:global_step/sec: 36.5395 +INFO:tensorflow:step = 63901, loss = 0.144278, precision = 1.0 (2.736 sec) +INFO:tensorflow:global_step/sec: 36.9086 +INFO:tensorflow:step = 64001, loss = 0.168153, precision = 0.984375 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.7696 +INFO:tensorflow:step = 64101, loss = 0.155173, precision = 0.992188 (2.720 sec) +Saved checkpoint after 164 epoch(s) to data/resnet20/checkpoints/00164... +INFO:tensorflow:global_step/sec: 31.0815 +INFO:tensorflow:step = 64201, loss = 0.149871, precision = 0.992188 (3.217 sec) +INFO:tensorflow:global_step/sec: 36.9111 +INFO:tensorflow:step = 64301, loss = 0.143736, precision = 1.0 (2.709 sec) +INFO:tensorflow:global_step/sec: 36.9 +INFO:tensorflow:step = 64401, loss = 0.13626, precision = 1.0 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.6204 +INFO:tensorflow:step = 64501, loss = 0.162311, precision = 0.992188 (2.730 sec) +Saved checkpoint after 165 epoch(s) to data/resnet20/checkpoints/00165... +INFO:tensorflow:global_step/sec: 31.6357 +INFO:tensorflow:step = 64601, loss = 0.169024, precision = 0.976562 (3.161 sec) +INFO:tensorflow:global_step/sec: 36.9702 +INFO:tensorflow:step = 64701, loss = 0.140007, precision = 0.992188 (2.704 sec) +INFO:tensorflow:global_step/sec: 37.1726 +INFO:tensorflow:step = 64801, loss = 0.162648, precision = 0.984375 (2.690 sec) +INFO:tensorflow:global_step/sec: 37.0903 +INFO:tensorflow:step = 64901, loss = 0.14493, precision = 1.0 (2.696 sec) +Saved checkpoint after 166 epoch(s) to data/resnet20/checkpoints/00166... +INFO:tensorflow:global_step/sec: 31.5275 +INFO:tensorflow:step = 65001, loss = 0.163912, precision = 0.992188 (3.172 sec) +INFO:tensorflow:global_step/sec: 36.8526 +INFO:tensorflow:step = 65101, loss = 0.155741, precision = 0.992188 (2.713 sec) +INFO:tensorflow:global_step/sec: 37.2503 +INFO:tensorflow:step = 65201, loss = 0.155643, precision = 0.992188 (2.685 sec) +Saved checkpoint after 167 epoch(s) to data/resnet20/checkpoints/00167... +INFO:tensorflow:global_step/sec: 31.4411 +INFO:tensorflow:step = 65301, loss = 0.143038, precision = 1.0 (3.181 sec) +INFO:tensorflow:global_step/sec: 36.8848 +INFO:tensorflow:step = 65401, loss = 0.140459, precision = 0.992188 (2.711 sec) +INFO:tensorflow:global_step/sec: 36.9205 +INFO:tensorflow:step = 65501, loss = 0.149279, precision = 0.992188 (2.708 sec) +INFO:tensorflow:global_step/sec: 36.9726 +INFO:tensorflow:step = 65601, loss = 0.149787, precision = 0.992188 (2.705 sec) +Saved checkpoint after 168 epoch(s) to data/resnet20/checkpoints/00168... +INFO:tensorflow:global_step/sec: 30.9995 +INFO:tensorflow:step = 65701, loss = 0.147315, precision = 0.992188 (3.226 sec) +INFO:tensorflow:global_step/sec: 36.7111 +INFO:tensorflow:step = 65801, loss = 0.138787, precision = 1.0 (2.724 sec) +INFO:tensorflow:global_step/sec: 36.7078 +INFO:tensorflow:step = 65901, loss = 0.132299, precision = 1.0 (2.724 sec) +INFO:tensorflow:global_step/sec: 36.9295 +INFO:tensorflow:step = 66001, loss = 0.157056, precision = 0.992188 (2.708 sec) +Saved checkpoint after 169 epoch(s) to data/resnet20/checkpoints/00169... +INFO:tensorflow:global_step/sec: 31.0517 +INFO:tensorflow:step = 66101, loss = 0.156413, precision = 0.984375 (3.220 sec) +INFO:tensorflow:global_step/sec: 36.88 +INFO:tensorflow:step = 66201, loss = 0.158351, precision = 0.992188 (2.712 sec) +INFO:tensorflow:global_step/sec: 36.9081 +INFO:tensorflow:step = 66301, loss = 0.166458, precision = 0.984375 (2.709 sec) +INFO:tensorflow:global_step/sec: 36.981 +INFO:tensorflow:step = 66401, loss = 0.15604, precision = 0.992188 (2.705 sec) +Saved checkpoint after 170 epoch(s) to data/resnet20/checkpoints/00170... +INFO:tensorflow:global_step/sec: 30.6276 +INFO:tensorflow:step = 66501, loss = 0.174248, precision = 0.976562 (3.265 sec) +INFO:tensorflow:global_step/sec: 36.8858 +INFO:tensorflow:step = 66601, loss = 0.142339, precision = 0.992188 (2.710 sec) +INFO:tensorflow:global_step/sec: 36.7903 +INFO:tensorflow:step = 66701, loss = 0.151185, precision = 0.992188 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.5078 +INFO:tensorflow:step = 66801, loss = 0.149078, precision = 0.992188 (2.739 sec) +Saved checkpoint after 171 epoch(s) to data/resnet20/checkpoints/00171... +INFO:tensorflow:global_step/sec: 31.213 +INFO:tensorflow:step = 66901, loss = 0.166579, precision = 0.992188 (3.204 sec) +INFO:tensorflow:global_step/sec: 36.7907 +INFO:tensorflow:step = 67001, loss = 0.167252, precision = 0.976562 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.7875 +INFO:tensorflow:step = 67101, loss = 0.167524, precision = 0.984375 (2.718 sec) +INFO:tensorflow:global_step/sec: 36.7903 +INFO:tensorflow:step = 67201, loss = 0.126259, precision = 1.0 (2.718 sec) +Saved checkpoint after 172 epoch(s) to data/resnet20/checkpoints/00172... +INFO:tensorflow:global_step/sec: 30.8482 +INFO:tensorflow:step = 67301, loss = 0.144421, precision = 0.992188 (3.242 sec) +INFO:tensorflow:global_step/sec: 36.8448 +INFO:tensorflow:step = 67401, loss = 0.162633, precision = 0.992188 (2.714 sec) +INFO:tensorflow:global_step/sec: 36.8026 +INFO:tensorflow:step = 67501, loss = 0.127064, precision = 1.0 (2.717 sec) +INFO:tensorflow:global_step/sec: 36.4812 +INFO:tensorflow:step = 67601, loss = 0.145251, precision = 1.0 (2.741 sec) +Saved checkpoint after 173 epoch(s) to data/resnet20/checkpoints/00173... +INFO:tensorflow:global_step/sec: 31.2214 +INFO:tensorflow:step = 67701, loss = 0.146114, precision = 1.0 (3.203 sec) +INFO:tensorflow:global_step/sec: 36.9656 +INFO:tensorflow:step = 67801, loss = 0.147929, precision = 0.984375 (2.705 sec) +INFO:tensorflow:global_step/sec: 36.8946 +INFO:tensorflow:step = 67901, loss = 0.133098, precision = 1.0 (2.711 sec) +INFO:tensorflow:global_step/sec: 36.7724 +INFO:tensorflow:step = 68001, loss = 0.147861, precision = 1.0 (2.719 sec) +Saved checkpoint after 174 epoch(s) to data/resnet20/checkpoints/00174... +INFO:tensorflow:global_step/sec: 31.2306 +INFO:tensorflow:step = 68101, loss = 0.136662, precision = 1.0 (3.202 sec) +INFO:tensorflow:global_step/sec: 36.9335 +INFO:tensorflow:step = 68201, loss = 0.179864, precision = 0.984375 (2.707 sec) +INFO:tensorflow:global_step/sec: 36.7134 +INFO:tensorflow:step = 68301, loss = 0.15277, precision = 0.984375 (2.724 sec) +INFO:tensorflow:global_step/sec: 36.8337 +INFO:tensorflow:step = 68401, loss = 0.145477, precision = 0.992188 (2.715 sec) +Saved checkpoint after 175 epoch(s) to data/resnet20/checkpoints/00175... +INFO:tensorflow:global_step/sec: 31.0402 +INFO:tensorflow:step = 68501, loss = 0.156269, precision = 0.984375 (3.222 sec) +INFO:tensorflow:global_step/sec: 36.6312 +INFO:tensorflow:step = 68601, loss = 0.136135, precision = 0.992188 (2.730 sec) +INFO:tensorflow:global_step/sec: 36.7241 +INFO:tensorflow:step = 68701, loss = 0.159448, precision = 0.984375 (2.723 sec) +INFO:tensorflow:global_step/sec: 36.867 +INFO:tensorflow:step = 68801, loss = 0.134074, precision = 0.992188 (2.712 sec) +Saved checkpoint after 176 epoch(s) to data/resnet20/checkpoints/00176... +INFO:tensorflow:global_step/sec: 31.1081 +INFO:tensorflow:step = 68901, loss = 0.123306, precision = 1.0 (3.214 sec) +INFO:tensorflow:global_step/sec: 36.8066 +INFO:tensorflow:step = 69001, loss = 0.129245, precision = 1.0 (2.717 sec) +INFO:tensorflow:global_step/sec: 36.8313 +INFO:tensorflow:step = 69101, loss = 0.153069, precision = 0.992188 (2.715 sec) +INFO:tensorflow:global_step/sec: 37.0659 +INFO:tensorflow:step = 69201, loss = 0.157617, precision = 0.992188 (2.697 sec) +Saved checkpoint after 177 epoch(s) to data/resnet20/checkpoints/00177... +INFO:tensorflow:global_step/sec: 31.2679 +INFO:tensorflow:step = 69301, loss = 0.137965, precision = 1.0 (3.199 sec) +INFO:tensorflow:global_step/sec: 37.0708 +INFO:tensorflow:step = 69401, loss = 0.127319, precision = 1.0 (2.697 sec) +INFO:tensorflow:global_step/sec: 36.7984 +INFO:tensorflow:step = 69501, loss = 0.127831, precision = 1.0 (2.718 sec) +Saved checkpoint after 178 epoch(s) to data/resnet20/checkpoints/00178... +INFO:tensorflow:global_step/sec: 30.8766 +INFO:tensorflow:step = 69601, loss = 0.13363, precision = 1.0 (3.239 sec) +INFO:tensorflow:global_step/sec: 36.8395 +INFO:tensorflow:step = 69701, loss = 0.133293, precision = 1.0 (2.715 sec) +INFO:tensorflow:global_step/sec: 36.7147 +INFO:tensorflow:step = 69801, loss = 0.163918, precision = 0.976562 (2.724 sec) +INFO:tensorflow:global_step/sec: 36.7929 +INFO:tensorflow:step = 69901, loss = 0.138131, precision = 1.0 (2.718 sec) +Saved checkpoint after 179 epoch(s) to data/resnet20/checkpoints/00179... +INFO:tensorflow:global_step/sec: 31.2248 +INFO:tensorflow:step = 70001, loss = 0.139061, precision = 1.0 (3.203 sec) +INFO:tensorflow:global_step/sec: 37.1066 +INFO:tensorflow:step = 70101, loss = 0.138351, precision = 1.0 (2.695 sec) +INFO:tensorflow:global_step/sec: 37.0827 +INFO:tensorflow:step = 70201, loss = 0.134516, precision = 1.0 (2.696 sec) +INFO:tensorflow:global_step/sec: 36.7449 +INFO:tensorflow:step = 70301, loss = 0.133771, precision = 0.992188 (2.722 sec) +Saved checkpoint after 180 epoch(s) to data/resnet20/checkpoints/00180... +INFO:tensorflow:global_step/sec: 31.5135 +INFO:tensorflow:step = 70401, loss = 0.135167, precision = 1.0 (3.173 sec) +INFO:tensorflow:global_step/sec: 37.004 +INFO:tensorflow:step = 70501, loss = 0.145283, precision = 0.992188 (2.702 sec) +INFO:tensorflow:global_step/sec: 36.9547 +INFO:tensorflow:step = 70601, loss = 0.129595, precision = 1.0 (2.706 sec) +INFO:tensorflow:global_step/sec: 37.1301 +INFO:tensorflow:step = 70701, loss = 0.123746, precision = 1.0 (2.694 sec) +Saved checkpoint after 181 epoch(s) to data/resnet20/checkpoints/00181... diff --git a/tensorflow/CIFAR10/logs/1p100_dawn/resnet56_train.log b/tensorflow/CIFAR10/logs/1p100_dawn/resnet56_train.log new file mode 100644 index 0000000..a0a7c30 --- /dev/null +++ b/tensorflow/CIFAR10/logs/1p100_dawn/resnet56_train.log @@ -0,0 +1,1835 @@ +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 0 +-device_regexes .* +-order_by name +-account_type_regexes _trainable_variables +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select params +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (--/850.99k params) + init/init_conv/DW (3x3x3x16, 432/432 params) + logit/DW (64x10, 640/640 params) + logit/biases (10, 10/10 params) + unit_1_0/shared_activation/init_bn/beta (16, 16/16 params) + unit_1_0/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_0/sub2/bn2/beta (16, 16/16 params) + unit_1_0/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_1/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_1/sub2/bn2/beta (16, 16/16 params) + unit_1_1/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_2/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_2/sub2/bn2/beta (16, 16/16 params) + unit_1_2/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_3/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_3/sub2/bn2/beta (16, 16/16 params) + unit_1_3/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_4/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_4/sub2/bn2/beta (16, 16/16 params) + unit_1_4/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_5/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_5/sub2/bn2/beta (16, 16/16 params) + unit_1_5/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_6/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_6/sub2/bn2/beta (16, 16/16 params) + unit_1_6/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_7/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_7/sub2/bn2/beta (16, 16/16 params) + unit_1_7/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_1_8/sub1/conv1/DW (3x3x16x16, 2.30k/2.30k params) + unit_1_8/sub2/bn2/beta (16, 16/16 params) + unit_1_8/sub2/conv2/DW (3x3x16x16, 2.30k/2.30k params) + unit_2_0/residual_only_activation/init_bn/beta (16, 16/16 params) + unit_2_0/sub1/conv1/DW (3x3x16x32, 4.61k/4.61k params) + unit_2_0/sub2/bn2/beta (32, 32/32 params) + unit_2_0/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_1/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_1/sub2/bn2/beta (32, 32/32 params) + unit_2_1/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_2/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_2/sub2/bn2/beta (32, 32/32 params) + unit_2_2/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_3/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_3/sub2/bn2/beta (32, 32/32 params) + unit_2_3/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_4/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_4/sub2/bn2/beta (32, 32/32 params) + unit_2_4/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_5/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_5/sub2/bn2/beta (32, 32/32 params) + unit_2_5/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_6/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_6/sub2/bn2/beta (32, 32/32 params) + unit_2_6/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_7/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_7/sub2/bn2/beta (32, 32/32 params) + unit_2_7/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_2_8/sub1/conv1/DW (3x3x32x32, 9.22k/9.22k params) + unit_2_8/sub2/bn2/beta (32, 32/32 params) + unit_2_8/sub2/conv2/DW (3x3x32x32, 9.22k/9.22k params) + unit_3_0/residual_only_activation/init_bn/beta (32, 32/32 params) + unit_3_0/sub1/conv1/DW (3x3x32x64, 18.43k/18.43k params) + unit_3_0/sub2/bn2/beta (64, 64/64 params) + unit_3_0/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_1/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_1/sub2/bn2/beta (64, 64/64 params) + unit_3_1/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_2/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_2/sub2/bn2/beta (64, 64/64 params) + unit_3_2/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_3/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_3/sub2/bn2/beta (64, 64/64 params) + unit_3_3/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_4/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_4/sub2/bn2/beta (64, 64/64 params) + unit_3_4/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_5/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_5/sub2/bn2/beta (64, 64/64 params) + unit_3_5/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_6/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_6/sub2/bn2/beta (64, 64/64 params) + unit_3_6/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_7/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_7/sub2/bn2/beta (64, 64/64 params) + unit_3_7/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/residual_only_activation/init_bn/beta (64, 64/64 params) + unit_3_8/sub1/conv1/DW (3x3x64x64, 36.86k/36.86k params) + unit_3_8/sub2/bn2/beta (64, 64/64 params) + unit_3_8/sub2/conv2/DW (3x3x64x64, 36.86k/36.86k params) + unit_last/final_bn/beta (64, 64/64 params) + +======================End of Report========================== +Parsing GraphDef... +Parsing OpLog... +Preparing Views... + +=========================Options============================= +-max_depth 10000 +-min_bytes 0 +-min_micros 0 +-min_params 0 +-min_float_ops 1 +-device_regexes .* +-order_by float_ops +-account_type_regexes .* +-start_name_regexes .* +-trim_name_regexes +-show_name_regexes .* +-hide_name_regexes +-account_displayed_op_only true +-select float_ops +-output stdout: + +==================Model Analysis Report====================== +_TFProfRoot (0/32.12b flops) + unit_3_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_3_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_3/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_3/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_4/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_4/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_5/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_5/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_6/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_6/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_7/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_7/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_1_8/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_1_8/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_0/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_1/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_1/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_2_2/sub1/conv1/Conv2D (603.98m/603.98m flops) + unit_2_2/sub2/conv2/Conv2D (603.98m/603.98m flops) + unit_3_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + unit_2_0/sub1/conv1/Conv2D (301.99m/301.99m flops) + init/init_conv/Conv2D (113.25m/113.25m flops) + logit/xw_plus_b (1.28k/165.12k flops) + logit/xw_plus_b/MatMul (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul_1 (163.84k/163.84k flops) + gradients/logit/xw_plus_b/MatMul_grad/MatMul (163.84k/163.84k flops) + +======================End of Report========================== +2017-08-03 00:43:43.457373: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: +name: Tesla P100-PCIE-16GB +major: 6 minor: 0 memoryClockRate (GHz) 1.3285 +pciBusID 0000:05:00.0 +Total memory: 15.89GiB +Free memory: 15.61GiB +2017-08-03 00:43:43.457452: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 +2017-08-03 00:43:43.457465: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y +2017-08-03 00:43:43.457490: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:05:00.0) +2017-08-03 00:43:44.071504: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-08-03 00:43:44.071582: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 56 visible devices +2017-08-03 00:43:44.092229: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x66670a0 executing computations on platform Host. Devices: +2017-08-03 00:43:44.092300: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): , +2017-08-03 00:43:44.092677: I tensorflow/compiler/xla/service/platform_util.cc:58] platform CUDA present with 1 visible devices +2017-08-03 00:43:44.092700: I tensorflow/compiler/xla/service/platform_util.cc:58] platform Host present with 56 visible devices +2017-08-03 00:43:44.110106: I tensorflow/compiler/xla/service/service.cc:198] XLA service 0x65a1d40 executing computations on platform CUDA. Devices: +2017-08-03 00:43:44.110194: I tensorflow/compiler/xla/service/service.cc:206] StreamExecutor device (0): Tesla P100-PCIE-16GB, Compute Capability 6.0 +INFO:tensorflow:step = 1, loss = 3.3552, precision = 0.125 +INFO:tensorflow:global_step/sec: 13.6052 +INFO:tensorflow:step = 101, loss = 2.76251, precision = 0.390625 (7.351 sec) +INFO:tensorflow:global_step/sec: 14.5254 +INFO:tensorflow:step = 201, loss = 2.41339, precision = 0.46875 (6.884 sec) +INFO:tensorflow:global_step/sec: 14.5226 +INFO:tensorflow:step = 301, loss = 2.34346, precision = 0.5 (6.886 sec) +total_params: 850986 +Saved checkpoint after 1 epoch(s) to data/resnet56/checkpoints/00001... +INFO:tensorflow:global_step/sec: 12.4344 +INFO:tensorflow:step = 401, loss = 2.69148, precision = 0.3125 (8.043 sec) +INFO:tensorflow:global_step/sec: 14.4498 +INFO:tensorflow:step = 501, loss = 2.05377, precision = 0.546875 (6.921 sec) +INFO:tensorflow:global_step/sec: 14.4428 +INFO:tensorflow:step = 601, loss = 1.98049, precision = 0.570312 (6.924 sec) +INFO:tensorflow:global_step/sec: 14.3681 +INFO:tensorflow:step = 701, loss = 1.7523, precision = 0.648438 (6.960 sec) +Saved checkpoint after 2 epoch(s) to data/resnet56/checkpoints/00002... +INFO:tensorflow:global_step/sec: 12.5131 +INFO:tensorflow:step = 801, loss = 1.86466, precision = 0.609375 (7.991 sec) +INFO:tensorflow:global_step/sec: 14.3161 +INFO:tensorflow:step = 901, loss = 1.78602, precision = 0.625 (6.985 sec) +INFO:tensorflow:global_step/sec: 14.4405 +INFO:tensorflow:step = 1001, loss = 1.42481, precision = 0.71875 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.3594 +INFO:tensorflow:step = 1101, loss = 1.29268, precision = 0.78125 (6.964 sec) +Saved checkpoint after 3 epoch(s) to data/resnet56/checkpoints/00003... +INFO:tensorflow:global_step/sec: 12.5221 +INFO:tensorflow:step = 1201, loss = 1.39661, precision = 0.695312 (7.986 sec) +INFO:tensorflow:global_step/sec: 14.4142 +INFO:tensorflow:step = 1301, loss = 1.29177, precision = 0.726562 (6.937 sec) +INFO:tensorflow:global_step/sec: 14.4152 +INFO:tensorflow:step = 1401, loss = 1.20355, precision = 0.789062 (6.937 sec) +INFO:tensorflow:global_step/sec: 14.3085 +INFO:tensorflow:step = 1501, loss = 1.19617, precision = 0.757812 (6.989 sec) +Saved checkpoint after 4 epoch(s) to data/resnet56/checkpoints/00004... +INFO:tensorflow:global_step/sec: 12.5435 +INFO:tensorflow:step = 1601, loss = 1.01513, precision = 0.851562 (7.972 sec) +INFO:tensorflow:global_step/sec: 14.4 +INFO:tensorflow:step = 1701, loss = 1.11359, precision = 0.757812 (6.944 sec) +INFO:tensorflow:global_step/sec: 14.4028 +INFO:tensorflow:step = 1801, loss = 0.934253, precision = 0.84375 (6.943 sec) +INFO:tensorflow:global_step/sec: 14.3632 +INFO:tensorflow:step = 1901, loss = 1.10071, precision = 0.75 (6.962 sec) +Saved checkpoint after 5 epoch(s) to data/resnet56/checkpoints/00005... +INFO:tensorflow:global_step/sec: 12.2895 +INFO:tensorflow:step = 2001, loss = 0.998392, precision = 0.796875 (8.137 sec) +INFO:tensorflow:global_step/sec: 14.4855 +INFO:tensorflow:step = 2101, loss = 0.989184, precision = 0.789062 (6.903 sec) +INFO:tensorflow:global_step/sec: 14.3715 +INFO:tensorflow:step = 2201, loss = 0.857736, precision = 0.84375 (6.959 sec) +INFO:tensorflow:global_step/sec: 14.3295 +INFO:tensorflow:step = 2301, loss = 0.903838, precision = 0.820312 (6.979 sec) +Saved checkpoint after 6 epoch(s) to data/resnet56/checkpoints/00006... +INFO:tensorflow:global_step/sec: 12.447 +INFO:tensorflow:step = 2401, loss = 0.988178, precision = 0.8125 (8.034 sec) +INFO:tensorflow:global_step/sec: 14.3911 +INFO:tensorflow:step = 2501, loss = 0.745362, precision = 0.828125 (6.948 sec) +INFO:tensorflow:global_step/sec: 14.3793 +INFO:tensorflow:step = 2601, loss = 0.920055, precision = 0.734375 (6.955 sec) +INFO:tensorflow:global_step/sec: 14.3194 +INFO:tensorflow:step = 2701, loss = 0.889815, precision = 0.8125 (6.984 sec) +Saved checkpoint after 7 epoch(s) to data/resnet56/checkpoints/00007... +INFO:tensorflow:global_step/sec: 12.3964 +INFO:tensorflow:step = 2801, loss = 0.748543, precision = 0.84375 (8.067 sec) +INFO:tensorflow:global_step/sec: 14.3632 +INFO:tensorflow:step = 2901, loss = 0.861842, precision = 0.804688 (6.962 sec) +INFO:tensorflow:global_step/sec: 14.3223 +INFO:tensorflow:step = 3001, loss = 1.16534, precision = 0.710938 (6.982 sec) +INFO:tensorflow:global_step/sec: 14.3593 +INFO:tensorflow:step = 3101, loss = 0.689368, precision = 0.851562 (6.964 sec) +Saved checkpoint after 8 epoch(s) to data/resnet56/checkpoints/00008... +INFO:tensorflow:global_step/sec: 12.4554 +INFO:tensorflow:step = 3201, loss = 0.88712, precision = 0.8125 (8.029 sec) +INFO:tensorflow:global_step/sec: 14.3993 +INFO:tensorflow:step = 3301, loss = 0.788028, precision = 0.796875 (6.945 sec) +INFO:tensorflow:global_step/sec: 14.3636 +INFO:tensorflow:step = 3401, loss = 0.853449, precision = 0.773438 (6.962 sec) +INFO:tensorflow:global_step/sec: 14.3316 +INFO:tensorflow:step = 3501, loss = 0.711074, precision = 0.851562 (6.978 sec) +Saved checkpoint after 9 epoch(s) to data/resnet56/checkpoints/00009... +INFO:tensorflow:global_step/sec: 12.5099 +INFO:tensorflow:step = 3601, loss = 0.806741, precision = 0.835938 (7.994 sec) +INFO:tensorflow:global_step/sec: 14.3616 +INFO:tensorflow:step = 3701, loss = 0.684088, precision = 0.859375 (6.963 sec) +INFO:tensorflow:global_step/sec: 14.3965 +INFO:tensorflow:step = 3801, loss = 0.868966, precision = 0.789062 (6.946 sec) +INFO:tensorflow:global_step/sec: 14.3709 +INFO:tensorflow:step = 3901, loss = 0.799195, precision = 0.828125 (6.958 sec) +Saved checkpoint after 10 epoch(s) to data/resnet56/checkpoints/00010... +INFO:tensorflow:global_step/sec: 12.4918 +INFO:tensorflow:step = 4001, loss = 0.858795, precision = 0.8125 (8.005 sec) +INFO:tensorflow:global_step/sec: 14.3276 +INFO:tensorflow:step = 4101, loss = 0.778481, precision = 0.828125 (6.979 sec) +INFO:tensorflow:global_step/sec: 14.3865 +INFO:tensorflow:step = 4201, loss = 0.73516, precision = 0.820312 (6.951 sec) +Saved checkpoint after 11 epoch(s) to data/resnet56/checkpoints/00011... +INFO:tensorflow:global_step/sec: 12.4907 +INFO:tensorflow:step = 4301, loss = 0.685631, precision = 0.875 (8.006 sec) +INFO:tensorflow:global_step/sec: 14.3492 +INFO:tensorflow:step = 4401, loss = 0.859196, precision = 0.789062 (6.969 sec) +INFO:tensorflow:global_step/sec: 14.3388 +INFO:tensorflow:step = 4501, loss = 0.775528, precision = 0.8125 (6.974 sec) +INFO:tensorflow:global_step/sec: 14.3523 +INFO:tensorflow:step = 4601, loss = 0.638044, precision = 0.859375 (6.968 sec) +Saved checkpoint after 12 epoch(s) to data/resnet56/checkpoints/00012... +INFO:tensorflow:global_step/sec: 12.4748 +INFO:tensorflow:step = 4701, loss = 0.854688, precision = 0.84375 (8.016 sec) +INFO:tensorflow:global_step/sec: 14.3877 +INFO:tensorflow:step = 4801, loss = 0.710855, precision = 0.859375 (6.950 sec) +INFO:tensorflow:global_step/sec: 14.3513 +INFO:tensorflow:step = 4901, loss = 0.642373, precision = 0.867188 (6.968 sec) +INFO:tensorflow:global_step/sec: 14.4239 +INFO:tensorflow:step = 5001, loss = 0.729918, precision = 0.835938 (6.933 sec) +Saved checkpoint after 13 epoch(s) to data/resnet56/checkpoints/00013... +INFO:tensorflow:global_step/sec: 12.4547 +INFO:tensorflow:step = 5101, loss = 0.843776, precision = 0.796875 (8.029 sec) +INFO:tensorflow:global_step/sec: 14.3727 +INFO:tensorflow:step = 5201, loss = 0.762775, precision = 0.835938 (6.958 sec) +INFO:tensorflow:global_step/sec: 14.4344 +INFO:tensorflow:step = 5301, loss = 0.686557, precision = 0.835938 (6.928 sec) +INFO:tensorflow:global_step/sec: 14.3919 +INFO:tensorflow:step = 5401, loss = 0.801012, precision = 0.789062 (6.948 sec) +Saved checkpoint after 14 epoch(s) to data/resnet56/checkpoints/00014... +INFO:tensorflow:global_step/sec: 12.4875 +INFO:tensorflow:step = 5501, loss = 0.638018, precision = 0.882812 (8.008 sec) +INFO:tensorflow:global_step/sec: 14.3914 +INFO:tensorflow:step = 5601, loss = 0.757734, precision = 0.804688 (6.948 sec) +INFO:tensorflow:global_step/sec: 14.3522 +INFO:tensorflow:step = 5701, loss = 0.678436, precision = 0.8125 (6.968 sec) +INFO:tensorflow:global_step/sec: 14.4127 +INFO:tensorflow:step = 5801, loss = 0.686967, precision = 0.867188 (6.938 sec) +Saved checkpoint after 15 epoch(s) to data/resnet56/checkpoints/00015... +INFO:tensorflow:global_step/sec: 12.4497 +INFO:tensorflow:step = 5901, loss = 0.809195, precision = 0.835938 (8.032 sec) +INFO:tensorflow:global_step/sec: 14.3667 +INFO:tensorflow:step = 6001, loss = 0.696595, precision = 0.867188 (6.960 sec) +INFO:tensorflow:global_step/sec: 14.3827 +INFO:tensorflow:step = 6101, loss = 0.734481, precision = 0.828125 (6.953 sec) +INFO:tensorflow:global_step/sec: 14.406 +INFO:tensorflow:step = 6201, loss = 0.705985, precision = 0.835938 (6.941 sec) +Saved checkpoint after 16 epoch(s) to data/resnet56/checkpoints/00016... +INFO:tensorflow:global_step/sec: 12.5358 +INFO:tensorflow:step = 6301, loss = 0.733058, precision = 0.820312 (7.977 sec) +INFO:tensorflow:global_step/sec: 14.4213 +INFO:tensorflow:step = 6401, loss = 0.634602, precision = 0.882812 (6.934 sec) +INFO:tensorflow:global_step/sec: 14.382 +INFO:tensorflow:step = 6501, loss = 0.850343, precision = 0.8125 (6.953 sec) +INFO:tensorflow:global_step/sec: 14.3785 +INFO:tensorflow:step = 6601, loss = 0.749046, precision = 0.867188 (6.955 sec) +Saved checkpoint after 17 epoch(s) to data/resnet56/checkpoints/00017... +INFO:tensorflow:global_step/sec: 12.4715 +INFO:tensorflow:step = 6701, loss = 0.597499, precision = 0.882812 (8.019 sec) +INFO:tensorflow:global_step/sec: 14.4009 +INFO:tensorflow:step = 6801, loss = 0.727188, precision = 0.804688 (6.944 sec) +INFO:tensorflow:global_step/sec: 14.3834 +INFO:tensorflow:step = 6901, loss = 0.687772, precision = 0.882812 (6.952 sec) +INFO:tensorflow:global_step/sec: 14.363 +INFO:tensorflow:step = 7001, loss = 0.762512, precision = 0.8125 (6.962 sec) +Saved checkpoint after 18 epoch(s) to data/resnet56/checkpoints/00018... +INFO:tensorflow:global_step/sec: 12.3255 +INFO:tensorflow:step = 7101, loss = 0.558911, precision = 0.914062 (8.113 sec) +INFO:tensorflow:global_step/sec: 14.434 +INFO:tensorflow:step = 7201, loss = 0.70711, precision = 0.84375 (6.928 sec) +INFO:tensorflow:global_step/sec: 14.4405 +INFO:tensorflow:step = 7301, loss = 0.707202, precision = 0.835938 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.3747 +INFO:tensorflow:step = 7401, loss = 0.681872, precision = 0.851562 (6.957 sec) +Saved checkpoint after 19 epoch(s) to data/resnet56/checkpoints/00019... +INFO:tensorflow:global_step/sec: 12.4588 +INFO:tensorflow:step = 7501, loss = 0.619319, precision = 0.890625 (8.027 sec) +INFO:tensorflow:global_step/sec: 14.359 +INFO:tensorflow:step = 7601, loss = 0.709476, precision = 0.84375 (6.964 sec) +INFO:tensorflow:global_step/sec: 14.3682 +INFO:tensorflow:step = 7701, loss = 0.70165, precision = 0.890625 (6.960 sec) +INFO:tensorflow:global_step/sec: 14.3937 +INFO:tensorflow:step = 7801, loss = 0.640967, precision = 0.875 (6.947 sec) +Saved checkpoint after 20 epoch(s) to data/resnet56/checkpoints/00020... +INFO:tensorflow:global_step/sec: 12.4449 +INFO:tensorflow:step = 7901, loss = 0.550504, precision = 0.90625 (8.036 sec) +INFO:tensorflow:global_step/sec: 14.4241 +INFO:tensorflow:step = 8001, loss = 0.6708, precision = 0.875 (6.933 sec) +INFO:tensorflow:global_step/sec: 14.383 +INFO:tensorflow:step = 8101, loss = 0.653333, precision = 0.890625 (6.953 sec) +INFO:tensorflow:global_step/sec: 14.3537 +INFO:tensorflow:step = 8201, loss = 0.715516, precision = 0.851562 (6.967 sec) +Saved checkpoint after 21 epoch(s) to data/resnet56/checkpoints/00021... +INFO:tensorflow:global_step/sec: 12.4592 +INFO:tensorflow:step = 8301, loss = 0.67591, precision = 0.882812 (8.026 sec) +INFO:tensorflow:global_step/sec: 14.3756 +INFO:tensorflow:step = 8401, loss = 0.768176, precision = 0.804688 (6.956 sec) +INFO:tensorflow:global_step/sec: 14.4216 +INFO:tensorflow:step = 8501, loss = 0.633458, precision = 0.859375 (6.934 sec) +INFO:tensorflow:global_step/sec: 14.3775 +INFO:tensorflow:step = 8601, loss = 0.626406, precision = 0.890625 (6.955 sec) +Saved checkpoint after 22 epoch(s) to data/resnet56/checkpoints/00022... +INFO:tensorflow:global_step/sec: 12.4464 +INFO:tensorflow:step = 8701, loss = 0.609048, precision = 0.875 (8.035 sec) +INFO:tensorflow:global_step/sec: 14.4285 +INFO:tensorflow:step = 8801, loss = 0.725175, precision = 0.84375 (6.931 sec) +INFO:tensorflow:global_step/sec: 14.3629 +INFO:tensorflow:step = 8901, loss = 0.759483, precision = 0.835938 (6.962 sec) +Saved checkpoint after 23 epoch(s) to data/resnet56/checkpoints/00023... +INFO:tensorflow:global_step/sec: 12.6669 +INFO:tensorflow:step = 9001, loss = 0.694714, precision = 0.859375 (7.895 sec) +INFO:tensorflow:global_step/sec: 14.4087 +INFO:tensorflow:step = 9101, loss = 0.625168, precision = 0.914062 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.4201 +INFO:tensorflow:step = 9201, loss = 0.619196, precision = 0.859375 (6.935 sec) +INFO:tensorflow:global_step/sec: 14.4542 +INFO:tensorflow:step = 9301, loss = 0.644618, precision = 0.898438 (6.918 sec) +Saved checkpoint after 24 epoch(s) to data/resnet56/checkpoints/00024... +INFO:tensorflow:global_step/sec: 12.5239 +INFO:tensorflow:step = 9401, loss = 0.695165, precision = 0.867188 (7.985 sec) +INFO:tensorflow:global_step/sec: 14.3722 +INFO:tensorflow:step = 9501, loss = 0.604112, precision = 0.921875 (6.958 sec) +INFO:tensorflow:global_step/sec: 14.4189 +INFO:tensorflow:step = 9601, loss = 0.745278, precision = 0.867188 (6.935 sec) +INFO:tensorflow:global_step/sec: 14.3883 +INFO:tensorflow:step = 9701, loss = 0.739409, precision = 0.835938 (6.950 sec) +Saved checkpoint after 25 epoch(s) to data/resnet56/checkpoints/00025... +INFO:tensorflow:global_step/sec: 12.535 +INFO:tensorflow:step = 9801, loss = 0.722668, precision = 0.84375 (7.978 sec) +INFO:tensorflow:global_step/sec: 14.4107 +INFO:tensorflow:step = 9901, loss = 0.652702, precision = 0.882812 (6.939 sec) +INFO:tensorflow:global_step/sec: 14.37 +INFO:tensorflow:step = 10001, loss = 0.713211, precision = 0.8125 (6.959 sec) +INFO:tensorflow:global_step/sec: 14.3797 +INFO:tensorflow:step = 10101, loss = 0.549094, precision = 0.914062 (6.955 sec) +Saved checkpoint after 26 epoch(s) to data/resnet56/checkpoints/00026... +INFO:tensorflow:global_step/sec: 12.4993 +INFO:tensorflow:step = 10201, loss = 0.684437, precision = 0.851562 (8.000 sec) +INFO:tensorflow:global_step/sec: 14.4194 +INFO:tensorflow:step = 10301, loss = 0.625154, precision = 0.882812 (6.935 sec) +INFO:tensorflow:global_step/sec: 14.4113 +INFO:tensorflow:step = 10401, loss = 0.610376, precision = 0.875 (6.939 sec) +INFO:tensorflow:global_step/sec: 14.3967 +INFO:tensorflow:step = 10501, loss = 0.780717, precision = 0.828125 (6.946 sec) +Saved checkpoint after 27 epoch(s) to data/resnet56/checkpoints/00027... +INFO:tensorflow:global_step/sec: 12.5771 +INFO:tensorflow:step = 10601, loss = 0.650081, precision = 0.882812 (7.951 sec) +INFO:tensorflow:global_step/sec: 14.4122 +INFO:tensorflow:step = 10701, loss = 0.676444, precision = 0.890625 (6.939 sec) +INFO:tensorflow:global_step/sec: 14.4174 +INFO:tensorflow:step = 10801, loss = 0.587773, precision = 0.890625 (6.936 sec) +INFO:tensorflow:global_step/sec: 14.4625 +INFO:tensorflow:step = 10901, loss = 0.658733, precision = 0.867188 (6.915 sec) +Saved checkpoint after 28 epoch(s) to data/resnet56/checkpoints/00028... +INFO:tensorflow:global_step/sec: 12.4973 +INFO:tensorflow:step = 11001, loss = 0.534806, precision = 0.898438 (8.002 sec) +INFO:tensorflow:global_step/sec: 14.4404 +INFO:tensorflow:step = 11101, loss = 0.554957, precision = 0.898438 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.3707 +INFO:tensorflow:step = 11201, loss = 0.690801, precision = 0.84375 (6.959 sec) +INFO:tensorflow:global_step/sec: 14.3934 +INFO:tensorflow:step = 11301, loss = 0.782605, precision = 0.835938 (6.948 sec) +Saved checkpoint after 29 epoch(s) to data/resnet56/checkpoints/00029... +INFO:tensorflow:global_step/sec: 12.4961 +INFO:tensorflow:step = 11401, loss = 0.6682, precision = 0.890625 (8.003 sec) +INFO:tensorflow:global_step/sec: 14.3789 +INFO:tensorflow:step = 11501, loss = 0.688307, precision = 0.835938 (6.955 sec) +INFO:tensorflow:global_step/sec: 14.3908 +INFO:tensorflow:step = 11601, loss = 0.618013, precision = 0.914062 (6.949 sec) +INFO:tensorflow:global_step/sec: 14.4031 +INFO:tensorflow:step = 11701, loss = 0.719713, precision = 0.828125 (6.943 sec) +Saved checkpoint after 30 epoch(s) to data/resnet56/checkpoints/00030... +INFO:tensorflow:global_step/sec: 12.487 +INFO:tensorflow:step = 11801, loss = 0.681434, precision = 0.84375 (8.009 sec) +INFO:tensorflow:global_step/sec: 14.4294 +INFO:tensorflow:step = 11901, loss = 0.613333, precision = 0.898438 (6.930 sec) +INFO:tensorflow:global_step/sec: 14.436 +INFO:tensorflow:step = 12001, loss = 0.788281, precision = 0.84375 (6.927 sec) +INFO:tensorflow:global_step/sec: 14.4345 +INFO:tensorflow:step = 12101, loss = 0.568309, precision = 0.890625 (6.927 sec) +Saved checkpoint after 31 epoch(s) to data/resnet56/checkpoints/00031... +INFO:tensorflow:global_step/sec: 12.4026 +INFO:tensorflow:step = 12201, loss = 0.721529, precision = 0.851562 (8.064 sec) +INFO:tensorflow:global_step/sec: 14.4074 +INFO:tensorflow:step = 12301, loss = 0.610239, precision = 0.898438 (6.941 sec) +INFO:tensorflow:global_step/sec: 14.4077 +INFO:tensorflow:step = 12401, loss = 0.684078, precision = 0.859375 (6.941 sec) +INFO:tensorflow:global_step/sec: 14.3861 +INFO:tensorflow:step = 12501, loss = 0.507741, precision = 0.929688 (6.951 sec) +Saved checkpoint after 32 epoch(s) to data/resnet56/checkpoints/00032... +INFO:tensorflow:global_step/sec: 12.5311 +INFO:tensorflow:step = 12601, loss = 0.660966, precision = 0.875 (7.980 sec) +INFO:tensorflow:global_step/sec: 14.403 +INFO:tensorflow:step = 12701, loss = 0.652309, precision = 0.890625 (6.943 sec) +INFO:tensorflow:global_step/sec: 14.3714 +INFO:tensorflow:step = 12801, loss = 0.73201, precision = 0.867188 (6.958 sec) +INFO:tensorflow:global_step/sec: 14.3903 +INFO:tensorflow:step = 12901, loss = 0.645272, precision = 0.921875 (6.949 sec) +Saved checkpoint after 33 epoch(s) to data/resnet56/checkpoints/00033... +INFO:tensorflow:global_step/sec: 12.5553 +INFO:tensorflow:step = 13001, loss = 0.777568, precision = 0.835938 (7.965 sec) +INFO:tensorflow:global_step/sec: 14.4464 +INFO:tensorflow:step = 13101, loss = 0.632445, precision = 0.90625 (6.921 sec) +INFO:tensorflow:global_step/sec: 14.4711 +INFO:tensorflow:step = 13201, loss = 0.699357, precision = 0.890625 (6.911 sec) +Saved checkpoint after 34 epoch(s) to data/resnet56/checkpoints/00034... +INFO:tensorflow:global_step/sec: 12.5714 +INFO:tensorflow:step = 13301, loss = 0.667637, precision = 0.882812 (7.954 sec) +INFO:tensorflow:global_step/sec: 14.4027 +INFO:tensorflow:step = 13401, loss = 0.671324, precision = 0.882812 (6.944 sec) +INFO:tensorflow:global_step/sec: 14.3929 +INFO:tensorflow:step = 13501, loss = 0.691098, precision = 0.875 (6.948 sec) +INFO:tensorflow:global_step/sec: 14.3825 +INFO:tensorflow:step = 13601, loss = 0.809308, precision = 0.804688 (6.953 sec) +Saved checkpoint after 35 epoch(s) to data/resnet56/checkpoints/00035... +INFO:tensorflow:global_step/sec: 12.4583 +INFO:tensorflow:step = 13701, loss = 0.574364, precision = 0.890625 (8.026 sec) +INFO:tensorflow:global_step/sec: 14.4291 +INFO:tensorflow:step = 13801, loss = 0.568596, precision = 0.90625 (6.931 sec) +INFO:tensorflow:global_step/sec: 14.3442 +INFO:tensorflow:step = 13901, loss = 0.766151, precision = 0.820312 (6.972 sec) +INFO:tensorflow:global_step/sec: 14.3565 +INFO:tensorflow:step = 14001, loss = 0.729844, precision = 0.867188 (6.965 sec) +Saved checkpoint after 36 epoch(s) to data/resnet56/checkpoints/00036... +INFO:tensorflow:global_step/sec: 12.5091 +INFO:tensorflow:step = 14101, loss = 0.537923, precision = 0.929688 (7.994 sec) +INFO:tensorflow:global_step/sec: 14.3623 +INFO:tensorflow:step = 14201, loss = 0.658048, precision = 0.890625 (6.963 sec) +INFO:tensorflow:global_step/sec: 14.3882 +INFO:tensorflow:step = 14301, loss = 0.541185, precision = 0.921875 (6.950 sec) +INFO:tensorflow:global_step/sec: 14.3668 +INFO:tensorflow:step = 14401, loss = 0.652842, precision = 0.882812 (6.961 sec) +Saved checkpoint after 37 epoch(s) to data/resnet56/checkpoints/00037... +INFO:tensorflow:global_step/sec: 12.4935 +INFO:tensorflow:step = 14501, loss = 0.750559, precision = 0.835938 (8.004 sec) +INFO:tensorflow:global_step/sec: 14.4284 +INFO:tensorflow:step = 14601, loss = 0.658758, precision = 0.890625 (6.931 sec) +INFO:tensorflow:global_step/sec: 14.4412 +INFO:tensorflow:step = 14701, loss = 0.673662, precision = 0.859375 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.3783 +INFO:tensorflow:step = 14801, loss = 0.636764, precision = 0.90625 (6.955 sec) +Saved checkpoint after 38 epoch(s) to data/resnet56/checkpoints/00038... +INFO:tensorflow:global_step/sec: 12.5462 +INFO:tensorflow:step = 14901, loss = 0.641287, precision = 0.867188 (7.970 sec) +INFO:tensorflow:global_step/sec: 14.4936 +INFO:tensorflow:step = 15001, loss = 0.518863, precision = 0.953125 (6.900 sec) +INFO:tensorflow:global_step/sec: 14.4874 +INFO:tensorflow:step = 15101, loss = 0.865746, precision = 0.8125 (6.902 sec) +INFO:tensorflow:global_step/sec: 14.4557 +INFO:tensorflow:step = 15201, loss = 0.66294, precision = 0.882812 (6.918 sec) +Saved checkpoint after 39 epoch(s) to data/resnet56/checkpoints/00039... +INFO:tensorflow:global_step/sec: 12.5126 +INFO:tensorflow:step = 15301, loss = 0.546064, precision = 0.914062 (7.992 sec) +INFO:tensorflow:global_step/sec: 14.4484 +INFO:tensorflow:step = 15401, loss = 0.677908, precision = 0.882812 (6.921 sec) +INFO:tensorflow:global_step/sec: 14.4127 +INFO:tensorflow:step = 15501, loss = 0.747406, precision = 0.851562 (6.938 sec) +INFO:tensorflow:global_step/sec: 14.3698 +INFO:tensorflow:step = 15601, loss = 0.76404, precision = 0.859375 (6.959 sec) +Saved checkpoint after 40 epoch(s) to data/resnet56/checkpoints/00040... +INFO:tensorflow:global_step/sec: 12.5596 +INFO:tensorflow:step = 15701, loss = 0.684981, precision = 0.890625 (7.962 sec) +INFO:tensorflow:global_step/sec: 14.3945 +INFO:tensorflow:step = 15801, loss = 0.702819, precision = 0.898438 (6.947 sec) +INFO:tensorflow:global_step/sec: 14.4541 +INFO:tensorflow:step = 15901, loss = 0.645774, precision = 0.867188 (6.918 sec) +INFO:tensorflow:global_step/sec: 14.4042 +INFO:tensorflow:step = 16001, loss = 0.534516, precision = 0.929688 (6.943 sec) +Saved checkpoint after 41 epoch(s) to data/resnet56/checkpoints/00041... +INFO:tensorflow:global_step/sec: 12.4962 +INFO:tensorflow:step = 16101, loss = 0.65376, precision = 0.851562 (8.003 sec) +INFO:tensorflow:global_step/sec: 14.4118 +INFO:tensorflow:step = 16201, loss = 0.628911, precision = 0.875 (6.939 sec) +INFO:tensorflow:global_step/sec: 14.4391 +INFO:tensorflow:step = 16301, loss = 0.677035, precision = 0.867188 (6.926 sec) +INFO:tensorflow:global_step/sec: 14.4353 +INFO:tensorflow:step = 16401, loss = 0.634451, precision = 0.914062 (6.927 sec) +Saved checkpoint after 42 epoch(s) to data/resnet56/checkpoints/00042... +INFO:tensorflow:global_step/sec: 12.5075 +INFO:tensorflow:step = 16501, loss = 0.689841, precision = 0.859375 (7.995 sec) +INFO:tensorflow:global_step/sec: 14.4822 +INFO:tensorflow:step = 16601, loss = 0.66198, precision = 0.890625 (6.905 sec) +INFO:tensorflow:global_step/sec: 14.465 +INFO:tensorflow:step = 16701, loss = 0.597779, precision = 0.890625 (6.914 sec) +INFO:tensorflow:global_step/sec: 14.4137 +INFO:tensorflow:step = 16801, loss = 0.599292, precision = 0.898438 (6.938 sec) +Saved checkpoint after 43 epoch(s) to data/resnet56/checkpoints/00043... +INFO:tensorflow:global_step/sec: 12.331 +INFO:tensorflow:step = 16901, loss = 0.657353, precision = 0.867188 (8.110 sec) +INFO:tensorflow:global_step/sec: 14.4921 +INFO:tensorflow:step = 17001, loss = 0.616967, precision = 0.898438 (6.900 sec) +INFO:tensorflow:global_step/sec: 14.4614 +INFO:tensorflow:step = 17101, loss = 0.810541, precision = 0.8125 (6.915 sec) +INFO:tensorflow:global_step/sec: 14.4045 +INFO:tensorflow:step = 17201, loss = 0.666645, precision = 0.898438 (6.942 sec) +Saved checkpoint after 44 epoch(s) to data/resnet56/checkpoints/00044... +INFO:tensorflow:global_step/sec: 12.5641 +INFO:tensorflow:step = 17301, loss = 0.701125, precision = 0.898438 (7.959 sec) +INFO:tensorflow:global_step/sec: 14.4245 +INFO:tensorflow:step = 17401, loss = 0.669334, precision = 0.875 (6.932 sec) +INFO:tensorflow:global_step/sec: 14.3519 +INFO:tensorflow:step = 17501, loss = 0.647525, precision = 0.867188 (6.968 sec) +Saved checkpoint after 45 epoch(s) to data/resnet56/checkpoints/00045... +INFO:tensorflow:global_step/sec: 12.5563 +INFO:tensorflow:step = 17601, loss = 0.652945, precision = 0.867188 (7.964 sec) +INFO:tensorflow:global_step/sec: 14.4223 +INFO:tensorflow:step = 17701, loss = 0.605744, precision = 0.875 (6.936 sec) +INFO:tensorflow:global_step/sec: 14.4183 +INFO:tensorflow:step = 17801, loss = 0.668478, precision = 0.851562 (6.933 sec) +INFO:tensorflow:global_step/sec: 14.3897 +INFO:tensorflow:step = 17901, loss = 0.49466, precision = 0.921875 (6.949 sec) +Saved checkpoint after 46 epoch(s) to data/resnet56/checkpoints/00046... +INFO:tensorflow:global_step/sec: 12.5009 +INFO:tensorflow:step = 18001, loss = 0.657363, precision = 0.828125 (8.000 sec) +INFO:tensorflow:global_step/sec: 14.4167 +INFO:tensorflow:step = 18101, loss = 0.572921, precision = 0.898438 (6.936 sec) +INFO:tensorflow:global_step/sec: 14.4379 +INFO:tensorflow:step = 18201, loss = 0.551847, precision = 0.9375 (6.926 sec) +INFO:tensorflow:global_step/sec: 14.3547 +INFO:tensorflow:step = 18301, loss = 0.671253, precision = 0.84375 (6.966 sec) +Saved checkpoint after 47 epoch(s) to data/resnet56/checkpoints/00047... +INFO:tensorflow:global_step/sec: 12.5445 +INFO:tensorflow:step = 18401, loss = 0.618043, precision = 0.875 (7.972 sec) +INFO:tensorflow:global_step/sec: 14.4973 +INFO:tensorflow:step = 18501, loss = 0.680359, precision = 0.851562 (6.898 sec) +INFO:tensorflow:global_step/sec: 14.443 +INFO:tensorflow:step = 18601, loss = 0.618072, precision = 0.90625 (6.924 sec) +INFO:tensorflow:global_step/sec: 14.4333 +INFO:tensorflow:step = 18701, loss = 0.636822, precision = 0.882812 (6.929 sec) +Saved checkpoint after 48 epoch(s) to data/resnet56/checkpoints/00048... +INFO:tensorflow:global_step/sec: 12.5511 +INFO:tensorflow:step = 18801, loss = 0.660755, precision = 0.851562 (7.967 sec) +INFO:tensorflow:global_step/sec: 14.4494 +INFO:tensorflow:step = 18901, loss = 0.573736, precision = 0.9375 (6.921 sec) +INFO:tensorflow:global_step/sec: 14.457 +INFO:tensorflow:step = 19001, loss = 0.656258, precision = 0.882812 (6.917 sec) +INFO:tensorflow:global_step/sec: 14.425 +INFO:tensorflow:step = 19101, loss = 0.651922, precision = 0.867188 (6.932 sec) +Saved checkpoint after 49 epoch(s) to data/resnet56/checkpoints/00049... +INFO:tensorflow:global_step/sec: 12.5445 +INFO:tensorflow:step = 19201, loss = 0.656263, precision = 0.914062 (7.972 sec) +INFO:tensorflow:global_step/sec: 14.4092 +INFO:tensorflow:step = 19301, loss = 0.760809, precision = 0.851562 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.4524 +INFO:tensorflow:step = 19401, loss = 0.714525, precision = 0.851562 (6.919 sec) +INFO:tensorflow:global_step/sec: 14.4423 +INFO:tensorflow:step = 19501, loss = 0.729754, precision = 0.859375 (6.924 sec) +Saved checkpoint after 50 epoch(s) to data/resnet56/checkpoints/00050... +INFO:tensorflow:global_step/sec: 12.5464 +INFO:tensorflow:step = 19601, loss = 0.727247, precision = 0.875 (7.971 sec) +INFO:tensorflow:global_step/sec: 14.4566 +INFO:tensorflow:step = 19701, loss = 0.606103, precision = 0.882812 (6.917 sec) +INFO:tensorflow:global_step/sec: 14.4102 +INFO:tensorflow:step = 19801, loss = 0.615251, precision = 0.859375 (6.939 sec) +INFO:tensorflow:global_step/sec: 14.388 +INFO:tensorflow:step = 19901, loss = 0.574565, precision = 0.90625 (6.950 sec) +Saved checkpoint after 51 epoch(s) to data/resnet56/checkpoints/00051... +INFO:tensorflow:global_step/sec: 12.5467 +INFO:tensorflow:step = 20001, loss = 0.551826, precision = 0.914062 (7.970 sec) +INFO:tensorflow:global_step/sec: 14.4227 +INFO:tensorflow:step = 20101, loss = 0.708434, precision = 0.851562 (6.933 sec) +INFO:tensorflow:global_step/sec: 14.4089 +INFO:tensorflow:step = 20201, loss = 0.752297, precision = 0.820312 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.4166 +INFO:tensorflow:step = 20301, loss = 0.796295, precision = 0.835938 (6.936 sec) +Saved checkpoint after 52 epoch(s) to data/resnet56/checkpoints/00052... +INFO:tensorflow:global_step/sec: 12.5528 +INFO:tensorflow:step = 20401, loss = 0.53811, precision = 0.921875 (7.967 sec) +INFO:tensorflow:global_step/sec: 14.4583 +INFO:tensorflow:step = 20501, loss = 0.652967, precision = 0.882812 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.4245 +INFO:tensorflow:step = 20601, loss = 0.557454, precision = 0.898438 (6.933 sec) +INFO:tensorflow:global_step/sec: 14.4353 +INFO:tensorflow:step = 20701, loss = 0.711372, precision = 0.859375 (6.927 sec) +Saved checkpoint after 53 epoch(s) to data/resnet56/checkpoints/00053... +INFO:tensorflow:global_step/sec: 12.5172 +INFO:tensorflow:step = 20801, loss = 0.61456, precision = 0.914062 (7.989 sec) +INFO:tensorflow:global_step/sec: 14.3871 +INFO:tensorflow:step = 20901, loss = 0.659566, precision = 0.84375 (6.951 sec) +INFO:tensorflow:global_step/sec: 14.4092 +INFO:tensorflow:step = 21001, loss = 0.686898, precision = 0.882812 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.4524 +INFO:tensorflow:step = 21101, loss = 0.525595, precision = 0.9375 (6.919 sec) +Saved checkpoint after 54 epoch(s) to data/resnet56/checkpoints/00054... +INFO:tensorflow:global_step/sec: 12.5099 +INFO:tensorflow:step = 21201, loss = 0.72701, precision = 0.859375 (7.994 sec) +INFO:tensorflow:global_step/sec: 14.4407 +INFO:tensorflow:step = 21301, loss = 0.630519, precision = 0.898438 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.3711 +INFO:tensorflow:step = 21401, loss = 0.738193, precision = 0.851562 (6.958 sec) +INFO:tensorflow:global_step/sec: 14.4384 +INFO:tensorflow:step = 21501, loss = 0.567369, precision = 0.90625 (6.926 sec) +Saved checkpoint after 55 epoch(s) to data/resnet56/checkpoints/00055... +INFO:tensorflow:global_step/sec: 12.4745 +INFO:tensorflow:step = 21601, loss = 0.627769, precision = 0.867188 (8.016 sec) +INFO:tensorflow:global_step/sec: 14.4301 +INFO:tensorflow:step = 21701, loss = 0.617195, precision = 0.898438 (6.930 sec) +INFO:tensorflow:global_step/sec: 14.4522 +INFO:tensorflow:step = 21801, loss = 0.724316, precision = 0.835938 (6.919 sec) +Saved checkpoint after 56 epoch(s) to data/resnet56/checkpoints/00056... +INFO:tensorflow:global_step/sec: 12.3688 +INFO:tensorflow:step = 21901, loss = 0.53132, precision = 0.945312 (8.084 sec) +INFO:tensorflow:global_step/sec: 14.4178 +INFO:tensorflow:step = 22001, loss = 0.666334, precision = 0.898438 (6.937 sec) +INFO:tensorflow:global_step/sec: 14.4066 +INFO:tensorflow:step = 22101, loss = 0.617862, precision = 0.890625 (6.941 sec) +INFO:tensorflow:global_step/sec: 14.4176 +INFO:tensorflow:step = 22201, loss = 0.681245, precision = 0.898438 (6.936 sec) +Saved checkpoint after 57 epoch(s) to data/resnet56/checkpoints/00057... +INFO:tensorflow:global_step/sec: 12.547 +INFO:tensorflow:step = 22301, loss = 0.682854, precision = 0.867188 (7.970 sec) +INFO:tensorflow:global_step/sec: 14.4057 +INFO:tensorflow:step = 22401, loss = 0.601235, precision = 0.90625 (6.942 sec) +INFO:tensorflow:global_step/sec: 14.4061 +INFO:tensorflow:step = 22501, loss = 0.650148, precision = 0.882812 (6.941 sec) +INFO:tensorflow:global_step/sec: 14.4229 +INFO:tensorflow:step = 22601, loss = 0.591319, precision = 0.890625 (6.933 sec) +Saved checkpoint after 58 epoch(s) to data/resnet56/checkpoints/00058... +INFO:tensorflow:global_step/sec: 12.5462 +INFO:tensorflow:step = 22701, loss = 0.585514, precision = 0.929688 (7.971 sec) +INFO:tensorflow:global_step/sec: 14.3735 +INFO:tensorflow:step = 22801, loss = 0.556175, precision = 0.921875 (6.957 sec) +INFO:tensorflow:global_step/sec: 14.443 +INFO:tensorflow:step = 22901, loss = 0.754173, precision = 0.835938 (6.924 sec) +INFO:tensorflow:global_step/sec: 14.4283 +INFO:tensorflow:step = 23001, loss = 0.697686, precision = 0.867188 (6.931 sec) +Saved checkpoint after 59 epoch(s) to data/resnet56/checkpoints/00059... +INFO:tensorflow:global_step/sec: 12.5302 +INFO:tensorflow:step = 23101, loss = 0.575386, precision = 0.914062 (7.981 sec) +INFO:tensorflow:global_step/sec: 14.4472 +INFO:tensorflow:step = 23201, loss = 0.753549, precision = 0.835938 (6.922 sec) +INFO:tensorflow:global_step/sec: 14.4521 +INFO:tensorflow:step = 23301, loss = 0.618111, precision = 0.890625 (6.920 sec) +INFO:tensorflow:global_step/sec: 14.4681 +INFO:tensorflow:step = 23401, loss = 0.590805, precision = 0.90625 (6.912 sec) +Saved checkpoint after 60 epoch(s) to data/resnet56/checkpoints/00060... +INFO:tensorflow:global_step/sec: 12.5466 +INFO:tensorflow:step = 23501, loss = 0.700034, precision = 0.890625 (7.970 sec) +INFO:tensorflow:global_step/sec: 14.4174 +INFO:tensorflow:step = 23601, loss = 0.679656, precision = 0.859375 (6.936 sec) +INFO:tensorflow:global_step/sec: 14.4222 +INFO:tensorflow:step = 23701, loss = 0.615072, precision = 0.890625 (6.934 sec) +INFO:tensorflow:global_step/sec: 14.4237 +INFO:tensorflow:step = 23801, loss = 0.773443, precision = 0.804688 (6.933 sec) +Saved checkpoint after 61 epoch(s) to data/resnet56/checkpoints/00061... +INFO:tensorflow:global_step/sec: 12.5706 +INFO:tensorflow:step = 23901, loss = 0.63321, precision = 0.867188 (7.955 sec) +INFO:tensorflow:global_step/sec: 14.4998 +INFO:tensorflow:step = 24001, loss = 0.592243, precision = 0.882812 (6.896 sec) +INFO:tensorflow:global_step/sec: 14.4484 +INFO:tensorflow:step = 24101, loss = 0.594596, precision = 0.898438 (6.921 sec) +INFO:tensorflow:global_step/sec: 14.4245 +INFO:tensorflow:step = 24201, loss = 0.751077, precision = 0.828125 (6.933 sec) +Saved checkpoint after 62 epoch(s) to data/resnet56/checkpoints/00062... +INFO:tensorflow:global_step/sec: 12.5764 +INFO:tensorflow:step = 24301, loss = 0.596017, precision = 0.898438 (7.952 sec) +INFO:tensorflow:global_step/sec: 14.4015 +INFO:tensorflow:step = 24401, loss = 0.562603, precision = 0.90625 (6.944 sec) +INFO:tensorflow:global_step/sec: 14.415 +INFO:tensorflow:step = 24501, loss = 0.489734, precision = 0.96875 (6.937 sec) +INFO:tensorflow:global_step/sec: 14.4238 +INFO:tensorflow:step = 24601, loss = 0.587095, precision = 0.90625 (6.933 sec) +Saved checkpoint after 63 epoch(s) to data/resnet56/checkpoints/00063... +INFO:tensorflow:global_step/sec: 12.4466 +INFO:tensorflow:step = 24701, loss = 0.66695, precision = 0.875 (8.034 sec) +INFO:tensorflow:global_step/sec: 14.4097 +INFO:tensorflow:step = 24801, loss = 0.562194, precision = 0.953125 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.4082 +INFO:tensorflow:step = 24901, loss = 0.731398, precision = 0.835938 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.3869 +INFO:tensorflow:step = 25001, loss = 0.527157, precision = 0.9375 (6.951 sec) +Saved checkpoint after 64 epoch(s) to data/resnet56/checkpoints/00064... +INFO:tensorflow:global_step/sec: 12.5592 +INFO:tensorflow:step = 25101, loss = 0.681797, precision = 0.890625 (7.962 sec) +INFO:tensorflow:global_step/sec: 14.4637 +INFO:tensorflow:step = 25201, loss = 0.585106, precision = 0.90625 (6.914 sec) +INFO:tensorflow:global_step/sec: 14.4051 +INFO:tensorflow:step = 25301, loss = 0.724161, precision = 0.84375 (6.942 sec) +INFO:tensorflow:global_step/sec: 14.4399 +INFO:tensorflow:step = 25401, loss = 0.615502, precision = 0.859375 (6.925 sec) +Saved checkpoint after 65 epoch(s) to data/resnet56/checkpoints/00065... +INFO:tensorflow:global_step/sec: 12.6298 +INFO:tensorflow:step = 25501, loss = 0.651613, precision = 0.859375 (7.918 sec) +INFO:tensorflow:global_step/sec: 14.369 +INFO:tensorflow:step = 25601, loss = 0.761597, precision = 0.859375 (6.959 sec) +INFO:tensorflow:global_step/sec: 14.451 +INFO:tensorflow:step = 25701, loss = 0.571824, precision = 0.890625 (6.920 sec) +INFO:tensorflow:global_step/sec: 14.4152 +INFO:tensorflow:step = 25801, loss = 0.559607, precision = 0.914062 (6.937 sec) +Saved checkpoint after 66 epoch(s) to data/resnet56/checkpoints/00066... +INFO:tensorflow:global_step/sec: 12.54 +INFO:tensorflow:step = 25901, loss = 0.537353, precision = 0.921875 (7.975 sec) +INFO:tensorflow:global_step/sec: 14.5015 +INFO:tensorflow:step = 26001, loss = 0.813898, precision = 0.84375 (6.896 sec) +INFO:tensorflow:global_step/sec: 14.506 +INFO:tensorflow:step = 26101, loss = 0.619596, precision = 0.898438 (6.894 sec) +Saved checkpoint after 67 epoch(s) to data/resnet56/checkpoints/00067... +INFO:tensorflow:global_step/sec: 12.5357 +INFO:tensorflow:step = 26201, loss = 0.531715, precision = 0.929688 (7.978 sec) +INFO:tensorflow:global_step/sec: 14.3977 +INFO:tensorflow:step = 26301, loss = 0.583364, precision = 0.9375 (6.945 sec) +INFO:tensorflow:global_step/sec: 14.4249 +INFO:tensorflow:step = 26401, loss = 0.553881, precision = 0.9375 (6.932 sec) +INFO:tensorflow:global_step/sec: 14.4006 +INFO:tensorflow:step = 26501, loss = 0.609982, precision = 0.90625 (6.944 sec) +Saved checkpoint after 68 epoch(s) to data/resnet56/checkpoints/00068... +INFO:tensorflow:global_step/sec: 12.5055 +INFO:tensorflow:step = 26601, loss = 0.562419, precision = 0.9375 (7.997 sec) +INFO:tensorflow:global_step/sec: 14.4092 +INFO:tensorflow:step = 26701, loss = 0.615776, precision = 0.859375 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.3625 +INFO:tensorflow:step = 26801, loss = 0.615582, precision = 0.890625 (6.963 sec) +INFO:tensorflow:global_step/sec: 14.4224 +INFO:tensorflow:step = 26901, loss = 0.519395, precision = 0.945312 (6.934 sec) +Saved checkpoint after 69 epoch(s) to data/resnet56/checkpoints/00069... +INFO:tensorflow:global_step/sec: 12.3668 +INFO:tensorflow:step = 27001, loss = 0.587624, precision = 0.90625 (8.085 sec) +INFO:tensorflow:global_step/sec: 14.4615 +INFO:tensorflow:step = 27101, loss = 0.634616, precision = 0.875 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.3671 +INFO:tensorflow:step = 27201, loss = 0.639735, precision = 0.828125 (6.960 sec) +INFO:tensorflow:global_step/sec: 14.4384 +INFO:tensorflow:step = 27301, loss = 0.58287, precision = 0.921875 (6.926 sec) +Saved checkpoint after 70 epoch(s) to data/resnet56/checkpoints/00070... +INFO:tensorflow:global_step/sec: 12.5417 +INFO:tensorflow:step = 27401, loss = 0.508853, precision = 0.914062 (7.974 sec) +INFO:tensorflow:global_step/sec: 14.4038 +INFO:tensorflow:step = 27501, loss = 0.572815, precision = 0.898438 (6.942 sec) +INFO:tensorflow:global_step/sec: 14.4534 +INFO:tensorflow:step = 27601, loss = 0.569181, precision = 0.929688 (6.919 sec) +INFO:tensorflow:global_step/sec: 14.4004 +INFO:tensorflow:step = 27701, loss = 0.672474, precision = 0.867188 (6.944 sec) +Saved checkpoint after 71 epoch(s) to data/resnet56/checkpoints/00071... +INFO:tensorflow:global_step/sec: 12.5549 +INFO:tensorflow:step = 27801, loss = 0.644935, precision = 0.875 (7.965 sec) +INFO:tensorflow:global_step/sec: 14.3957 +INFO:tensorflow:step = 27901, loss = 0.565872, precision = 0.929688 (6.947 sec) +INFO:tensorflow:global_step/sec: 14.3987 +INFO:tensorflow:step = 28001, loss = 0.603912, precision = 0.859375 (6.945 sec) +INFO:tensorflow:global_step/sec: 14.4549 +INFO:tensorflow:step = 28101, loss = 0.669383, precision = 0.84375 (6.918 sec) +Saved checkpoint after 72 epoch(s) to data/resnet56/checkpoints/00072... +INFO:tensorflow:global_step/sec: 12.5186 +INFO:tensorflow:step = 28201, loss = 0.684222, precision = 0.890625 (7.988 sec) +INFO:tensorflow:global_step/sec: 14.3774 +INFO:tensorflow:step = 28301, loss = 0.691228, precision = 0.875 (6.955 sec) +INFO:tensorflow:global_step/sec: 14.471 +INFO:tensorflow:step = 28401, loss = 0.599505, precision = 0.921875 (6.910 sec) +INFO:tensorflow:global_step/sec: 14.4871 +INFO:tensorflow:step = 28501, loss = 0.659276, precision = 0.84375 (6.902 sec) +Saved checkpoint after 73 epoch(s) to data/resnet56/checkpoints/00073... +INFO:tensorflow:global_step/sec: 12.5384 +INFO:tensorflow:step = 28601, loss = 0.696679, precision = 0.859375 (7.976 sec) +INFO:tensorflow:global_step/sec: 14.3979 +INFO:tensorflow:step = 28701, loss = 0.642113, precision = 0.890625 (6.945 sec) +INFO:tensorflow:global_step/sec: 14.4589 +INFO:tensorflow:step = 28801, loss = 0.592235, precision = 0.921875 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.4736 +INFO:tensorflow:step = 28901, loss = 0.710656, precision = 0.867188 (6.909 sec) +Saved checkpoint after 74 epoch(s) to data/resnet56/checkpoints/00074... +INFO:tensorflow:global_step/sec: 12.4946 +INFO:tensorflow:step = 29001, loss = 0.698518, precision = 0.875 (8.004 sec) +INFO:tensorflow:global_step/sec: 14.4076 +INFO:tensorflow:step = 29101, loss = 0.676127, precision = 0.851562 (6.941 sec) +INFO:tensorflow:global_step/sec: 14.444 +INFO:tensorflow:step = 29201, loss = 0.532997, precision = 0.914062 (6.923 sec) +INFO:tensorflow:global_step/sec: 14.436 +INFO:tensorflow:step = 29301, loss = 0.616027, precision = 0.90625 (6.927 sec) +Saved checkpoint after 75 epoch(s) to data/resnet56/checkpoints/00075... +INFO:tensorflow:global_step/sec: 12.5295 +INFO:tensorflow:step = 29401, loss = 0.552292, precision = 0.921875 (7.981 sec) +INFO:tensorflow:global_step/sec: 14.4196 +INFO:tensorflow:step = 29501, loss = 0.600697, precision = 0.890625 (6.935 sec) +INFO:tensorflow:global_step/sec: 14.4707 +INFO:tensorflow:step = 29601, loss = 0.597691, precision = 0.898438 (6.910 sec) +INFO:tensorflow:global_step/sec: 14.4818 +INFO:tensorflow:step = 29701, loss = 0.628651, precision = 0.875 (6.905 sec) +Saved checkpoint after 76 epoch(s) to data/resnet56/checkpoints/00076... +INFO:tensorflow:global_step/sec: 12.5189 +INFO:tensorflow:step = 29801, loss = 0.665823, precision = 0.867188 (7.988 sec) +INFO:tensorflow:global_step/sec: 14.4189 +INFO:tensorflow:step = 29901, loss = 0.620502, precision = 0.914062 (6.935 sec) +INFO:tensorflow:global_step/sec: 14.4284 +INFO:tensorflow:step = 30001, loss = 0.539279, precision = 0.929688 (6.931 sec) +INFO:tensorflow:global_step/sec: 14.4355 +INFO:tensorflow:step = 30101, loss = 0.645214, precision = 0.867188 (6.927 sec) +Saved checkpoint after 77 epoch(s) to data/resnet56/checkpoints/00077... +INFO:tensorflow:global_step/sec: 12.4454 +INFO:tensorflow:step = 30201, loss = 0.654991, precision = 0.867188 (8.035 sec) +INFO:tensorflow:global_step/sec: 14.4199 +INFO:tensorflow:step = 30301, loss = 0.734104, precision = 0.851562 (6.935 sec) +INFO:tensorflow:global_step/sec: 14.4481 +INFO:tensorflow:step = 30401, loss = 0.663456, precision = 0.875 (6.921 sec) +Saved checkpoint after 78 epoch(s) to data/resnet56/checkpoints/00078... +INFO:tensorflow:global_step/sec: 12.5691 +INFO:tensorflow:step = 30501, loss = 0.515556, precision = 0.929688 (7.956 sec) +INFO:tensorflow:global_step/sec: 14.3525 +INFO:tensorflow:step = 30601, loss = 0.531572, precision = 0.929688 (6.967 sec) +INFO:tensorflow:global_step/sec: 14.433 +INFO:tensorflow:step = 30701, loss = 0.619698, precision = 0.898438 (6.929 sec) +INFO:tensorflow:global_step/sec: 14.4765 +INFO:tensorflow:step = 30801, loss = 0.576839, precision = 0.914062 (6.908 sec) +Saved checkpoint after 79 epoch(s) to data/resnet56/checkpoints/00079... +INFO:tensorflow:global_step/sec: 12.5622 +INFO:tensorflow:step = 30901, loss = 0.485915, precision = 0.953125 (7.960 sec) +INFO:tensorflow:global_step/sec: 14.3361 +INFO:tensorflow:step = 31001, loss = 0.626217, precision = 0.90625 (6.975 sec) +INFO:tensorflow:global_step/sec: 14.4891 +INFO:tensorflow:step = 31101, loss = 0.702168, precision = 0.875 (6.902 sec) +INFO:tensorflow:global_step/sec: 14.4415 +INFO:tensorflow:step = 31201, loss = 0.568474, precision = 0.898438 (6.924 sec) +Saved checkpoint after 80 epoch(s) to data/resnet56/checkpoints/00080... +INFO:tensorflow:global_step/sec: 12.5926 +INFO:tensorflow:step = 31301, loss = 0.615014, precision = 0.851562 (7.941 sec) +INFO:tensorflow:global_step/sec: 14.4757 +INFO:tensorflow:step = 31401, loss = 0.525655, precision = 0.929688 (6.908 sec) +INFO:tensorflow:global_step/sec: 14.4842 +INFO:tensorflow:step = 31501, loss = 0.625303, precision = 0.90625 (6.904 sec) +INFO:tensorflow:global_step/sec: 14.3881 +INFO:tensorflow:step = 31601, loss = 0.581603, precision = 0.882812 (6.950 sec) +Saved checkpoint after 81 epoch(s) to data/resnet56/checkpoints/00081... +INFO:tensorflow:global_step/sec: 12.4915 +INFO:tensorflow:step = 31701, loss = 0.637754, precision = 0.90625 (8.006 sec) +INFO:tensorflow:global_step/sec: 14.4359 +INFO:tensorflow:step = 31801, loss = 0.578071, precision = 0.921875 (6.927 sec) +INFO:tensorflow:global_step/sec: 14.5054 +INFO:tensorflow:step = 31901, loss = 0.571424, precision = 0.898438 (6.894 sec) +INFO:tensorflow:global_step/sec: 14.4444 +INFO:tensorflow:step = 32001, loss = 0.62745, precision = 0.890625 (6.923 sec) +Saved checkpoint after 82 epoch(s) to data/resnet56/checkpoints/00082... +INFO:tensorflow:global_step/sec: 12.1121 +INFO:tensorflow:step = 32101, loss = 0.565429, precision = 0.914062 (8.256 sec) +INFO:tensorflow:global_step/sec: 14.4233 +INFO:tensorflow:step = 32201, loss = 0.682407, precision = 0.875 (6.934 sec) +INFO:tensorflow:global_step/sec: 14.4569 +INFO:tensorflow:step = 32301, loss = 0.58467, precision = 0.898438 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.4982 +INFO:tensorflow:step = 32401, loss = 0.654294, precision = 0.875 (6.899 sec) +Saved checkpoint after 83 epoch(s) to data/resnet56/checkpoints/00083... +INFO:tensorflow:global_step/sec: 12.411 +INFO:tensorflow:step = 32501, loss = 0.506089, precision = 0.9375 (8.056 sec) +INFO:tensorflow:global_step/sec: 14.4616 +INFO:tensorflow:step = 32601, loss = 0.683348, precision = 0.875 (6.915 sec) +INFO:tensorflow:global_step/sec: 14.4498 +INFO:tensorflow:step = 32701, loss = 0.631142, precision = 0.882812 (6.920 sec) +INFO:tensorflow:global_step/sec: 14.4058 +INFO:tensorflow:step = 32801, loss = 0.618421, precision = 0.898438 (6.942 sec) +Saved checkpoint after 84 epoch(s) to data/resnet56/checkpoints/00084... +INFO:tensorflow:global_step/sec: 12.4403 +INFO:tensorflow:step = 32901, loss = 0.815444, precision = 0.835938 (8.038 sec) +INFO:tensorflow:global_step/sec: 14.4877 +INFO:tensorflow:step = 33001, loss = 0.579123, precision = 0.914062 (6.902 sec) +INFO:tensorflow:global_step/sec: 14.4592 +INFO:tensorflow:step = 33101, loss = 0.675279, precision = 0.867188 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.4902 +INFO:tensorflow:step = 33201, loss = 0.573342, precision = 0.890625 (6.901 sec) +Saved checkpoint after 85 epoch(s) to data/resnet56/checkpoints/00085... +INFO:tensorflow:global_step/sec: 12.3564 +INFO:tensorflow:step = 33301, loss = 0.587982, precision = 0.90625 (8.093 sec) +INFO:tensorflow:global_step/sec: 14.4564 +INFO:tensorflow:step = 33401, loss = 0.52836, precision = 0.921875 (6.917 sec) +INFO:tensorflow:global_step/sec: 14.4909 +INFO:tensorflow:step = 33501, loss = 0.561633, precision = 0.914062 (6.901 sec) +INFO:tensorflow:global_step/sec: 14.4796 +INFO:tensorflow:step = 33601, loss = 0.612584, precision = 0.90625 (6.906 sec) +Saved checkpoint after 86 epoch(s) to data/resnet56/checkpoints/00086... +INFO:tensorflow:global_step/sec: 12.5008 +INFO:tensorflow:step = 33701, loss = 0.668936, precision = 0.859375 (8.000 sec) +INFO:tensorflow:global_step/sec: 14.4096 +INFO:tensorflow:step = 33801, loss = 0.720488, precision = 0.867188 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.4633 +INFO:tensorflow:step = 33901, loss = 0.551656, precision = 0.890625 (6.914 sec) +INFO:tensorflow:global_step/sec: 14.4601 +INFO:tensorflow:step = 34001, loss = 0.56556, precision = 0.921875 (6.915 sec) +Saved checkpoint after 87 epoch(s) to data/resnet56/checkpoints/00087... +INFO:tensorflow:global_step/sec: 12.6031 +INFO:tensorflow:step = 34101, loss = 0.596889, precision = 0.90625 (7.935 sec) +INFO:tensorflow:global_step/sec: 14.4614 +INFO:tensorflow:step = 34201, loss = 0.558008, precision = 0.914062 (6.915 sec) +INFO:tensorflow:global_step/sec: 14.4124 +INFO:tensorflow:step = 34301, loss = 0.504818, precision = 0.929688 (6.939 sec) +INFO:tensorflow:global_step/sec: 14.4055 +INFO:tensorflow:step = 34401, loss = 0.622318, precision = 0.898438 (6.942 sec) +Saved checkpoint after 88 epoch(s) to data/resnet56/checkpoints/00088... +INFO:tensorflow:global_step/sec: 12.4901 +INFO:tensorflow:step = 34501, loss = 0.711159, precision = 0.867188 (8.006 sec) +INFO:tensorflow:global_step/sec: 14.4971 +INFO:tensorflow:step = 34601, loss = 0.682046, precision = 0.867188 (6.898 sec) +INFO:tensorflow:global_step/sec: 14.44 +INFO:tensorflow:step = 34701, loss = 0.696709, precision = 0.851562 (6.925 sec) +Saved checkpoint after 89 epoch(s) to data/resnet56/checkpoints/00089... +INFO:tensorflow:global_step/sec: 12.5107 +INFO:tensorflow:step = 34801, loss = 0.527577, precision = 0.9375 (7.993 sec) +INFO:tensorflow:global_step/sec: 14.3893 +INFO:tensorflow:step = 34901, loss = 0.750956, precision = 0.84375 (6.950 sec) +INFO:tensorflow:global_step/sec: 14.4299 +INFO:tensorflow:step = 35001, loss = 0.600741, precision = 0.890625 (6.930 sec) +INFO:tensorflow:global_step/sec: 14.4217 +INFO:tensorflow:step = 35101, loss = 0.640542, precision = 0.898438 (6.934 sec) +Saved checkpoint after 90 epoch(s) to data/resnet56/checkpoints/00090... +INFO:tensorflow:global_step/sec: 12.4996 +INFO:tensorflow:step = 35201, loss = 0.679764, precision = 0.90625 (8.000 sec) +INFO:tensorflow:global_step/sec: 14.4362 +INFO:tensorflow:step = 35301, loss = 0.636654, precision = 0.90625 (6.927 sec) +INFO:tensorflow:global_step/sec: 14.4622 +INFO:tensorflow:step = 35401, loss = 0.51344, precision = 0.9375 (6.915 sec) +INFO:tensorflow:global_step/sec: 14.4518 +INFO:tensorflow:step = 35501, loss = 0.704695, precision = 0.835938 (6.920 sec) +Saved checkpoint after 91 epoch(s) to data/resnet56/checkpoints/00091... +INFO:tensorflow:global_step/sec: 12.4958 +INFO:tensorflow:step = 35601, loss = 0.47875, precision = 0.9375 (8.002 sec) +INFO:tensorflow:global_step/sec: 14.5434 +INFO:tensorflow:step = 35701, loss = 0.496748, precision = 0.945312 (6.875 sec) +INFO:tensorflow:global_step/sec: 14.4187 +INFO:tensorflow:step = 35801, loss = 0.469646, precision = 0.9375 (6.936 sec) +INFO:tensorflow:global_step/sec: 14.4366 +INFO:tensorflow:step = 35901, loss = 0.504615, precision = 0.921875 (6.927 sec) +Saved checkpoint after 92 epoch(s) to data/resnet56/checkpoints/00092... +INFO:tensorflow:global_step/sec: 12.419 +INFO:tensorflow:step = 36001, loss = 0.403491, precision = 0.953125 (8.052 sec) +INFO:tensorflow:global_step/sec: 14.4209 +INFO:tensorflow:step = 36101, loss = 0.429825, precision = 0.953125 (6.934 sec) +INFO:tensorflow:global_step/sec: 14.4696 +INFO:tensorflow:step = 36201, loss = 0.390842, precision = 0.960938 (6.911 sec) +INFO:tensorflow:global_step/sec: 14.4449 +INFO:tensorflow:step = 36301, loss = 0.406678, precision = 0.9375 (6.923 sec) +Saved checkpoint after 93 epoch(s) to data/resnet56/checkpoints/00093... +INFO:tensorflow:global_step/sec: 12.4361 +INFO:tensorflow:step = 36401, loss = 0.484292, precision = 0.945312 (8.041 sec) +INFO:tensorflow:global_step/sec: 14.4188 +INFO:tensorflow:step = 36501, loss = 0.452133, precision = 0.929688 (6.935 sec) +INFO:tensorflow:global_step/sec: 14.4583 +INFO:tensorflow:step = 36601, loss = 0.460896, precision = 0.960938 (6.917 sec) +INFO:tensorflow:global_step/sec: 14.4655 +INFO:tensorflow:step = 36701, loss = 0.387779, precision = 0.960938 (6.913 sec) +Saved checkpoint after 94 epoch(s) to data/resnet56/checkpoints/00094... +INFO:tensorflow:global_step/sec: 12.4431 +INFO:tensorflow:step = 36801, loss = 0.469654, precision = 0.9375 (8.037 sec) +INFO:tensorflow:global_step/sec: 14.4567 +INFO:tensorflow:step = 36901, loss = 0.39316, precision = 0.960938 (6.917 sec) +INFO:tensorflow:global_step/sec: 14.4467 +INFO:tensorflow:step = 37001, loss = 0.427562, precision = 0.945312 (6.922 sec) +INFO:tensorflow:global_step/sec: 14.4289 +INFO:tensorflow:step = 37101, loss = 0.394956, precision = 0.953125 (6.931 sec) +Saved checkpoint after 95 epoch(s) to data/resnet56/checkpoints/00095... +INFO:tensorflow:global_step/sec: 12.4481 +INFO:tensorflow:step = 37201, loss = 0.351889, precision = 0.960938 (8.037 sec) +INFO:tensorflow:global_step/sec: 14.4195 +INFO:tensorflow:step = 37301, loss = 0.331704, precision = 0.984375 (6.932 sec) +INFO:tensorflow:global_step/sec: 14.4323 +INFO:tensorflow:step = 37401, loss = 0.427608, precision = 0.9375 (6.929 sec) +INFO:tensorflow:global_step/sec: 14.4446 +INFO:tensorflow:step = 37501, loss = 0.408804, precision = 0.953125 (6.923 sec) +Saved checkpoint after 96 epoch(s) to data/resnet56/checkpoints/00096... +INFO:tensorflow:global_step/sec: 12.492 +INFO:tensorflow:step = 37601, loss = 0.321221, precision = 0.976562 (8.005 sec) +INFO:tensorflow:global_step/sec: 14.4618 +INFO:tensorflow:step = 37701, loss = 0.332197, precision = 0.96875 (6.914 sec) +INFO:tensorflow:global_step/sec: 14.4221 +INFO:tensorflow:step = 37801, loss = 0.286236, precision = 1.0 (6.934 sec) +INFO:tensorflow:global_step/sec: 14.4412 +INFO:tensorflow:step = 37901, loss = 0.356354, precision = 0.960938 (6.925 sec) +Saved checkpoint after 97 epoch(s) to data/resnet56/checkpoints/00097... +INFO:tensorflow:global_step/sec: 12.4186 +INFO:tensorflow:step = 38001, loss = 0.365232, precision = 0.945312 (8.053 sec) +INFO:tensorflow:global_step/sec: 14.4502 +INFO:tensorflow:step = 38101, loss = 0.324683, precision = 0.984375 (6.920 sec) +INFO:tensorflow:global_step/sec: 14.3727 +INFO:tensorflow:step = 38201, loss = 0.335034, precision = 0.976562 (6.959 sec) +INFO:tensorflow:global_step/sec: 14.3568 +INFO:tensorflow:step = 38301, loss = 0.293278, precision = 0.992188 (6.964 sec) +Saved checkpoint after 98 epoch(s) to data/resnet56/checkpoints/00098... +INFO:tensorflow:global_step/sec: 12.3868 +INFO:tensorflow:step = 38401, loss = 0.350353, precision = 0.96875 (8.073 sec) +INFO:tensorflow:global_step/sec: 14.399 +INFO:tensorflow:step = 38501, loss = 0.3327, precision = 0.960938 (6.945 sec) +INFO:tensorflow:global_step/sec: 14.4201 +INFO:tensorflow:step = 38601, loss = 0.343223, precision = 0.960938 (6.935 sec) +INFO:tensorflow:global_step/sec: 14.4375 +INFO:tensorflow:step = 38701, loss = 0.308454, precision = 0.976562 (6.926 sec) +Saved checkpoint after 99 epoch(s) to data/resnet56/checkpoints/00099... +INFO:tensorflow:global_step/sec: 12.3732 +INFO:tensorflow:step = 38801, loss = 0.281165, precision = 0.984375 (8.082 sec) +INFO:tensorflow:global_step/sec: 14.422 +INFO:tensorflow:step = 38901, loss = 0.310168, precision = 0.984375 (6.934 sec) +INFO:tensorflow:global_step/sec: 14.4243 +INFO:tensorflow:step = 39001, loss = 0.306406, precision = 0.992188 (6.933 sec) +Saved checkpoint after 100 epoch(s) to data/resnet56/checkpoints/00100... +INFO:tensorflow:global_step/sec: 12.3274 +INFO:tensorflow:step = 39101, loss = 0.301743, precision = 0.976562 (8.113 sec) +INFO:tensorflow:global_step/sec: 14.4082 +INFO:tensorflow:step = 39201, loss = 0.286185, precision = 0.96875 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.419 +INFO:tensorflow:step = 39301, loss = 0.395804, precision = 0.9375 (6.936 sec) +INFO:tensorflow:global_step/sec: 14.3521 +INFO:tensorflow:step = 39401, loss = 0.323202, precision = 0.96875 (6.968 sec) +Saved checkpoint after 101 epoch(s) to data/resnet56/checkpoints/00101... +INFO:tensorflow:global_step/sec: 12.3604 +INFO:tensorflow:step = 39501, loss = 0.298583, precision = 0.984375 (8.090 sec) +INFO:tensorflow:global_step/sec: 14.4353 +INFO:tensorflow:step = 39601, loss = 0.274957, precision = 1.0 (6.928 sec) +INFO:tensorflow:global_step/sec: 14.4314 +INFO:tensorflow:step = 39701, loss = 0.300244, precision = 0.960938 (6.929 sec) +INFO:tensorflow:global_step/sec: 14.3173 +INFO:tensorflow:step = 39801, loss = 0.252912, precision = 0.992188 (6.985 sec) +Saved checkpoint after 102 epoch(s) to data/resnet56/checkpoints/00102... +INFO:tensorflow:global_step/sec: 12.373 +INFO:tensorflow:step = 39901, loss = 0.278584, precision = 0.992188 (8.082 sec) +INFO:tensorflow:global_step/sec: 14.4438 +INFO:tensorflow:step = 40001, loss = 0.280127, precision = 0.984375 (6.923 sec) +INFO:tensorflow:global_step/sec: 14.4309 +INFO:tensorflow:step = 40101, loss = 0.272486, precision = 0.976562 (6.930 sec) +INFO:tensorflow:global_step/sec: 14.4451 +INFO:tensorflow:step = 40201, loss = 0.265197, precision = 0.960938 (6.923 sec) +Saved checkpoint after 103 epoch(s) to data/resnet56/checkpoints/00103... +INFO:tensorflow:global_step/sec: 12.4276 +INFO:tensorflow:step = 40301, loss = 0.325843, precision = 0.953125 (8.047 sec) +INFO:tensorflow:global_step/sec: 14.455 +INFO:tensorflow:step = 40401, loss = 0.270157, precision = 0.984375 (6.918 sec) +INFO:tensorflow:global_step/sec: 14.4647 +INFO:tensorflow:step = 40501, loss = 0.270566, precision = 0.976562 (6.913 sec) +INFO:tensorflow:global_step/sec: 14.4598 +INFO:tensorflow:step = 40601, loss = 0.335207, precision = 0.976562 (6.916 sec) +Saved checkpoint after 104 epoch(s) to data/resnet56/checkpoints/00104... +INFO:tensorflow:global_step/sec: 12.4584 +INFO:tensorflow:step = 40701, loss = 0.234159, precision = 1.0 (8.026 sec) +INFO:tensorflow:global_step/sec: 14.4738 +INFO:tensorflow:step = 40801, loss = 0.314318, precision = 0.960938 (6.909 sec) +INFO:tensorflow:global_step/sec: 14.4458 +INFO:tensorflow:step = 40901, loss = 0.297349, precision = 0.960938 (6.922 sec) +INFO:tensorflow:global_step/sec: 14.458 +INFO:tensorflow:step = 41001, loss = 0.325257, precision = 0.96875 (6.917 sec) +Saved checkpoint after 105 epoch(s) to data/resnet56/checkpoints/00105... +INFO:tensorflow:global_step/sec: 12.3129 +INFO:tensorflow:step = 41101, loss = 0.299471, precision = 0.953125 (8.122 sec) +INFO:tensorflow:global_step/sec: 14.4377 +INFO:tensorflow:step = 41201, loss = 0.266066, precision = 0.984375 (6.926 sec) +INFO:tensorflow:global_step/sec: 14.5028 +INFO:tensorflow:step = 41301, loss = 0.253765, precision = 0.976562 (6.895 sec) +INFO:tensorflow:global_step/sec: 14.4572 +INFO:tensorflow:step = 41401, loss = 0.258323, precision = 0.976562 (6.917 sec) +Saved checkpoint after 106 epoch(s) to data/resnet56/checkpoints/00106... +INFO:tensorflow:global_step/sec: 12.3712 +INFO:tensorflow:step = 41501, loss = 0.253028, precision = 0.984375 (8.083 sec) +INFO:tensorflow:global_step/sec: 14.4716 +INFO:tensorflow:step = 41601, loss = 0.223773, precision = 0.992188 (6.910 sec) +INFO:tensorflow:global_step/sec: 14.3976 +INFO:tensorflow:step = 41701, loss = 0.276552, precision = 0.960938 (6.946 sec) +INFO:tensorflow:global_step/sec: 14.4811 +INFO:tensorflow:step = 41801, loss = 0.273822, precision = 0.960938 (6.905 sec) +Saved checkpoint after 107 epoch(s) to data/resnet56/checkpoints/00107... +INFO:tensorflow:global_step/sec: 12.5678 +INFO:tensorflow:step = 41901, loss = 0.248595, precision = 0.96875 (7.957 sec) +INFO:tensorflow:global_step/sec: 14.4469 +INFO:tensorflow:step = 42001, loss = 0.274476, precision = 0.976562 (6.922 sec) +INFO:tensorflow:global_step/sec: 14.4573 +INFO:tensorflow:step = 42101, loss = 0.2876, precision = 0.953125 (6.917 sec) +INFO:tensorflow:global_step/sec: 14.4159 +INFO:tensorflow:step = 42201, loss = 0.253329, precision = 0.984375 (6.937 sec) +Saved checkpoint after 108 epoch(s) to data/resnet56/checkpoints/00108... +INFO:tensorflow:global_step/sec: 12.3411 +INFO:tensorflow:step = 42301, loss = 0.262976, precision = 0.976562 (8.103 sec) +INFO:tensorflow:global_step/sec: 14.4546 +INFO:tensorflow:step = 42401, loss = 0.263587, precision = 0.976562 (6.918 sec) +INFO:tensorflow:global_step/sec: 14.4163 +INFO:tensorflow:step = 42501, loss = 0.224054, precision = 0.984375 (6.937 sec) +INFO:tensorflow:global_step/sec: 14.4503 +INFO:tensorflow:step = 42601, loss = 0.260475, precision = 0.976562 (6.920 sec) +Saved checkpoint after 109 epoch(s) to data/resnet56/checkpoints/00109... +INFO:tensorflow:global_step/sec: 12.434 +INFO:tensorflow:step = 42701, loss = 0.251282, precision = 0.953125 (8.042 sec) +INFO:tensorflow:global_step/sec: 14.3633 +INFO:tensorflow:step = 42801, loss = 0.233391, precision = 0.976562 (6.962 sec) +INFO:tensorflow:global_step/sec: 14.4401 +INFO:tensorflow:step = 42901, loss = 0.212685, precision = 1.0 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.4513 +INFO:tensorflow:step = 43001, loss = 0.196022, precision = 1.0 (6.920 sec) +Saved checkpoint after 110 epoch(s) to data/resnet56/checkpoints/00110... +INFO:tensorflow:global_step/sec: 12.5598 +INFO:tensorflow:step = 43101, loss = 0.247769, precision = 0.96875 (7.962 sec) +INFO:tensorflow:global_step/sec: 14.4297 +INFO:tensorflow:step = 43201, loss = 0.205891, precision = 0.992188 (6.930 sec) +INFO:tensorflow:global_step/sec: 14.4421 +INFO:tensorflow:step = 43301, loss = 0.25631, precision = 0.960938 (6.924 sec) +Saved checkpoint after 111 epoch(s) to data/resnet56/checkpoints/00111... +INFO:tensorflow:global_step/sec: 12.5159 +INFO:tensorflow:step = 43401, loss = 0.223733, precision = 0.992188 (7.989 sec) +INFO:tensorflow:global_step/sec: 14.3679 +INFO:tensorflow:step = 43501, loss = 0.214298, precision = 0.992188 (6.961 sec) +INFO:tensorflow:global_step/sec: 14.4539 +INFO:tensorflow:step = 43601, loss = 0.215812, precision = 0.984375 (6.918 sec) +INFO:tensorflow:global_step/sec: 14.4678 +INFO:tensorflow:step = 43701, loss = 0.248235, precision = 0.976562 (6.912 sec) +Saved checkpoint after 112 epoch(s) to data/resnet56/checkpoints/00112... +INFO:tensorflow:global_step/sec: 12.4726 +INFO:tensorflow:step = 43801, loss = 0.224179, precision = 0.984375 (8.017 sec) +INFO:tensorflow:global_step/sec: 14.3504 +INFO:tensorflow:step = 43901, loss = 0.186936, precision = 1.0 (6.968 sec) +INFO:tensorflow:global_step/sec: 14.4558 +INFO:tensorflow:step = 44001, loss = 0.268706, precision = 0.960938 (6.918 sec) +INFO:tensorflow:global_step/sec: 14.4499 +INFO:tensorflow:step = 44101, loss = 0.245352, precision = 0.960938 (6.921 sec) +Saved checkpoint after 113 epoch(s) to data/resnet56/checkpoints/00113... +INFO:tensorflow:global_step/sec: 12.4732 +INFO:tensorflow:step = 44201, loss = 0.196721, precision = 0.992188 (8.017 sec) +INFO:tensorflow:global_step/sec: 14.4106 +INFO:tensorflow:step = 44301, loss = 0.271562, precision = 0.953125 (6.939 sec) +INFO:tensorflow:global_step/sec: 14.408 +INFO:tensorflow:step = 44401, loss = 0.280848, precision = 0.96875 (6.941 sec) +INFO:tensorflow:global_step/sec: 14.4586 +INFO:tensorflow:step = 44501, loss = 0.187177, precision = 1.0 (6.916 sec) +Saved checkpoint after 114 epoch(s) to data/resnet56/checkpoints/00114... +INFO:tensorflow:global_step/sec: 12.5682 +INFO:tensorflow:step = 44601, loss = 0.25798, precision = 0.976562 (7.957 sec) +INFO:tensorflow:global_step/sec: 14.3943 +INFO:tensorflow:step = 44701, loss = 0.194703, precision = 0.992188 (6.947 sec) +INFO:tensorflow:global_step/sec: 14.4785 +INFO:tensorflow:step = 44801, loss = 0.204789, precision = 0.992188 (6.907 sec) +INFO:tensorflow:global_step/sec: 14.4532 +INFO:tensorflow:step = 44901, loss = 0.268898, precision = 0.976562 (6.919 sec) +Saved checkpoint after 115 epoch(s) to data/resnet56/checkpoints/00115... +INFO:tensorflow:global_step/sec: 12.5205 +INFO:tensorflow:step = 45001, loss = 0.195035, precision = 1.0 (7.987 sec) +INFO:tensorflow:global_step/sec: 14.4745 +INFO:tensorflow:step = 45101, loss = 0.210209, precision = 0.984375 (6.908 sec) +INFO:tensorflow:global_step/sec: 14.4551 +INFO:tensorflow:step = 45201, loss = 0.248911, precision = 0.960938 (6.919 sec) +INFO:tensorflow:global_step/sec: 14.4751 +INFO:tensorflow:step = 45301, loss = 0.220673, precision = 0.984375 (6.909 sec) +Saved checkpoint after 116 epoch(s) to data/resnet56/checkpoints/00116... +INFO:tensorflow:global_step/sec: 12.4791 +INFO:tensorflow:step = 45401, loss = 0.197641, precision = 0.984375 (8.013 sec) +INFO:tensorflow:global_step/sec: 14.439 +INFO:tensorflow:step = 45501, loss = 0.282687, precision = 0.945312 (6.926 sec) +INFO:tensorflow:global_step/sec: 14.4076 +INFO:tensorflow:step = 45601, loss = 0.220919, precision = 0.976562 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.4683 +INFO:tensorflow:step = 45701, loss = 0.201237, precision = 0.984375 (6.912 sec) +Saved checkpoint after 117 epoch(s) to data/resnet56/checkpoints/00117... +INFO:tensorflow:global_step/sec: 12.4352 +INFO:tensorflow:step = 45801, loss = 0.215851, precision = 0.984375 (8.042 sec) +INFO:tensorflow:global_step/sec: 14.4294 +INFO:tensorflow:step = 45901, loss = 0.198106, precision = 0.984375 (6.938 sec) +INFO:tensorflow:global_step/sec: 14.4166 +INFO:tensorflow:step = 46001, loss = 0.218494, precision = 0.96875 (6.928 sec) +INFO:tensorflow:global_step/sec: 14.3593 +INFO:tensorflow:step = 46101, loss = 0.230651, precision = 0.960938 (6.965 sec) +Saved checkpoint after 118 epoch(s) to data/resnet56/checkpoints/00118... +INFO:tensorflow:global_step/sec: 12.6551 +INFO:tensorflow:step = 46201, loss = 0.237919, precision = 0.984375 (7.901 sec) +INFO:tensorflow:global_step/sec: 14.4075 +INFO:tensorflow:step = 46301, loss = 0.283037, precision = 0.953125 (6.941 sec) +INFO:tensorflow:global_step/sec: 14.4058 +INFO:tensorflow:step = 46401, loss = 0.205464, precision = 0.992188 (6.942 sec) +INFO:tensorflow:global_step/sec: 14.4958 +INFO:tensorflow:step = 46501, loss = 0.24523, precision = 0.96875 (6.898 sec) +Saved checkpoint after 119 epoch(s) to data/resnet56/checkpoints/00119... +INFO:tensorflow:global_step/sec: 12.4673 +INFO:tensorflow:step = 46601, loss = 0.180776, precision = 0.992188 (8.021 sec) +INFO:tensorflow:global_step/sec: 14.4384 +INFO:tensorflow:step = 46701, loss = 0.219811, precision = 0.976562 (6.926 sec) +INFO:tensorflow:global_step/sec: 14.4319 +INFO:tensorflow:step = 46801, loss = 0.213762, precision = 0.96875 (6.929 sec) +INFO:tensorflow:global_step/sec: 14.4648 +INFO:tensorflow:step = 46901, loss = 0.243675, precision = 0.96875 (6.913 sec) +Saved checkpoint after 120 epoch(s) to data/resnet56/checkpoints/00120... +INFO:tensorflow:global_step/sec: 12.4792 +INFO:tensorflow:step = 47001, loss = 0.260702, precision = 0.96875 (8.013 sec) +INFO:tensorflow:global_step/sec: 14.4025 +INFO:tensorflow:step = 47101, loss = 0.181016, precision = 1.0 (6.943 sec) +INFO:tensorflow:global_step/sec: 14.4193 +INFO:tensorflow:step = 47201, loss = 0.187007, precision = 1.0 (6.935 sec) +INFO:tensorflow:global_step/sec: 14.4368 +INFO:tensorflow:step = 47301, loss = 0.299468, precision = 0.960938 (6.927 sec) +Saved checkpoint after 121 epoch(s) to data/resnet56/checkpoints/00121... +INFO:tensorflow:global_step/sec: 12.3512 +INFO:tensorflow:step = 47401, loss = 0.181331, precision = 0.984375 (8.097 sec) +INFO:tensorflow:global_step/sec: 14.418 +INFO:tensorflow:step = 47501, loss = 0.203589, precision = 0.984375 (6.936 sec) +INFO:tensorflow:global_step/sec: 14.4704 +INFO:tensorflow:step = 47601, loss = 0.196601, precision = 0.984375 (6.911 sec) +INFO:tensorflow:global_step/sec: 14.4459 +INFO:tensorflow:step = 47701, loss = 0.208082, precision = 0.976562 (6.922 sec) +Saved checkpoint after 122 epoch(s) to data/resnet56/checkpoints/00122... +INFO:tensorflow:global_step/sec: 12.6451 +INFO:tensorflow:step = 47801, loss = 0.205955, precision = 0.984375 (7.908 sec) +INFO:tensorflow:global_step/sec: 14.5515 +INFO:tensorflow:step = 47901, loss = 0.190683, precision = 0.976562 (6.872 sec) +INFO:tensorflow:global_step/sec: 14.4207 +INFO:tensorflow:step = 48001, loss = 0.189691, precision = 0.976562 (6.935 sec) +Saved checkpoint after 123 epoch(s) to data/resnet56/checkpoints/00123... +INFO:tensorflow:global_step/sec: 12.547 +INFO:tensorflow:step = 48101, loss = 0.243001, precision = 0.976562 (7.970 sec) +INFO:tensorflow:global_step/sec: 14.376 +INFO:tensorflow:step = 48201, loss = 0.235314, precision = 0.976562 (6.956 sec) +INFO:tensorflow:global_step/sec: 14.4263 +INFO:tensorflow:step = 48301, loss = 0.19673, precision = 0.984375 (6.931 sec) +INFO:tensorflow:global_step/sec: 14.4738 +INFO:tensorflow:step = 48401, loss = 0.209593, precision = 0.976562 (6.909 sec) +Saved checkpoint after 124 epoch(s) to data/resnet56/checkpoints/00124... +INFO:tensorflow:global_step/sec: 12.5362 +INFO:tensorflow:step = 48501, loss = 0.200532, precision = 0.976562 (7.977 sec) +INFO:tensorflow:global_step/sec: 14.3832 +INFO:tensorflow:step = 48601, loss = 0.237809, precision = 0.96875 (6.953 sec) +INFO:tensorflow:global_step/sec: 14.4322 +INFO:tensorflow:step = 48701, loss = 0.166781, precision = 1.0 (6.929 sec) +INFO:tensorflow:global_step/sec: 14.4609 +INFO:tensorflow:step = 48801, loss = 0.203261, precision = 0.96875 (6.915 sec) +Saved checkpoint after 125 epoch(s) to data/resnet56/checkpoints/00125... +INFO:tensorflow:global_step/sec: 12.5079 +INFO:tensorflow:step = 48901, loss = 0.206463, precision = 0.984375 (7.995 sec) +INFO:tensorflow:global_step/sec: 14.45 +INFO:tensorflow:step = 49001, loss = 0.217519, precision = 0.96875 (6.921 sec) +INFO:tensorflow:global_step/sec: 14.4318 +INFO:tensorflow:step = 49101, loss = 0.245479, precision = 0.960938 (6.929 sec) +INFO:tensorflow:global_step/sec: 14.4274 +INFO:tensorflow:step = 49201, loss = 0.210676, precision = 0.96875 (6.931 sec) +Saved checkpoint after 126 epoch(s) to data/resnet56/checkpoints/00126... +INFO:tensorflow:global_step/sec: 12.5744 +INFO:tensorflow:step = 49301, loss = 0.174472, precision = 0.984375 (7.953 sec) +INFO:tensorflow:global_step/sec: 14.4519 +INFO:tensorflow:step = 49401, loss = 0.172653, precision = 0.992188 (6.919 sec) +INFO:tensorflow:global_step/sec: 14.4332 +INFO:tensorflow:step = 49501, loss = 0.176667, precision = 0.992188 (6.928 sec) +INFO:tensorflow:global_step/sec: 14.4798 +INFO:tensorflow:step = 49601, loss = 0.172457, precision = 0.992188 (6.907 sec) +Saved checkpoint after 127 epoch(s) to data/resnet56/checkpoints/00127... +INFO:tensorflow:global_step/sec: 12.4966 +INFO:tensorflow:step = 49701, loss = 0.201663, precision = 0.984375 (8.002 sec) +INFO:tensorflow:global_step/sec: 14.3623 +INFO:tensorflow:step = 49801, loss = 0.206688, precision = 0.976562 (6.963 sec) +INFO:tensorflow:global_step/sec: 14.4015 +INFO:tensorflow:step = 49901, loss = 0.198962, precision = 0.976562 (6.944 sec) +INFO:tensorflow:global_step/sec: 14.4462 +INFO:tensorflow:step = 50001, loss = 0.197217, precision = 0.96875 (6.922 sec) +Saved checkpoint after 128 epoch(s) to data/resnet56/checkpoints/00128... +INFO:tensorflow:global_step/sec: 12.5542 +INFO:tensorflow:step = 50101, loss = 0.22015, precision = 0.976562 (7.966 sec) +INFO:tensorflow:global_step/sec: 14.4027 +INFO:tensorflow:step = 50201, loss = 0.223824, precision = 0.96875 (6.943 sec) +INFO:tensorflow:global_step/sec: 14.3814 +INFO:tensorflow:step = 50301, loss = 0.196196, precision = 0.976562 (6.953 sec) +INFO:tensorflow:global_step/sec: 14.4408 +INFO:tensorflow:step = 50401, loss = 0.201276, precision = 0.976562 (6.925 sec) +Saved checkpoint after 129 epoch(s) to data/resnet56/checkpoints/00129... +INFO:tensorflow:global_step/sec: 12.4801 +INFO:tensorflow:step = 50501, loss = 0.223438, precision = 0.976562 (8.013 sec) +INFO:tensorflow:global_step/sec: 14.458 +INFO:tensorflow:step = 50601, loss = 0.172683, precision = 0.992188 (6.917 sec) +INFO:tensorflow:global_step/sec: 14.3926 +INFO:tensorflow:step = 50701, loss = 0.24392, precision = 0.96875 (6.948 sec) +INFO:tensorflow:global_step/sec: 14.4528 +INFO:tensorflow:step = 50801, loss = 0.214655, precision = 0.984375 (6.920 sec) +Saved checkpoint after 130 epoch(s) to data/resnet56/checkpoints/00130... +INFO:tensorflow:global_step/sec: 12.4684 +INFO:tensorflow:step = 50901, loss = 0.19497, precision = 0.976562 (8.019 sec) +INFO:tensorflow:global_step/sec: 14.4228 +INFO:tensorflow:step = 51001, loss = 0.174263, precision = 0.992188 (6.933 sec) +INFO:tensorflow:global_step/sec: 14.4635 +INFO:tensorflow:step = 51101, loss = 0.167916, precision = 0.992188 (6.914 sec) +INFO:tensorflow:global_step/sec: 14.4043 +INFO:tensorflow:step = 51201, loss = 0.208279, precision = 0.96875 (6.942 sec) +Saved checkpoint after 131 epoch(s) to data/resnet56/checkpoints/00131... +INFO:tensorflow:global_step/sec: 12.4375 +INFO:tensorflow:step = 51301, loss = 0.220595, precision = 0.976562 (8.042 sec) +INFO:tensorflow:global_step/sec: 14.439 +INFO:tensorflow:step = 51401, loss = 0.161422, precision = 1.0 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.4252 +INFO:tensorflow:step = 51501, loss = 0.172028, precision = 0.992188 (6.933 sec) +INFO:tensorflow:global_step/sec: 14.3393 +INFO:tensorflow:step = 51601, loss = 0.193636, precision = 0.976562 (6.973 sec) +Saved checkpoint after 132 epoch(s) to data/resnet56/checkpoints/00132... +INFO:tensorflow:global_step/sec: 12.4448 +INFO:tensorflow:step = 51701, loss = 0.248359, precision = 0.960938 (8.035 sec) +INFO:tensorflow:global_step/sec: 14.4769 +INFO:tensorflow:step = 51801, loss = 0.238182, precision = 0.945312 (6.908 sec) +INFO:tensorflow:global_step/sec: 14.4728 +INFO:tensorflow:step = 51901, loss = 0.212741, precision = 0.984375 (6.909 sec) +INFO:tensorflow:global_step/sec: 14.4713 +INFO:tensorflow:step = 52001, loss = 0.218244, precision = 0.976562 (6.910 sec) +Saved checkpoint after 133 epoch(s) to data/resnet56/checkpoints/00133... +INFO:tensorflow:global_step/sec: 12.2582 +INFO:tensorflow:step = 52101, loss = 0.209252, precision = 0.976562 (8.157 sec) +INFO:tensorflow:global_step/sec: 14.374 +INFO:tensorflow:step = 52201, loss = 0.184481, precision = 0.976562 (6.957 sec) +INFO:tensorflow:global_step/sec: 14.4249 +INFO:tensorflow:step = 52301, loss = 0.172599, precision = 1.0 (6.932 sec) +Saved checkpoint after 134 epoch(s) to data/resnet56/checkpoints/00134... +INFO:tensorflow:global_step/sec: 12.42 +INFO:tensorflow:step = 52401, loss = 0.172412, precision = 0.992188 (8.051 sec) +INFO:tensorflow:global_step/sec: 14.4323 +INFO:tensorflow:step = 52501, loss = 0.152515, precision = 0.992188 (6.929 sec) +INFO:tensorflow:global_step/sec: 14.3979 +INFO:tensorflow:step = 52601, loss = 0.17198, precision = 0.984375 (6.946 sec) +INFO:tensorflow:global_step/sec: 14.4012 +INFO:tensorflow:step = 52701, loss = 0.185233, precision = 0.992188 (6.944 sec) +Saved checkpoint after 135 epoch(s) to data/resnet56/checkpoints/00135... +INFO:tensorflow:global_step/sec: 12.4715 +INFO:tensorflow:step = 52801, loss = 0.169079, precision = 1.0 (8.018 sec) +INFO:tensorflow:global_step/sec: 14.4271 +INFO:tensorflow:step = 52901, loss = 0.20549, precision = 0.992188 (6.931 sec) +INFO:tensorflow:global_step/sec: 14.3642 +INFO:tensorflow:step = 53001, loss = 0.20609, precision = 0.960938 (6.962 sec) +INFO:tensorflow:global_step/sec: 14.4344 +INFO:tensorflow:step = 53101, loss = 0.181211, precision = 0.984375 (6.928 sec) +Saved checkpoint after 136 epoch(s) to data/resnet56/checkpoints/00136... +INFO:tensorflow:global_step/sec: 12.4273 +INFO:tensorflow:step = 53201, loss = 0.178428, precision = 0.984375 (8.047 sec) +INFO:tensorflow:global_step/sec: 14.4268 +INFO:tensorflow:step = 53301, loss = 0.162914, precision = 0.992188 (6.932 sec) +INFO:tensorflow:global_step/sec: 14.4402 +INFO:tensorflow:step = 53401, loss = 0.152324, precision = 1.0 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.4837 +INFO:tensorflow:step = 53501, loss = 0.206659, precision = 0.96875 (6.904 sec) +Saved checkpoint after 137 epoch(s) to data/resnet56/checkpoints/00137... +INFO:tensorflow:global_step/sec: 12.4693 +INFO:tensorflow:step = 53601, loss = 0.182835, precision = 0.984375 (8.020 sec) +INFO:tensorflow:global_step/sec: 14.4621 +INFO:tensorflow:step = 53701, loss = 0.14166, precision = 1.0 (6.914 sec) +INFO:tensorflow:global_step/sec: 14.4931 +INFO:tensorflow:step = 53801, loss = 0.155455, precision = 1.0 (6.901 sec) +INFO:tensorflow:global_step/sec: 14.4739 +INFO:tensorflow:step = 53901, loss = 0.14439, precision = 1.0 (6.908 sec) +Saved checkpoint after 138 epoch(s) to data/resnet56/checkpoints/00138... +INFO:tensorflow:global_step/sec: 12.5386 +INFO:tensorflow:step = 54001, loss = 0.144411, precision = 1.0 (7.976 sec) +INFO:tensorflow:global_step/sec: 14.4665 +INFO:tensorflow:step = 54101, loss = 0.137974, precision = 1.0 (6.912 sec) +INFO:tensorflow:global_step/sec: 14.5084 +INFO:tensorflow:step = 54201, loss = 0.167386, precision = 0.992188 (6.892 sec) +INFO:tensorflow:global_step/sec: 14.4957 +INFO:tensorflow:step = 54301, loss = 0.147021, precision = 1.0 (6.899 sec) +Saved checkpoint after 139 epoch(s) to data/resnet56/checkpoints/00139... +INFO:tensorflow:global_step/sec: 12.5135 +INFO:tensorflow:step = 54401, loss = 0.14254, precision = 1.0 (7.991 sec) +INFO:tensorflow:global_step/sec: 14.4602 +INFO:tensorflow:step = 54501, loss = 0.143289, precision = 1.0 (6.915 sec) +INFO:tensorflow:global_step/sec: 14.4844 +INFO:tensorflow:step = 54601, loss = 0.161081, precision = 0.992188 (6.904 sec) +INFO:tensorflow:global_step/sec: 14.5199 +INFO:tensorflow:step = 54701, loss = 0.157948, precision = 1.0 (6.887 sec) +Saved checkpoint after 140 epoch(s) to data/resnet56/checkpoints/00140... +INFO:tensorflow:global_step/sec: 12.4408 +INFO:tensorflow:step = 54801, loss = 0.145701, precision = 1.0 (8.039 sec) +INFO:tensorflow:global_step/sec: 14.4561 +INFO:tensorflow:step = 54901, loss = 0.158478, precision = 0.984375 (6.917 sec) +INFO:tensorflow:global_step/sec: 14.501 +INFO:tensorflow:step = 55001, loss = 0.141549, precision = 1.0 (6.896 sec) +INFO:tensorflow:global_step/sec: 14.3807 +INFO:tensorflow:step = 55101, loss = 0.142393, precision = 1.0 (6.956 sec) +Saved checkpoint after 141 epoch(s) to data/resnet56/checkpoints/00141... +INFO:tensorflow:global_step/sec: 12.4657 +INFO:tensorflow:step = 55201, loss = 0.139895, precision = 1.0 (8.020 sec) +INFO:tensorflow:global_step/sec: 14.426 +INFO:tensorflow:step = 55301, loss = 0.152508, precision = 0.992188 (6.932 sec) +INFO:tensorflow:global_step/sec: 14.4607 +INFO:tensorflow:step = 55401, loss = 0.146428, precision = 1.0 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.4676 +INFO:tensorflow:step = 55501, loss = 0.150392, precision = 0.992188 (6.913 sec) +Saved checkpoint after 142 epoch(s) to data/resnet56/checkpoints/00142... +INFO:tensorflow:global_step/sec: 12.4481 +INFO:tensorflow:step = 55601, loss = 0.145344, precision = 1.0 (8.032 sec) +INFO:tensorflow:global_step/sec: 14.4372 +INFO:tensorflow:step = 55701, loss = 0.140177, precision = 1.0 (6.927 sec) +INFO:tensorflow:global_step/sec: 14.4639 +INFO:tensorflow:step = 55801, loss = 0.142948, precision = 1.0 (6.914 sec) +INFO:tensorflow:global_step/sec: 14.4522 +INFO:tensorflow:step = 55901, loss = 0.136858, precision = 1.0 (6.919 sec) +Saved checkpoint after 143 epoch(s) to data/resnet56/checkpoints/00143... +INFO:tensorflow:global_step/sec: 12.5232 +INFO:tensorflow:step = 56001, loss = 0.158048, precision = 0.992188 (7.985 sec) +INFO:tensorflow:global_step/sec: 14.4453 +INFO:tensorflow:step = 56101, loss = 0.14424, precision = 1.0 (6.922 sec) +INFO:tensorflow:global_step/sec: 14.4353 +INFO:tensorflow:step = 56201, loss = 0.144235, precision = 1.0 (6.927 sec) +INFO:tensorflow:global_step/sec: 14.4838 +INFO:tensorflow:step = 56301, loss = 0.13684, precision = 1.0 (6.904 sec) +Saved checkpoint after 144 epoch(s) to data/resnet56/checkpoints/00144... +INFO:tensorflow:global_step/sec: 12.4993 +INFO:tensorflow:step = 56401, loss = 0.147035, precision = 0.992188 (8.001 sec) +INFO:tensorflow:global_step/sec: 14.4531 +INFO:tensorflow:step = 56501, loss = 0.147619, precision = 0.992188 (6.919 sec) +INFO:tensorflow:global_step/sec: 14.4284 +INFO:tensorflow:step = 56601, loss = 0.147931, precision = 0.992188 (6.931 sec) +Saved checkpoint after 145 epoch(s) to data/resnet56/checkpoints/00145... +INFO:tensorflow:global_step/sec: 12.4818 +INFO:tensorflow:step = 56701, loss = 0.147048, precision = 0.992188 (8.011 sec) +INFO:tensorflow:global_step/sec: 14.3979 +INFO:tensorflow:step = 56801, loss = 0.134268, precision = 1.0 (6.946 sec) +INFO:tensorflow:global_step/sec: 14.4027 +INFO:tensorflow:step = 56901, loss = 0.137119, precision = 1.0 (6.943 sec) +INFO:tensorflow:global_step/sec: 14.4242 +INFO:tensorflow:step = 57001, loss = 0.151474, precision = 0.992188 (6.935 sec) +Saved checkpoint after 146 epoch(s) to data/resnet56/checkpoints/00146... +INFO:tensorflow:global_step/sec: 12.2335 +INFO:tensorflow:step = 57101, loss = 0.139334, precision = 1.0 (8.172 sec) +INFO:tensorflow:global_step/sec: 14.528 +INFO:tensorflow:step = 57201, loss = 0.134475, precision = 1.0 (6.883 sec) +INFO:tensorflow:global_step/sec: 14.4005 +INFO:tensorflow:step = 57301, loss = 0.133083, precision = 1.0 (6.945 sec) +INFO:tensorflow:global_step/sec: 14.4132 +INFO:tensorflow:step = 57401, loss = 0.14653, precision = 0.992188 (6.938 sec) +Saved checkpoint after 147 epoch(s) to data/resnet56/checkpoints/00147... +INFO:tensorflow:global_step/sec: 12.4306 +INFO:tensorflow:step = 57501, loss = 0.134402, precision = 1.0 (8.044 sec) +INFO:tensorflow:global_step/sec: 14.5098 +INFO:tensorflow:step = 57601, loss = 0.14015, precision = 0.992188 (6.892 sec) +INFO:tensorflow:global_step/sec: 14.4545 +INFO:tensorflow:step = 57701, loss = 0.134688, precision = 1.0 (6.919 sec) +INFO:tensorflow:global_step/sec: 14.4417 +INFO:tensorflow:step = 57801, loss = 0.138932, precision = 1.0 (6.924 sec) +Saved checkpoint after 148 epoch(s) to data/resnet56/checkpoints/00148... +INFO:tensorflow:global_step/sec: 12.4448 +INFO:tensorflow:step = 57901, loss = 0.135367, precision = 1.0 (8.035 sec) +INFO:tensorflow:global_step/sec: 14.5139 +INFO:tensorflow:step = 58001, loss = 0.143884, precision = 0.992188 (6.890 sec) +INFO:tensorflow:global_step/sec: 14.4793 +INFO:tensorflow:step = 58101, loss = 0.136727, precision = 1.0 (6.907 sec) +INFO:tensorflow:global_step/sec: 14.5168 +INFO:tensorflow:step = 58201, loss = 0.132584, precision = 1.0 (6.889 sec) +Saved checkpoint after 149 epoch(s) to data/resnet56/checkpoints/00149... +INFO:tensorflow:global_step/sec: 12.4674 +INFO:tensorflow:step = 58301, loss = 0.14097, precision = 1.0 (8.021 sec) +INFO:tensorflow:global_step/sec: 14.4091 +INFO:tensorflow:step = 58401, loss = 0.145587, precision = 0.992188 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.4619 +INFO:tensorflow:step = 58501, loss = 0.132719, precision = 1.0 (6.915 sec) +INFO:tensorflow:global_step/sec: 14.4719 +INFO:tensorflow:step = 58601, loss = 0.132672, precision = 1.0 (6.910 sec) +Saved checkpoint after 150 epoch(s) to data/resnet56/checkpoints/00150... +INFO:tensorflow:global_step/sec: 12.4438 +INFO:tensorflow:step = 58701, loss = 0.132481, precision = 1.0 (8.036 sec) +INFO:tensorflow:global_step/sec: 14.5055 +INFO:tensorflow:step = 58801, loss = 0.13921, precision = 1.0 (6.894 sec) +INFO:tensorflow:global_step/sec: 14.4794 +INFO:tensorflow:step = 58901, loss = 0.1311, precision = 1.0 (6.906 sec) +INFO:tensorflow:global_step/sec: 14.4631 +INFO:tensorflow:step = 59001, loss = 0.147582, precision = 0.992188 (6.914 sec) +Saved checkpoint after 151 epoch(s) to data/resnet56/checkpoints/00151... +INFO:tensorflow:global_step/sec: 12.5669 +INFO:tensorflow:step = 59101, loss = 0.136153, precision = 1.0 (7.957 sec) +INFO:tensorflow:global_step/sec: 14.4403 +INFO:tensorflow:step = 59201, loss = 0.136325, precision = 1.0 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.4301 +INFO:tensorflow:step = 59301, loss = 0.131321, precision = 1.0 (6.930 sec) +INFO:tensorflow:global_step/sec: 14.4668 +INFO:tensorflow:step = 59401, loss = 0.131438, precision = 1.0 (6.912 sec) +Saved checkpoint after 152 epoch(s) to data/resnet56/checkpoints/00152... +INFO:tensorflow:global_step/sec: 12.5282 +INFO:tensorflow:step = 59501, loss = 0.159215, precision = 0.992188 (7.982 sec) +INFO:tensorflow:global_step/sec: 14.3978 +INFO:tensorflow:step = 59601, loss = 0.130302, precision = 1.0 (6.945 sec) +INFO:tensorflow:global_step/sec: 14.4817 +INFO:tensorflow:step = 59701, loss = 0.130307, precision = 1.0 (6.905 sec) +INFO:tensorflow:global_step/sec: 14.4111 +INFO:tensorflow:step = 59801, loss = 0.131409, precision = 1.0 (6.939 sec) +Saved checkpoint after 153 epoch(s) to data/resnet56/checkpoints/00153... +INFO:tensorflow:global_step/sec: 12.4347 +INFO:tensorflow:step = 59901, loss = 0.150103, precision = 0.992188 (8.042 sec) +INFO:tensorflow:global_step/sec: 14.46 +INFO:tensorflow:step = 60001, loss = 0.13469, precision = 0.992188 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.4584 +INFO:tensorflow:step = 60101, loss = 0.132582, precision = 1.0 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.4843 +INFO:tensorflow:step = 60201, loss = 0.130649, precision = 1.0 (6.904 sec) +Saved checkpoint after 154 epoch(s) to data/resnet56/checkpoints/00154... +INFO:tensorflow:global_step/sec: 12.3628 +INFO:tensorflow:step = 60301, loss = 0.131232, precision = 1.0 (8.089 sec) +INFO:tensorflow:global_step/sec: 14.465 +INFO:tensorflow:step = 60401, loss = 0.13344, precision = 1.0 (6.913 sec) +INFO:tensorflow:global_step/sec: 14.4486 +INFO:tensorflow:step = 60501, loss = 0.154401, precision = 0.992188 (6.921 sec) +INFO:tensorflow:global_step/sec: 14.4508 +INFO:tensorflow:step = 60601, loss = 0.138218, precision = 0.992188 (6.920 sec) +Saved checkpoint after 155 epoch(s) to data/resnet56/checkpoints/00155... +INFO:tensorflow:global_step/sec: 12.4325 +INFO:tensorflow:step = 60701, loss = 0.126682, precision = 1.0 (8.044 sec) +INFO:tensorflow:global_step/sec: 14.4014 +INFO:tensorflow:step = 60801, loss = 0.131447, precision = 1.0 (6.944 sec) +INFO:tensorflow:global_step/sec: 14.4684 +INFO:tensorflow:step = 60901, loss = 0.139052, precision = 0.992188 (6.912 sec) +Saved checkpoint after 156 epoch(s) to data/resnet56/checkpoints/00156... +INFO:tensorflow:global_step/sec: 12.5296 +INFO:tensorflow:step = 61001, loss = 0.131403, precision = 1.0 (7.981 sec) +INFO:tensorflow:global_step/sec: 14.4121 +INFO:tensorflow:step = 61101, loss = 0.127927, precision = 1.0 (6.939 sec) +INFO:tensorflow:global_step/sec: 14.4376 +INFO:tensorflow:step = 61201, loss = 0.126893, precision = 1.0 (6.926 sec) +INFO:tensorflow:global_step/sec: 14.4335 +INFO:tensorflow:step = 61301, loss = 0.128442, precision = 1.0 (6.928 sec) +Saved checkpoint after 157 epoch(s) to data/resnet56/checkpoints/00157... +INFO:tensorflow:global_step/sec: 12.4565 +INFO:tensorflow:step = 61401, loss = 0.128822, precision = 1.0 (8.028 sec) +INFO:tensorflow:global_step/sec: 14.4042 +INFO:tensorflow:step = 61501, loss = 0.142795, precision = 0.992188 (6.942 sec) +INFO:tensorflow:global_step/sec: 14.4503 +INFO:tensorflow:step = 61601, loss = 0.128494, precision = 1.0 (6.920 sec) +INFO:tensorflow:global_step/sec: 14.517 +INFO:tensorflow:step = 61701, loss = 0.127612, precision = 1.0 (6.888 sec) +Saved checkpoint after 158 epoch(s) to data/resnet56/checkpoints/00158... +INFO:tensorflow:global_step/sec: 12.3597 +INFO:tensorflow:step = 61801, loss = 0.133719, precision = 1.0 (8.091 sec) +INFO:tensorflow:global_step/sec: 14.4739 +INFO:tensorflow:step = 61901, loss = 0.13216, precision = 1.0 (6.909 sec) +INFO:tensorflow:global_step/sec: 14.5243 +INFO:tensorflow:step = 62001, loss = 0.125081, precision = 1.0 (6.885 sec) +INFO:tensorflow:global_step/sec: 14.5741 +INFO:tensorflow:step = 62101, loss = 0.130899, precision = 1.0 (6.862 sec) +Saved checkpoint after 159 epoch(s) to data/resnet56/checkpoints/00159... +INFO:tensorflow:global_step/sec: 12.4482 +INFO:tensorflow:step = 62201, loss = 0.132652, precision = 0.992188 (8.034 sec) +INFO:tensorflow:global_step/sec: 14.4939 +INFO:tensorflow:step = 62301, loss = 0.126362, precision = 1.0 (6.899 sec) +INFO:tensorflow:global_step/sec: 14.482 +INFO:tensorflow:step = 62401, loss = 0.127115, precision = 1.0 (6.905 sec) +INFO:tensorflow:global_step/sec: 14.4596 +INFO:tensorflow:step = 62501, loss = 0.127745, precision = 1.0 (6.916 sec) +Saved checkpoint after 160 epoch(s) to data/resnet56/checkpoints/00160... +INFO:tensorflow:global_step/sec: 12.4823 +INFO:tensorflow:step = 62601, loss = 0.128324, precision = 1.0 (8.011 sec) +INFO:tensorflow:global_step/sec: 14.4697 +INFO:tensorflow:step = 62701, loss = 0.137888, precision = 0.992188 (6.911 sec) +INFO:tensorflow:global_step/sec: 14.5074 +INFO:tensorflow:step = 62801, loss = 0.13715, precision = 0.992188 (6.893 sec) +INFO:tensorflow:global_step/sec: 14.4711 +INFO:tensorflow:step = 62901, loss = 0.131156, precision = 1.0 (6.910 sec) +Saved checkpoint after 161 epoch(s) to data/resnet56/checkpoints/00161... +INFO:tensorflow:global_step/sec: 12.4223 +INFO:tensorflow:step = 63001, loss = 0.143552, precision = 0.992188 (8.050 sec) +INFO:tensorflow:global_step/sec: 14.4133 +INFO:tensorflow:step = 63101, loss = 0.130571, precision = 0.992188 (6.938 sec) +INFO:tensorflow:global_step/sec: 14.4675 +INFO:tensorflow:step = 63201, loss = 0.126352, precision = 1.0 (6.912 sec) +INFO:tensorflow:global_step/sec: 14.4193 +INFO:tensorflow:step = 63301, loss = 0.126525, precision = 1.0 (6.935 sec) +Saved checkpoint after 162 epoch(s) to data/resnet56/checkpoints/00162... +INFO:tensorflow:global_step/sec: 12.4425 +INFO:tensorflow:step = 63401, loss = 0.131296, precision = 0.992188 (8.037 sec) +INFO:tensorflow:global_step/sec: 14.4385 +INFO:tensorflow:step = 63501, loss = 0.143465, precision = 0.992188 (6.926 sec) +INFO:tensorflow:global_step/sec: 14.3825 +INFO:tensorflow:step = 63601, loss = 0.129613, precision = 1.0 (6.953 sec) +INFO:tensorflow:global_step/sec: 14.4751 +INFO:tensorflow:step = 63701, loss = 0.125146, precision = 1.0 (6.908 sec) +Saved checkpoint after 163 epoch(s) to data/resnet56/checkpoints/00163... +INFO:tensorflow:global_step/sec: 12.4837 +INFO:tensorflow:step = 63801, loss = 0.125051, precision = 1.0 (8.011 sec) +INFO:tensorflow:global_step/sec: 14.4398 +INFO:tensorflow:step = 63901, loss = 0.125913, precision = 1.0 (6.925 sec) +INFO:tensorflow:global_step/sec: 14.3877 +INFO:tensorflow:step = 64001, loss = 0.124728, precision = 1.0 (6.950 sec) +INFO:tensorflow:global_step/sec: 14.528 +INFO:tensorflow:step = 64101, loss = 0.139697, precision = 0.992188 (6.883 sec) +Saved checkpoint after 164 epoch(s) to data/resnet56/checkpoints/00164... +INFO:tensorflow:global_step/sec: 12.5769 +INFO:tensorflow:step = 64201, loss = 0.123551, precision = 1.0 (7.951 sec) +INFO:tensorflow:global_step/sec: 14.4386 +INFO:tensorflow:step = 64301, loss = 0.144965, precision = 0.992188 (6.926 sec) +INFO:tensorflow:global_step/sec: 14.5145 +INFO:tensorflow:step = 64401, loss = 0.126635, precision = 1.0 (6.890 sec) +INFO:tensorflow:global_step/sec: 14.4252 +INFO:tensorflow:step = 64501, loss = 0.127404, precision = 1.0 (6.933 sec) +Saved checkpoint after 165 epoch(s) to data/resnet56/checkpoints/00165... +INFO:tensorflow:global_step/sec: 12.3997 +INFO:tensorflow:step = 64601, loss = 0.122559, precision = 1.0 (8.065 sec) +INFO:tensorflow:global_step/sec: 14.4453 +INFO:tensorflow:step = 64701, loss = 0.128832, precision = 1.0 (6.923 sec) +INFO:tensorflow:global_step/sec: 14.4531 +INFO:tensorflow:step = 64801, loss = 0.124165, precision = 1.0 (6.919 sec) +INFO:tensorflow:global_step/sec: 14.4557 +INFO:tensorflow:step = 64901, loss = 0.134202, precision = 0.992188 (6.918 sec) +Saved checkpoint after 166 epoch(s) to data/resnet56/checkpoints/00166... +INFO:tensorflow:global_step/sec: 12.4472 +INFO:tensorflow:step = 65001, loss = 0.122589, precision = 1.0 (8.034 sec) +INFO:tensorflow:global_step/sec: 14.4056 +INFO:tensorflow:step = 65101, loss = 0.123148, precision = 1.0 (6.942 sec) +INFO:tensorflow:global_step/sec: 14.4693 +INFO:tensorflow:step = 65201, loss = 0.123064, precision = 1.0 (6.911 sec) +Saved checkpoint after 167 epoch(s) to data/resnet56/checkpoints/00167... +INFO:tensorflow:global_step/sec: 12.4665 +INFO:tensorflow:step = 65301, loss = 0.123187, precision = 1.0 (8.021 sec) +INFO:tensorflow:global_step/sec: 14.367 +INFO:tensorflow:step = 65401, loss = 0.140728, precision = 0.992188 (6.961 sec) +INFO:tensorflow:global_step/sec: 14.5013 +INFO:tensorflow:step = 65501, loss = 0.12518, precision = 1.0 (6.896 sec) +INFO:tensorflow:global_step/sec: 14.4819 +INFO:tensorflow:step = 65601, loss = 0.121548, precision = 1.0 (6.905 sec) +Saved checkpoint after 168 epoch(s) to data/resnet56/checkpoints/00168... +INFO:tensorflow:global_step/sec: 12.4485 +INFO:tensorflow:step = 65701, loss = 0.125072, precision = 1.0 (8.033 sec) +INFO:tensorflow:global_step/sec: 14.4082 +INFO:tensorflow:step = 65801, loss = 0.122125, precision = 1.0 (6.940 sec) +INFO:tensorflow:global_step/sec: 14.4331 +INFO:tensorflow:step = 65901, loss = 0.125026, precision = 1.0 (6.929 sec) +INFO:tensorflow:global_step/sec: 14.4666 +INFO:tensorflow:step = 66001, loss = 0.124506, precision = 1.0 (6.912 sec) +Saved checkpoint after 169 epoch(s) to data/resnet56/checkpoints/00169... +INFO:tensorflow:global_step/sec: 12.4903 +INFO:tensorflow:step = 66101, loss = 0.120529, precision = 1.0 (8.006 sec) +INFO:tensorflow:global_step/sec: 14.4652 +INFO:tensorflow:step = 66201, loss = 0.122351, precision = 1.0 (6.913 sec) +INFO:tensorflow:global_step/sec: 14.414 +INFO:tensorflow:step = 66301, loss = 0.121068, precision = 1.0 (6.938 sec) +INFO:tensorflow:global_step/sec: 14.4895 +INFO:tensorflow:step = 66401, loss = 0.126548, precision = 1.0 (6.901 sec) +Saved checkpoint after 170 epoch(s) to data/resnet56/checkpoints/00170... +INFO:tensorflow:global_step/sec: 12.4987 +INFO:tensorflow:step = 66501, loss = 0.121984, precision = 1.0 (8.001 sec) +INFO:tensorflow:global_step/sec: 14.4448 +INFO:tensorflow:step = 66601, loss = 0.125491, precision = 1.0 (6.922 sec) +INFO:tensorflow:global_step/sec: 14.4881 +INFO:tensorflow:step = 66701, loss = 0.120697, precision = 1.0 (6.903 sec) +INFO:tensorflow:global_step/sec: 14.4113 +INFO:tensorflow:step = 66801, loss = 0.11915, precision = 1.0 (6.939 sec) +Saved checkpoint after 171 epoch(s) to data/resnet56/checkpoints/00171... +INFO:tensorflow:global_step/sec: 12.3378 +INFO:tensorflow:step = 66901, loss = 0.123308, precision = 1.0 (8.105 sec) +INFO:tensorflow:global_step/sec: 14.4119 +INFO:tensorflow:step = 67001, loss = 0.128841, precision = 0.992188 (6.939 sec) +INFO:tensorflow:global_step/sec: 14.4356 +INFO:tensorflow:step = 67101, loss = 0.128086, precision = 1.0 (6.927 sec) +INFO:tensorflow:global_step/sec: 14.4444 +INFO:tensorflow:step = 67201, loss = 0.142305, precision = 0.992188 (6.923 sec) +Saved checkpoint after 172 epoch(s) to data/resnet56/checkpoints/00172... +INFO:tensorflow:global_step/sec: 12.5151 +INFO:tensorflow:step = 67301, loss = 0.120214, precision = 1.0 (7.990 sec) +INFO:tensorflow:global_step/sec: 14.4932 +INFO:tensorflow:step = 67401, loss = 0.119841, precision = 1.0 (6.900 sec) +INFO:tensorflow:global_step/sec: 14.5256 +INFO:tensorflow:step = 67501, loss = 0.123074, precision = 1.0 (6.884 sec) +INFO:tensorflow:global_step/sec: 14.4777 +INFO:tensorflow:step = 67601, loss = 0.120415, precision = 1.0 (6.907 sec) +Saved checkpoint after 173 epoch(s) to data/resnet56/checkpoints/00173... +INFO:tensorflow:global_step/sec: 12.4862 +INFO:tensorflow:step = 67701, loss = 0.120939, precision = 1.0 (8.009 sec) +INFO:tensorflow:global_step/sec: 14.4844 +INFO:tensorflow:step = 67801, loss = 0.128132, precision = 1.0 (6.904 sec) +INFO:tensorflow:global_step/sec: 14.445 +INFO:tensorflow:step = 67901, loss = 0.141256, precision = 0.992188 (6.923 sec) +INFO:tensorflow:global_step/sec: 14.443 +INFO:tensorflow:step = 68001, loss = 0.122619, precision = 1.0 (6.924 sec) +Saved checkpoint after 174 epoch(s) to data/resnet56/checkpoints/00174... +INFO:tensorflow:global_step/sec: 12.4286 +INFO:tensorflow:step = 68101, loss = 0.121948, precision = 1.0 (8.046 sec) +INFO:tensorflow:global_step/sec: 14.432 +INFO:tensorflow:step = 68201, loss = 0.119037, precision = 1.0 (6.929 sec) +INFO:tensorflow:global_step/sec: 14.4588 +INFO:tensorflow:step = 68301, loss = 0.118827, precision = 1.0 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.4949 +INFO:tensorflow:step = 68401, loss = 0.1212, precision = 1.0 (6.899 sec) +Saved checkpoint after 175 epoch(s) to data/resnet56/checkpoints/00175... +INFO:tensorflow:global_step/sec: 12.501 +INFO:tensorflow:step = 68501, loss = 0.117777, precision = 1.0 (7.999 sec) +INFO:tensorflow:global_step/sec: 14.5004 +INFO:tensorflow:step = 68601, loss = 0.117773, precision = 1.0 (6.896 sec) +INFO:tensorflow:global_step/sec: 14.4304 +INFO:tensorflow:step = 68701, loss = 0.120981, precision = 1.0 (6.930 sec) +INFO:tensorflow:global_step/sec: 14.4541 +INFO:tensorflow:step = 68801, loss = 0.117743, precision = 1.0 (6.918 sec) +Saved checkpoint after 176 epoch(s) to data/resnet56/checkpoints/00176... +INFO:tensorflow:global_step/sec: 12.5463 +INFO:tensorflow:step = 68901, loss = 0.118769, precision = 1.0 (7.971 sec) +INFO:tensorflow:global_step/sec: 14.4582 +INFO:tensorflow:step = 69001, loss = 0.119229, precision = 1.0 (6.916 sec) +INFO:tensorflow:global_step/sec: 14.4356 +INFO:tensorflow:step = 69101, loss = 0.124422, precision = 1.0 (6.927 sec) +INFO:tensorflow:global_step/sec: 14.4396 +INFO:tensorflow:step = 69201, loss = 0.118451, precision = 1.0 (6.925 sec) +Saved checkpoint after 177 epoch(s) to data/resnet56/checkpoints/00177... +INFO:tensorflow:global_step/sec: 12.4635 +INFO:tensorflow:step = 69301, loss = 0.118792, precision = 1.0 (8.023 sec) +INFO:tensorflow:global_step/sec: 14.4546 +INFO:tensorflow:step = 69401, loss = 0.119287, precision = 1.0 (6.918 sec) +INFO:tensorflow:global_step/sec: 14.4313 +INFO:tensorflow:step = 69501, loss = 0.118832, precision = 1.0 (6.929 sec) +Saved checkpoint after 178 epoch(s) to data/resnet56/checkpoints/00178... +INFO:tensorflow:global_step/sec: 12.6383 +INFO:tensorflow:step = 69601, loss = 0.117208, precision = 1.0 (7.913 sec) +INFO:tensorflow:global_step/sec: 14.4333 +INFO:tensorflow:step = 69701, loss = 0.121302, precision = 1.0 (6.928 sec) +INFO:tensorflow:global_step/sec: 14.489 +INFO:tensorflow:step = 69801, loss = 0.122815, precision = 1.0 (6.902 sec) +INFO:tensorflow:global_step/sec: 14.4036 +INFO:tensorflow:step = 69901, loss = 0.118083, precision = 1.0 (6.943 sec) +Saved checkpoint after 179 epoch(s) to data/resnet56/checkpoints/00179... +INFO:tensorflow:global_step/sec: 12.5201 +INFO:tensorflow:step = 70001, loss = 0.130754, precision = 0.992188 (7.987 sec) +INFO:tensorflow:global_step/sec: 14.4923 +INFO:tensorflow:step = 70101, loss = 0.115948, precision = 1.0 (6.900 sec) +INFO:tensorflow:global_step/sec: 14.422 +INFO:tensorflow:step = 70201, loss = 0.118991, precision = 1.0 (6.934 sec) +INFO:tensorflow:global_step/sec: 14.4792 +INFO:tensorflow:step = 70301, loss = 0.116414, precision = 1.0 (6.906 sec) +Saved checkpoint after 180 epoch(s) to data/resnet56/checkpoints/00180... +INFO:tensorflow:global_step/sec: 12.4902 +INFO:tensorflow:step = 70401, loss = 0.126046, precision = 1.0 (8.006 sec) +INFO:tensorflow:global_step/sec: 14.4912 +INFO:tensorflow:step = 70501, loss = 0.117374, precision = 1.0 (6.901 sec) +INFO:tensorflow:global_step/sec: 14.447 +INFO:tensorflow:step = 70601, loss = 0.116976, precision = 1.0 (6.921 sec) +INFO:tensorflow:global_step/sec: 14.4251 +INFO:tensorflow:step = 70701, loss = 0.117075, precision = 1.0 (6.933 sec) +Saved checkpoint after 181 epoch(s) to data/resnet56/checkpoints/00181... diff --git a/tensorflow/CIFAR10/resnet/README.md b/tensorflow/CIFAR10/resnet/README.md new file mode 100644 index 0000000..b8027c1 --- /dev/null +++ b/tensorflow/CIFAR10/resnet/README.md @@ -0,0 +1,88 @@ +# ResNet on CIFAR10 and CIFAR100 + +(Borrowed from the tensorflow/models repository) + +## Dataset + +https://www.cs.toronto.edu/~kriz/cifar.html + +## Related papers + +- [Identity Mappings in Deep Residual Networks](https://arxiv.org/pdf/1603.05027v2.pdf) +- [Deep Residual Learning for Image Recognition](https://arxiv.org/pdf/1512.03385v1.pdf) +- [Wide Residual Networks](https://arxiv.org/pdf/1605.07146v1.pdf) + +## Setting + +* Pad to 36x36 and random crop. Horizontal flip. Per-image whitening. +* Momentum optimizer (momentum = 0.9). +* Learning rate schedule: 0.01 (1 epoch), 0.1 (90 epochs), 0.01 (45 epochs), 0.001 (45 epochs). +* L2 weight decay: 0.005. +* Batch size: 128. (28-10 wide and 1001 layer bottleneck use 64) + +## Results + +CIFAR-10 Model|Best Precision|Steps +--------------|--------------|------ +32 layer|92.5%|~80k +110 layer|93.6%|~80k +164 layer bottleneck|94.5%|~80k +1001 layer bottleneck|94.9%|~80k +28-10 wide|95%|~90k + +CIFAR-100 Model|Best Precision|Steps +---------------|--------------|----- +32 layer|68.1%|~45k +110 layer|71.3%|~60k +164 layer bottleneck|75.7%|~50k +1001 layer bottleneck|78.2%|~70k +28-10 wide|78.3%|~70k + +## Prerequisites + +1. Install TensorFlow 1.2 (preferably from source for higher performance) and Python 3.6.2. + +2. Download CIFAR-10/CIFAR-100 dataset. + +```shell +curl -o cifar-10-binary.tar.gz https://www.cs.toronto.edu/~kriz/cifar-10-binary.tar.gz +curl -o cifar-100-binary.tar.gz https://www.cs.toronto.edu/~kriz/cifar-100-binary.tar.gz +``` + +## How to run + +```shell +# cd to the models repository and run with bash. Expected command output shown. +# The directory should contain an empty WORKSPACE file, the resnet code, and the cifar10 dataset. +# Note: The user can split 5k from train set for eval set. +$ ls -R +.: +cifar10 resnet WORKSPACE + +./cifar10: +data_batch_1.bin data_batch_2.bin data_batch_3.bin data_batch_4.bin +data_batch_5.bin test_batch.bin + +./resnet: +cifar_input.py README.md resnet_main.py resnet_model.py + +# Train the model. +$ python3 resnet/resnet_main.py --train_data_path=cifar10/data_batch* \ + --log_root=/tmp/resnet_model \ + --train_dir=/tmp/resnet_model/train \ + --dataset='cifar10' \ + --num_gpus=1 + +# While the model is training, you can also check on its progress using tensorboard: +$ tensorboard --logdir=/tmp/resnet_model + +# Evaluate the model. +# Avoid running on the same GPU as the training job at the same time, +# otherwise, you might run out of memory. +$ python3 resnet/resnet_main.py --eval_data_path=cifar10/test_batch.bin \ + --log_root=/tmp/resnet_model \ + --eval_dir=/tmp/resnet_model/test \ + --mode=eval \ + --dataset='cifar10' \ + --num_gpus=0 +``` diff --git a/tensorflow/CIFAR10/resnet/cifar_input.py b/tensorflow/CIFAR10/resnet/cifar_input.py new file mode 100644 index 0000000..b0c5a07 --- /dev/null +++ b/tensorflow/CIFAR10/resnet/cifar_input.py @@ -0,0 +1,121 @@ +# Copyright 2016 The TensorFlow Authors. All Rights Reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# ============================================================================== + +"""CIFAR dataset input module. +""" + +import tensorflow as tf + +def build_input(dataset, data_path, batch_size, mode, data_format): + """Build CIFAR image and labels. + + Args: + dataset: Either 'cifar10' or 'cifar100'. + data_path: Filename for data. + batch_size: Input batch size. + mode: Either 'train' or 'eval'. + data_format: Either 'NCHW' or 'NHWC'. + Returns: + images: Batches of images. [batch_size, image_size, image_size, 3] + labels: Batches of labels. [batch_size, num_classes] + Raises: + ValueError: when the specified dataset is not supported. + """ + with tf.device('/cpu:0'): + image_size = 32 + if dataset == 'cifar10': + label_bytes = 1 + label_offset = 0 + num_classes = 10 + elif dataset == 'cifar100': + label_bytes = 1 + label_offset = 1 + num_classes = 100 + else: + raise ValueError('Not supported dataset %s', dataset) + + depth = 3 + image_bytes = image_size * image_size * depth + record_bytes = label_bytes + label_offset + image_bytes + + data_files = tf.gfile.Glob(data_path) + file_queue = tf.train.string_input_producer(data_files, shuffle=True) + # Read examples from files in the filename queue. + reader = tf.FixedLengthRecordReader(record_bytes=record_bytes) + _, value = reader.read(file_queue) + + # Convert these examples to dense labels and processed images. + record = tf.reshape(tf.decode_raw(value, tf.uint8), [record_bytes]) + label = tf.cast(tf.slice(record, [label_offset], [label_bytes]), tf.int32) + # Convert from string to [depth * height * width] to [depth, height, width]. + depth_major = tf.reshape(tf.slice(record, [label_bytes], [image_bytes]), + [depth, image_size, image_size]) + # Convert from [depth, height, width] to [height, width, depth]. + image = tf.cast(tf.transpose(depth_major, [1, 2, 0]), tf.float32) + + if mode == 'train': + image = tf.image.resize_image_with_crop_or_pad( + image, image_size+4, image_size+4) + image = tf.random_crop(image, [image_size, image_size, 3]) + image = tf.image.random_flip_left_right(image) + # Brightness/saturation/constrast provides small gains .2%~.5% on cifar. + # image = tf.image.random_brightness(image, max_delta=63. / 255.) + # image = tf.image.random_saturation(image, lower=0.5, upper=1.5) + # image = tf.image.random_contrast(image, lower=0.2, upper=1.8) + image = tf.image.per_image_standardization(image) + + example_queue = tf.RandomShuffleQueue( + capacity=16 * batch_size, + min_after_dequeue=8 * batch_size, + dtypes=[tf.float32, tf.int32], + shapes=[[image_size, image_size, depth], [1]]) + num_threads = 16 + else: + image = tf.image.resize_image_with_crop_or_pad( + image, image_size, image_size) + image = tf.image.per_image_standardization(image) + + example_queue = tf.FIFOQueue( + 3 * batch_size, + dtypes=[tf.float32, tf.int32], + shapes=[[image_size, image_size, depth], [1]]) + num_threads = 1 + + example_enqueue_op = example_queue.enqueue([image, label]) + tf.train.add_queue_runner(tf.train.queue_runner.QueueRunner( + example_queue, [example_enqueue_op] * num_threads)) + + # Read 'batch' labels + images from the example queue. + images, labels = example_queue.dequeue_many(batch_size) + labels = tf.reshape(labels, [batch_size, 1]) + indices = tf.reshape(tf.range(0, batch_size, 1), [batch_size, 1]) + labels = tf.sparse_to_dense( + tf.concat(values=[indices, labels], axis=1), + [batch_size, num_classes], 1.0, 0.0) + + if data_format == 'NCHW': + images = tf.transpose(images, [0, 3, 1, 2]) + + assert len(images.get_shape()) == 4 + assert images.get_shape()[0] == batch_size + if data_format == 'NCHW': + assert images.get_shape()[1] == 3 + else: + assert images.get_shape()[-1] == 3 + assert len(labels.get_shape()) == 2 + assert labels.get_shape()[0] == batch_size + assert labels.get_shape()[1] == num_classes + + return images, labels diff --git a/tensorflow/CIFAR10/resnet/resnet_main.py b/tensorflow/CIFAR10/resnet/resnet_main.py new file mode 100644 index 0000000..847eeda --- /dev/null +++ b/tensorflow/CIFAR10/resnet/resnet_main.py @@ -0,0 +1,302 @@ +# Copyright 2016 The TensorFlow Authors. All Rights Reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# ============================================================================== + +"""ResNet Train/Eval module. +""" +import os +import six +import subprocess +import sys +import time + +import cifar_input +import numpy as np +import resnet_model +import tensorflow as tf + +FLAGS = tf.app.flags.FLAGS +tf.app.flags.DEFINE_string('dataset', 'cifar10', 'cifar10 or cifar100.') +tf.app.flags.DEFINE_string('mode', 'train', 'train or eval.') +tf.app.flags.DEFINE_string('model', '', 'model to train.') +tf.app.flags.DEFINE_string('data_format', 'NHWC', + """Data layout to use: NHWC (TF native) + or NCHW (cuDNN native).""") +tf.app.flags.DEFINE_string('train_data_path', '', + 'Filepattern for training data.') +tf.app.flags.DEFINE_string('eval_data_path', '', + 'Filepattern for eval data') +tf.app.flags.DEFINE_integer('image_size', 32, 'Image side length.') +tf.app.flags.DEFINE_string('train_dir', '', + 'Directory to keep training outputs.') +tf.app.flags.DEFINE_string('eval_dir', '', + 'Directory to keep eval outputs.') +tf.app.flags.DEFINE_integer('eval_batch_count', 50, + 'Number of batches to eval.') +tf.app.flags.DEFINE_bool('eval_once', False, + 'Whether evaluate the model only once.') +tf.app.flags.DEFINE_string('log_root', '', + 'Should be a parent directory of FLAGS.train_dir/eval_dir.') +tf.app.flags.DEFINE_string('checkpoint_dir', '', + 'Directory to store the checkpoints') +tf.app.flags.DEFINE_integer('num_gpus', 0, + 'Number of gpus used for training. (0 or 1)') +tf.app.flags.DEFINE_bool('use_bottleneck', False, + 'Use bottleneck module or not.') +tf.app.flags.DEFINE_bool('time_inference', False, + 'Time inference.') +tf.app.flags.DEFINE_integer('batch_size', -1, + 'Batch size to use.') + + +def train(hps): + """Training loop.""" + images, labels = cifar_input.build_input( + FLAGS.dataset, FLAGS.train_data_path, hps.batch_size, FLAGS.mode, hps.data_format) + model = resnet_model.ResNet(hps, images, labels, FLAGS.mode) + model.build_graph() + + param_stats = tf.contrib.tfprof.model_analyzer.print_model_analysis( + tf.get_default_graph(), + tfprof_options=tf.contrib.tfprof.model_analyzer. + TRAINABLE_VARS_PARAMS_STAT_OPTIONS) + sys.stdout.write('total_params: %d\n' % param_stats.total_parameters) + + tf.contrib.tfprof.model_analyzer.print_model_analysis( + tf.get_default_graph(), + tfprof_options=tf.contrib.tfprof.model_analyzer.FLOAT_OPS_OPTIONS) + + truth = tf.argmax(model.labels, axis=1) + predictions = tf.argmax(model.predictions, axis=1) + precision = tf.reduce_mean(tf.to_float(tf.equal(predictions, truth))) + + summary_hook = tf.train.SummarySaverHook( + save_steps=100, + output_dir=FLAGS.train_dir, + summary_op=tf.summary.merge([model.summaries, + tf.summary.scalar('Precision', precision)])) + + num_steps_per_epoch = 391 # TODO: Don't hardcode this. + + logging_hook = tf.train.LoggingTensorHook( + tensors={'step': model.global_step, + 'loss': model.cost, + 'precision': precision}, + every_n_iter=100) + + class _LearningRateSetterHook(tf.train.SessionRunHook): + """Sets learning_rate based on global step.""" + + def begin(self): + self._lrn_rate = 0.01 + + def before_run(self, run_context): + return tf.train.SessionRunArgs( + model.global_step, # Asks for global step value. + feed_dict={model.lrn_rate: self._lrn_rate}) # Sets learning rate + + def after_run(self, run_context, run_values): + train_step = run_values.results + if train_step < num_steps_per_epoch: + self._lrn_rate = 0.01 + elif train_step < (91 * num_steps_per_epoch): + self._lrn_rate = 0.1 + elif train_step < (136 * num_steps_per_epoch): + self._lrn_rate = 0.01 + elif train_step < (181 * num_steps_per_epoch): + self._lrn_rate = 0.001 + else: + self._lrn_rate = 0.0001 + + class _SaverHook(tf.train.SessionRunHook): + """Sets learning_rate based on global step.""" + + def begin(self): + self.saver = tf.train.Saver(max_to_keep=10000) + subprocess.call("rm -rf %s; mkdir -p %s" % (FLAGS.checkpoint_dir, + FLAGS.checkpoint_dir), shell=True) + self.f = open(os.path.join(FLAGS.checkpoint_dir, "times.log"), 'w') + + def after_create_session(self, sess, coord): + self.sess = sess + self.start_time = time.time() + + def before_run(self, run_context): + return tf.train.SessionRunArgs( + model.global_step # Asks for global step value. + ) + + def after_run(self, run_context, run_values): + train_step = run_values.results + epoch = train_step / num_steps_per_epoch + if train_step % num_steps_per_epoch == 0: + end_time = time.time() + directory = os.path.join(FLAGS.checkpoint_dir, ("%5d" % epoch).replace(' ', '0')) + subprocess.call("mkdir -p %s" % directory, shell=True) + ckpt_name = 'model.ckpt' + self.saver.save(self.sess, os.path.join(directory, ckpt_name), + global_step=train_step) + self.f.write("Step: %d\tTime: %s\n" % (train_step, end_time - self.start_time)) + print("Saved checkpoint after %d epoch(s) to %s..." % (epoch, directory)) + sys.stdout.flush() + self.start_time = time.time() + + def end(self, sess): + self.f.close() + + with tf.train.MonitoredTrainingSession( + checkpoint_dir=FLAGS.log_root, + hooks=[logging_hook, _LearningRateSetterHook()], + chief_only_hooks=[summary_hook, _SaverHook()], + save_checkpoint_secs=None, + # Since we provide a SummarySaverHook, we need to disable default + # SummarySaverHook. To do that we set save_summaries_steps to 0. + save_summaries_steps=None, + save_summaries_secs=None, + config=tf.ConfigProto(allow_soft_placement=True)) as mon_sess: + for i in range(num_steps_per_epoch * 181): + mon_sess.run(model.train_op) + +def evaluate(hps): + """Eval loop.""" + images, labels = cifar_input.build_input( + FLAGS.dataset, FLAGS.eval_data_path, hps.batch_size, FLAGS.mode, hps.data_format) + model = resnet_model.ResNet(hps, images, labels, FLAGS.mode) + model.build_graph() + saver = tf.train.Saver() + summary_writer = tf.summary.FileWriter(FLAGS.eval_dir) + + sess = tf.Session(config=tf.ConfigProto(allow_soft_placement=True)) + tf.train.start_queue_runners(sess) + + best_precision = 0.0 + while True: + try: + ckpt_state = tf.train.get_checkpoint_state(FLAGS.log_root) + except tf.errors.OutOfRangeError as e: + tf.logging.error('Cannot restore checkpoint: %s', e) + continue + if not (ckpt_state and ckpt_state.model_checkpoint_path): + tf.logging.info('No model to eval yet at %s', FLAGS.log_root) + break + tf.logging.info('Loading checkpoint %s', ckpt_state.model_checkpoint_path) + saver.restore(sess, ckpt_state.model_checkpoint_path) + + global_step = ckpt_state.model_checkpoint_path.split('/')[-1].split('-')[-1] + if not global_step.isdigit(): + global_step = 0 + else: + global_step = int(global_step) + + total_prediction, correct_prediction, correct_prediction_top5 = 0, 0, 0 + start_time = time.time() + for _ in six.moves.range(FLAGS.eval_batch_count): + (summaries, loss, predictions, truth, train_step) = sess.run( + [model.summaries, model.cost, model.predictions, + model.labels, model.global_step]) + + if not FLAGS.time_inference: + for (indiv_truth, indiv_prediction) in zip(truth, predictions): + indiv_truth = np.argmax(indiv_truth) + top5_prediction = np.argsort(indiv_prediction)[-5:] + top1_prediction = np.argsort(indiv_prediction)[-1] + correct_prediction += (indiv_truth == top1_prediction) + if indiv_truth in top5_prediction: + correct_prediction_top5 += 1 + total_prediction += 1 + + if FLAGS.time_inference: + print("Time for inference: %.4f" % (time.time() - start_time)) + else: + precision = 1.0 * correct_prediction / total_prediction + precision_top5 = 1.0 * correct_prediction_top5 / total_prediction + best_precision = max(precision, best_precision) + + precision_summ = tf.Summary() + precision_summ.value.add( + tag='Precision', simple_value=precision) + summary_writer.add_summary(precision_summ, train_step) + best_precision_summ = tf.Summary() + best_precision_summ.value.add( + tag='Best Precision', simple_value=best_precision) + summary_writer.add_summary(best_precision_summ, train_step) + summary_writer.add_summary(summaries, train_step) + print('Precision @ 1 = %.4f, Recall @ 5 = %.4f, Global step = %d' % + (precision, precision_top5, global_step)) + summary_writer.flush() + + if FLAGS.eval_once: + break + + time.sleep(60) + + +def main(_): + if FLAGS.model == '': + raise Exception('--model must be specified.') + + if FLAGS.num_gpus == 0: + dev = '/cpu:0' + elif FLAGS.num_gpus == 1: + dev = '/gpu:0' + else: + raise ValueError('Only support 0 or 1 gpu.') + + if FLAGS.batch_size == -1: + if FLAGS.mode == 'train': + batch_size = 128 + elif FLAGS.mode == 'eval': + batch_size = 100 + else: + batch_size = FLAGS.batch_size + + if FLAGS.dataset == 'cifar10': + num_classes = 10 + elif FLAGS.dataset == 'cifar100': + num_classes = 100 + + if FLAGS.model == 'resnet20': + num_residual_units = 3 + elif FLAGS.model == 'resnet56': + num_residual_units = 9 + elif FLAGS.model == 'resnet164' and FLAGS.use_bottleneck: + num_residual_units = 18 + elif FLAGS.model == 'resnet164' and not FLAGS.use_bottleneck: + num_residual_units = 27 + else: + raise Exception("Invalid model -- only resnet20, resnet56 and resnet164 supported") + + data_format = FLAGS.data_format + + hps = resnet_model.HParams(batch_size=batch_size, + num_classes=num_classes, + min_lrn_rate=0.0001, + lrn_rate=0.1, + num_residual_units=num_residual_units, + use_bottleneck=FLAGS.use_bottleneck, + weight_decay_rate=0.0005, + relu_leakiness=0.1, + optimizer='mom', + data_format=data_format) + + with tf.device(dev): + if FLAGS.mode == 'train': + train(hps) + elif FLAGS.mode == 'eval': + evaluate(hps) + + +if __name__ == '__main__': + tf.logging.set_verbosity(tf.logging.INFO) + tf.app.run() diff --git a/tensorflow/CIFAR10/resnet/resnet_model.py b/tensorflow/CIFAR10/resnet/resnet_model.py new file mode 100644 index 0000000..54c1447 --- /dev/null +++ b/tensorflow/CIFAR10/resnet/resnet_model.py @@ -0,0 +1,281 @@ +# Copyright 2016 The TensorFlow Authors. All Rights Reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# ============================================================================== + +"""ResNet model. + +Related papers: +https://arxiv.org/pdf/1603.05027v2.pdf +https://arxiv.org/pdf/1512.03385v1.pdf +https://arxiv.org/pdf/1605.07146v1.pdf +""" +from collections import namedtuple + +import numpy as np +import tensorflow as tf +import six + +from tensorflow.python.training import moving_averages + + +HParams = namedtuple('HParams', + 'batch_size, num_classes, min_lrn_rate, lrn_rate, ' + 'num_residual_units, use_bottleneck, weight_decay_rate, ' + 'relu_leakiness, optimizer, data_format') + + +class ResNet(object): + """ResNet model.""" + + def __init__(self, hps, images, labels, mode): + """ResNet constructor. + + Args: + hps: Hyperparameters. + images: Batches of images. [batch_size, image_size, image_size, 3] + labels: Batches of labels. [batch_size, num_classes] + mode: One of 'train' and 'eval'. + """ + self.hps = hps + self._images = images + self.labels = labels + self.mode = mode + + self._extra_train_ops = [] + + def build_graph(self): + """Build a whole graph for the model.""" + self.global_step = tf.contrib.framework.get_or_create_global_step() + self._build_model() + if self.mode == 'train': + self._build_train_op() + self.summaries = tf.summary.merge_all() + + def _stride_arr(self, stride): + """Map a stride scalar to the stride array for tf.nn.conv2d.""" + if self.hps.data_format == 'NHWC': + return [1, stride, stride, 1] + elif self.hps.data_format == 'NCHW': + return [1, 1, stride, stride] + else: + raise Exception("Invalid data_format") + + def _build_model(self): + """Build the core model within the graph.""" + with tf.variable_scope('init'): + x = self._images + x = self._conv('init_conv', x, 3, 3, 16, self._stride_arr(1)) + + strides = [1, 2, 2] + activate_before_residual = [True, False, False] + if self.hps.use_bottleneck: + res_func = self._bottleneck_residual + filters = [16, 64, 128, 256] + else: + res_func = self._residual + filters = [16, 16, 32, 64] + # Uncomment the following codes to use w28-10 wide residual network. + # It is more memory efficient than very deep residual network and has + # comparably good performance. + # https://arxiv.org/pdf/1605.07146v1.pdf + # filters = [16, 160, 320, 640] + # Update hps.num_residual_units to 4 + + with tf.variable_scope('unit_1_0'): + x = res_func(x, filters[0], filters[1], self._stride_arr(strides[0]), + activate_before_residual[0]) + for i in six.moves.range(1, self.hps.num_residual_units): + with tf.variable_scope('unit_1_%d' % i): + x = res_func(x, filters[1], filters[1], self._stride_arr(1), False) + + with tf.variable_scope('unit_2_0'): + x = res_func(x, filters[1], filters[2], self._stride_arr(strides[1]), + activate_before_residual[1]) + for i in six.moves.range(1, self.hps.num_residual_units): + with tf.variable_scope('unit_2_%d' % i): + x = res_func(x, filters[2], filters[2], self._stride_arr(1), False) + + with tf.variable_scope('unit_3_0'): + x = res_func(x, filters[2], filters[3], self._stride_arr(strides[2]), + activate_before_residual[2]) + for i in six.moves.range(1, self.hps.num_residual_units): + with tf.variable_scope('unit_3_%d' % i): + x = res_func(x, filters[3], filters[3], self._stride_arr(1), False) + + with tf.variable_scope('unit_last'): + x = self._batch_norm('final_bn', x) + x = self._relu(x, self.hps.relu_leakiness) + x = self._global_avg_pool(x) + + with tf.variable_scope('logit'): + logits = self._fully_connected(x, self.hps.num_classes) + self.predictions = tf.nn.softmax(logits) + + with tf.variable_scope('costs'): + xent = tf.nn.softmax_cross_entropy_with_logits( + logits=logits, labels=self.labels) + self.cost = tf.reduce_mean(xent, name='xent') + self.cost += self._decay() + + tf.summary.scalar('cost', self.cost) + + def _build_train_op(self): + """Build training specific ops for the graph.""" + self.lrn_rate = tf.constant(self.hps.lrn_rate, tf.float32) + tf.summary.scalar('learning_rate', self.lrn_rate) + + trainable_variables = tf.trainable_variables() + grads = tf.gradients(self.cost, trainable_variables) + + if self.hps.optimizer == 'sgd': + optimizer = tf.train.GradientDescentOptimizer(self.lrn_rate) + elif self.hps.optimizer == 'mom': + optimizer = tf.train.MomentumOptimizer(self.lrn_rate, 0.9) + + apply_op = optimizer.apply_gradients( + zip(grads, trainable_variables), + global_step=self.global_step, name='train_step') + + train_ops = [apply_op] + self._extra_train_ops + self.train_op = tf.group(*train_ops) + + # TODO(xpan): Consider batch_norm in contrib/layers/python/layers/layers.py + def _batch_norm(self, name, x): + """Batch normalization.""" + with tf.variable_scope(name) as scope: + output = tf.contrib.layers.batch_norm(x, + decay=0.9, + epsilon=0.001, + data_format=self.hps.data_format, + scope=scope, + is_training=(self.mode == 'train'), + fused=True, + updates_collections=None) + return output + + def _residual(self, x, in_filter, out_filter, stride, + activate_before_residual=False): + """Residual unit with 2 sub layers.""" + if activate_before_residual: + with tf.variable_scope('shared_activation'): + x = self._batch_norm('init_bn', x) + x = self._relu(x, self.hps.relu_leakiness) + orig_x = x + else: + with tf.variable_scope('residual_only_activation'): + orig_x = x + x = self._batch_norm('init_bn', x) + x = self._relu(x, self.hps.relu_leakiness) + + with tf.variable_scope('sub1'): + x = self._conv('conv1', x, 3, in_filter, out_filter, stride) + + with tf.variable_scope('sub2'): + x = self._batch_norm('bn2', x) + x = self._relu(x, self.hps.relu_leakiness) + x = self._conv('conv2', x, 3, out_filter, out_filter, [1, 1, 1, 1]) + + with tf.variable_scope('sub_add'): + if in_filter != out_filter: + orig_x = tf.nn.avg_pool(orig_x, stride, stride, 'VALID', + data_format=self.hps.data_format) + if self.hps.data_format == 'NHWC': + orig_x = tf.pad( + orig_x, [[0, 0], [0, 0], [0, 0], + [(out_filter-in_filter)//2, (out_filter-in_filter)//2]]) + elif self.hps.data_format == 'NCHW': + orig_x = tf.pad( + orig_x, [[0, 0], [(out_filter-in_filter)//2, (out_filter-in_filter)//2], + [0, 0], [0, 0]]) + x += orig_x + + tf.logging.debug('image after unit %s', x.get_shape()) + return x + + def _bottleneck_residual(self, x, in_filter, out_filter, stride, + activate_before_residual=False): + """Bottleneck residual unit with 3 sub layers.""" + if activate_before_residual: + with tf.variable_scope('common_bn_relu'): + x = self._batch_norm('init_bn', x) + x = self._relu(x, self.hps.relu_leakiness) + orig_x = x + else: + with tf.variable_scope('residual_bn_relu'): + orig_x = x + x = self._batch_norm('init_bn', x) + x = self._relu(x, self.hps.relu_leakiness) + + with tf.variable_scope('sub1'): + x = self._conv('conv1', x, 1, in_filter, out_filter/4, stride) + + with tf.variable_scope('sub2'): + x = self._batch_norm('bn2', x) + x = self._relu(x, self.hps.relu_leakiness) + x = self._conv('conv2', x, 3, out_filter/4, out_filter/4, [1, 1, 1, 1]) + + with tf.variable_scope('sub3'): + x = self._batch_norm('bn3', x) + x = self._relu(x, self.hps.relu_leakiness) + x = self._conv('conv3', x, 1, out_filter/4, out_filter, [1, 1, 1, 1]) + + with tf.variable_scope('sub_add'): + if in_filter != out_filter: + orig_x = self._conv('project', orig_x, 1, in_filter, out_filter, stride) + x += orig_x + + tf.logging.info('image after unit %s', x.get_shape()) + return x + + def _decay(self): + """L2 weight decay loss.""" + costs = [] + for var in tf.trainable_variables(): + if var.op.name.find(r'DW') > 0: + costs.append(tf.nn.l2_loss(var)) + # tf.summary.histogram(var.op.name, var) + + return tf.multiply(self.hps.weight_decay_rate, tf.add_n(costs)) + + def _conv(self, name, x, filter_size, in_filters, out_filters, strides): + """Convolution.""" + with tf.variable_scope(name): + n = filter_size * filter_size * out_filters + kernel = tf.get_variable( + 'DW', [filter_size, filter_size, in_filters, out_filters], + tf.float32, initializer=tf.random_normal_initializer( + stddev=np.sqrt(2.0/n))) + return tf.nn.conv2d(x, kernel, strides, padding='SAME', + data_format=self.hps.data_format) + + def _relu(self, x, leakiness=0.0): + """Relu, with optional leaky support.""" + return tf.where(tf.less(x, 0.0), leakiness * x, x, name='leaky_relu') + + def _fully_connected(self, x, out_dim): + """FullyConnected layer for final output.""" + x = tf.reshape(x, [self.hps.batch_size, -1]) + w = tf.get_variable( + 'DW', [x.get_shape()[1], out_dim], + initializer=tf.uniform_unit_scaling_initializer(factor=1.0)) + b = tf.get_variable('biases', [out_dim], + initializer=tf.constant_initializer()) + return tf.nn.xw_plus_b(x, w, b) + + def _global_avg_pool(self, x): + assert x.get_shape().ndims == 4 + if self.hps.data_format == 'NHWC': + return tf.reduce_mean(x, [1, 2]) + elif self.hps.data_format == 'NCHW': + return tf.reduce_mean(x, [2, 3]) diff --git a/tensorflow/CIFAR10/time_inference.py b/tensorflow/CIFAR10/time_inference.py new file mode 100644 index 0000000..e53173d --- /dev/null +++ b/tensorflow/CIFAR10/time_inference.py @@ -0,0 +1,51 @@ +import argparse +import os +import subprocess +import sys + +def main(checkpoint_path, model, use_bottleneck): + print("Number of images\tInference time") + num_trials = 10 + for batch_size in [1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, 4096, 8192]: + command = ("python3 resnet/resnet_main.py --mode=eval --eval_data_path=cifar10/test_batch.bin " + "--eval_dir=data/%(model)s/log_root/eval --dataset='cifar10' --model=%(model)s " + "--use_bottleneck=%(use_bottleneck)s --eval_batch_count=%(num_trials)d --eval_once=True --num_gpus=1 " + "--data_format=NHWC --time_inference=True --eval_batch_count=1 --batch_size=%(batch_size)d" % + {"model": model, "use_bottleneck": "True" if use_bottleneck else "False", "batch_size": batch_size, + "num_trials": num_trials}) + full_command = command + " --log_root=%s 2>/dev/null" % checkpoint_path + try: + output = subprocess.check_output(full_command, shell=True) + output = output.decode('utf8').strip() + for line in output.split('\n'): + if "Time for inference" in line: + line = line.strip() + inference_time = float(line.split(": ")[1]) / num_trials + stats = [batch_size, inference_time] + print("\t".join([str(stat) for stat in stats])) + sys.stdout.flush() + except: + stats = [batch_size, ""] + print("\t".join([str(stat) for stat in stats])) + sys.stdout.flush() + + +if __name__ == '__main__': + parser = argparse.ArgumentParser( + description=("Backup model checkpoints periodically") + ) + parser.add_argument('-i', "--checkpoint_path", type=str, required=True, + help="Path to dumped model checkpoints") + parser.add_argument('-m', "--model", type=str, required=True, + help="Model name") + parser.add_argument('-b', "--use_bottleneck", type=bool, default=False, + help="Use bottleneck") + + cmdline_args = parser.parse_args() + opt_dict = vars(cmdline_args) + + checkpoint_path = opt_dict["checkpoint_path"] + model = opt_dict["model"] + use_bottleneck = opt_dict["use_bottleneck"] + + main(checkpoint_path, model, use_bottleneck) diff --git a/tensorflow/SQuAD/.gitignore b/tensorflow/SQuAD/.gitignore new file mode 100644 index 0000000..376bc75 --- /dev/null +++ b/tensorflow/SQuAD/.gitignore @@ -0,0 +1,3 @@ +out/ +data/ +*/__pycache__/ diff --git a/tensorflow/SQuAD/README.md b/tensorflow/SQuAD/README.md new file mode 100644 index 0000000..62f9b75 --- /dev/null +++ b/tensorflow/SQuAD/README.md @@ -0,0 +1,165 @@ +# Bi-directional Attention Flow for Machine Comprehension + +- This the original implementation of [Bi-directional Attention Flow for Machine Comprehension][paper] (Seo et al., 2016). +- This is tensorflow v1.1.0 comaptible version. This is not compatible with previous trained models, +so if you want to use them, go to [v0.2.1][v0.2.1]. +- The CodaLab worksheet for the [SQuAD Leaderboard][squad] submission is available [here][worksheet]. +- Please contact [Minjoon Seo][minjoon] ([@seominjoon][minjoon-github]) for questions and suggestions. + +## 0. Requirements +#### General +- Python (developed on 3.5.2. Issues have been reported with Python 2!) +- unzip + +#### Python Packages +- tensorflow (deep learning library, verified on 1.1.0) +- nltk (NLP tools, verified on 3.2.1) +- tqdm (progress bar, verified on 4.7.4) +- jinja2 (for visaulization; if you only train and test, not needed) + +## 1. Pre-processing +First, prepare data. Donwload SQuAD data and GloVe and nltk corpus +(~850 MB, this will download files to `$HOME/data`): +``` +chmod +x download.sh; ./download.sh +``` + +Second, Preprocess Stanford QA dataset (along with GloVe vectors) and save them in `$PWD/data/squad` (~5 minutes): +``` +python -m squad.prepro +``` + +## 2. Training +The model was trained with NVidia Titan X (Pascal Architecture, 2016). +The model requires at least 12GB of GPU RAM. +If your GPU RAM is smaller than 12GB, you can either decrease batch size (performance might degrade), +or you can use multi GPU (see below). +The training converges at ~18k steps, and it took ~4s per step (i.e. ~20 hours). + +Before training, it is recommended to first try the following code to verify everything is okay and memory is sufficient: +``` +python -m basic.cli --mode train --noload --debug +``` + +Then to fully train, run: +``` +python -m basic.cli --mode train --noload +``` + +You can speed up the training process with optimization flags: +``` +python -m basic.cli --mode train --noload --len_opt --cluster +``` +You can still omit them, but training will be much slower. + + +## 3. Test +To test, run: +``` +python -m basic.cli +``` + +Similarly to training, you can give the optimization flags to speed up test (5 minutes on dev data): +``` +python -m basic.cli --len_opt --cluster +``` + +This command loads the most recently saved model during training and begins testing on the test data. +After the process ends, it prints F1 and EM scores, and also outputs a json file (`$PWD/out/basic/00/answer/test-####.json`, +where `####` is the step # that the model was saved). +Note that the printed scores are not official (our scoring scheme is a bit harsher). +To obtain the official number, use the official evaluator (copied in `squad` folder) and the output json file: + +``` +python squad/evaluate-v1.1.py $HOME/data/squad/dev-v1.1.json out/basic/00/answer/test-####.json +``` + +### 3.1 Loading from pre-trained weights +NOTE: this version is not compatible with the following trained models. +For compatibility, use [v0.2.1][v0.2.1]. + +Instead of training the model yourself, you can choose to use pre-trained weights that were used for [SQuAD Leaderboard][squad] submission. +Refer to [this worksheet][worksheet] in CodaLab to reproduce the results. +If you are unfamiliar with CodaLab, follow these simple steps (given that you met all prereqs above): + +1. Download `save.zip` from the [worksheet][worksheet] and unzip it in the current directory. +2. Copy `glove.6B.100d.txt` from your glove data folder (`$HOME/data/glove/`) to the current directory. +3. To reproduce single model: + + ``` + basic/run_single.sh $HOME/data/squad/dev-v1.1.json single.json + ``` + + This writes the answers to `single.json` in the current directory. You can then use the official evaluator to obtain EM and F1 scores. If you want to run on GPU (~5 mins), change the value of batch_size flag in the shell file to a higher number (60 for 12GB GPU RAM). +4. Similarly, to reproduce ensemble method: + + ``` + basic/run_ensemble.sh $HOME/data/squad/dev-v1.1.json ensemble.json + ``` + If you want to run on GPU, you should run the script sequentially by removing '&' in the forloop, or you will need to specify different GPUs for each run of the for loop. + +## Results + +### Dev Data + +| | EM (%) | F1 (%) | +| -------- |:------:|:------:| +| single | 67.8 | 77.4 | + +###Dev Data (old) +NOTE: These numbers are from [v0.2.1][v0.2.1]. + +| | EM (%) | F1 (%) | +| -------- |:------:|:------:| +| single | 67.7 | 77.3 | +| ensemble | 72.6 | 80.7 | + + +###Test Data (old) +NOTE: These numbers are from [v0.2.1][v0.2.1]. + +| | EM (%) | F1 (%) | +| -------- |:------:|:------:| +| single | 68.0 | 77.3 | +| ensemble | 73.3 | 81.1 | + +Refer to [our paper][paper] for more details. +See [SQuAD Leaderboard][squad] to compare with other models. + + + + + +## Multi-GPU Training & Testing +Our model supports multi-GPU training. +We follow the parallelization paradigm described in [TensorFlow Tutorial][multi-gpu]. +In short, if you want to use batch size of 60 (default) but if you have 3 GPUs with 4GB of RAM, +then you initialize each GPU with batch size of 20, and combine the gradients on CPU. +This can be easily done by running: +``` +python -m basic.cli --mode train --noload --num_gpus 3 --batch_size 20 +``` + +Similarly, you can speed up your testing by: +``` +python -m basic.cli --num_gpus 3 --batch_size 20 +``` + + +[multi-gpu]: https://www.tensorflow.org/versions/r0.11/tutorials/deep_cnn/index.html#training-a-model-using-multiple-gpu-cards +[squad]: http://stanford-qa.com +[paper]: https://arxiv.org/abs/1611.01603 +[worksheet]: https://worksheets.codalab.org/worksheets/0x37a9b8c44f6845c28866267ef941c89d/ +[minjoon]: https://seominjoon.github.io +[minjoon-github]: https://github.com/seominjoon +[v0.2.1]: https://github.com/allenai/bi-att-flow/tree/v0.2.1 diff --git a/tensorflow/SQuAD/basic/__init__.py b/tensorflow/SQuAD/basic/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/tensorflow/SQuAD/basic/cli.py b/tensorflow/SQuAD/basic/cli.py new file mode 100644 index 0000000..5365612 --- /dev/null +++ b/tensorflow/SQuAD/basic/cli.py @@ -0,0 +1,112 @@ +import os + +import tensorflow as tf + +from basic.main import main as m + +flags = tf.app.flags + +# Names and directories +flags.DEFINE_string("model_name", "basic", "Model name [basic]") +flags.DEFINE_string("data_dir", "data/squad", "Data dir [data/squad]") +flags.DEFINE_string("run_id", "0", "Run ID [0]") +flags.DEFINE_string("out_base_dir", "out", "out base dir [out]") +flags.DEFINE_string("forward_name", "single", "Forward name [single]") +flags.DEFINE_string("answer_path", "", "Answer path []") +flags.DEFINE_string("eval_path", "", "Eval path []") +flags.DEFINE_string("load_path", "", "Load path []") +flags.DEFINE_string("shared_path", "", "Shared path []") + +# Device placement +flags.DEFINE_string("device", "/cpu:0", "default device for summing gradients. [/cpu:0]") +flags.DEFINE_string("device_type", "gpu", "device for computing gradients (parallelization). cpu | gpu [gpu]") +flags.DEFINE_integer("num_gpus", 1, "num of gpus or cpus for computing gradients [1]") + +# Essential training and test options +flags.DEFINE_string("mode", "test", "trains | test | forward [test]") +flags.DEFINE_boolean("load", True, "load saved data? [True]") +flags.DEFINE_bool("single", False, "supervise only the answer sentence? [False]") +flags.DEFINE_boolean("debug", False, "Debugging mode? [False]") +flags.DEFINE_bool('load_ema', True, "load exponential average of variables when testing? [True]") +flags.DEFINE_bool("eval", True, "eval? [True]") +flags.DEFINE_bool("wy", False, "Use wy for loss / eval? [False]") +flags.DEFINE_bool("na", False, "Enable no answer strategy and learn bias? [False]") +flags.DEFINE_float("th", 0.5, "Threshold [0.5]") + +# Training / test parameters +flags.DEFINE_integer("batch_size", 60, "Batch size [60]") +flags.DEFINE_integer("val_num_batches", 100, "validation num batches [100]") +flags.DEFINE_integer("test_num_batches", 0, "test num batches [0]") +flags.DEFINE_integer("num_epochs", 12, "Total number of epochs for training [12]") +flags.DEFINE_integer("num_steps", 20000, "Number of steps [20000]") +flags.DEFINE_integer("load_step", 0, "load step [0]") +flags.DEFINE_float("init_lr", 0.001, "Initial learning rate [0.001]") +flags.DEFINE_float("input_keep_prob", 0.8, "Input keep prob for the dropout of LSTM weights [0.8]") +flags.DEFINE_float("keep_prob", 0.8, "Keep prob for the dropout of Char-CNN weights [0.8]") +flags.DEFINE_float("wd", 0.0, "L2 weight decay for regularization [0.0]") +flags.DEFINE_integer("hidden_size", 100, "Hidden size [100]") +flags.DEFINE_integer("char_out_size", 100, "char-level word embedding size [100]") +flags.DEFINE_integer("char_emb_size", 8, "Char emb size [8]") +flags.DEFINE_string("out_channel_dims", "100", "Out channel dims of Char-CNN, separated by commas [100]") +flags.DEFINE_string("filter_heights", "5", "Filter heights of Char-CNN, separated by commas [5]") +flags.DEFINE_bool("finetune", False, "Finetune word embeddings? [False]") +flags.DEFINE_bool("highway", True, "Use highway? [True]") +flags.DEFINE_integer("highway_num_layers", 2, "highway num layers [2]") +flags.DEFINE_bool("share_cnn_weights", True, "Share Char-CNN weights [True]") +flags.DEFINE_bool("share_lstm_weights", True, "Share pre-processing (phrase-level) LSTM weights [True]") +flags.DEFINE_float("var_decay", 0.999, "Exponential moving average decay for variables [0.999]") + +# Optimizations +flags.DEFINE_bool("cluster", False, "Cluster data for faster training [False]") +flags.DEFINE_bool("len_opt", False, "Length optimization? [False]") +flags.DEFINE_bool("cpu_opt", False, "CPU optimization? GPU computation can be slower [False]") + +# Logging and saving options +flags.DEFINE_boolean("progress", True, "Show progress? [True]") +flags.DEFINE_integer("log_period", 100, "Log period [100]") +flags.DEFINE_integer("eval_period", 1000, "Eval period [1000]") +flags.DEFINE_integer("save_period", 1000, "Save Period [1000]") +flags.DEFINE_integer("max_to_keep", 20, "Max recent saves to keep [20]") +flags.DEFINE_bool("dump_eval", True, "dump eval? [True]") +flags.DEFINE_bool("dump_answer", True, "dump answer? [True]") +flags.DEFINE_bool("vis", False, "output visualization numbers? [False]") +flags.DEFINE_bool("dump_pickle", True, "Dump pickle instead of json? [True]") +flags.DEFINE_float("decay", 0.9, "Exponential moving average decay for logging values [0.9]") + +# Thresholds for speed and less memory usage +flags.DEFINE_integer("word_count_th", 10, "word count th [100]") +flags.DEFINE_integer("char_count_th", 50, "char count th [500]") +flags.DEFINE_integer("sent_size_th", 400, "sent size th [64]") +flags.DEFINE_integer("num_sents_th", 8, "num sents th [8]") +flags.DEFINE_integer("ques_size_th", 30, "ques size th [32]") +flags.DEFINE_integer("word_size_th", 16, "word size th [16]") +flags.DEFINE_integer("para_size_th", 256, "para size th [256]") + +# Advanced training options +flags.DEFINE_bool("lower_word", True, "lower word [True]") +flags.DEFINE_bool("squash", False, "squash the sentences into one? [False]") +flags.DEFINE_bool("swap_memory", True, "swap memory? [True]") +flags.DEFINE_string("data_filter", "max", "max | valid | semi [max]") +flags.DEFINE_bool("use_glove_for_unk", True, "use glove for unk [False]") +flags.DEFINE_bool("known_if_glove", True, "consider as known if present in glove [False]") +flags.DEFINE_string("logit_func", "tri_linear", "logit func [tri_linear]") +flags.DEFINE_string("answer_func", "linear", "answer logit func [linear]") +flags.DEFINE_string("sh_logit_func", "tri_linear", "sh logit func [tri_linear]") + +# Ablation options +flags.DEFINE_bool("use_char_emb", True, "use char emb? [True]") +flags.DEFINE_bool("use_word_emb", True, "use word embedding? [True]") +flags.DEFINE_bool("q2c_att", True, "question-to-context attention? [True]") +flags.DEFINE_bool("c2q_att", True, "context-to-question attention? [True]") +flags.DEFINE_bool("dynamic_att", False, "Dynamic attention [False]") + + +def main(_): + config = flags.FLAGS + + config.out_dir = os.path.join(config.out_base_dir, config.model_name, str(config.run_id).zfill(2)) + + m(config) + +if __name__ == "__main__": + tf.app.run() diff --git a/tensorflow/SQuAD/basic/ensemble.py b/tensorflow/SQuAD/basic/ensemble.py new file mode 100644 index 0000000..cbcbd4c --- /dev/null +++ b/tensorflow/SQuAD/basic/ensemble.py @@ -0,0 +1,116 @@ +import argparse +import functools +import gzip +import json +import pickle +from collections import defaultdict +from operator import mul + +from tqdm import tqdm +from squad.utils import get_phrase, get_best_span, get_span_score_pairs + + +def get_args(): + parser = argparse.ArgumentParser() + parser.add_argument('paths', nargs='+') + parser.add_argument('-o', '--out', default='ensemble.json') + parser.add_argument("--data_path", default="data/squad/data_test.json") + parser.add_argument("--shared_path", default="data/squad/shared_test.json") + args = parser.parse_args() + return args + + +def ensemble(args): + e_list = [] + for path in tqdm(args.paths): + with gzip.open(path, 'r') as fh: + e = pickle.load(fh) + e_list.append(e) + + with open(args.data_path, 'r') as fh: + data = json.load(fh) + + with open(args.shared_path, 'r') as fh: + shared = json.load(fh) + + out = {} + for idx, (id_, rx) in tqdm(enumerate(zip(data['ids'], data['*x'])), total=len(e['yp'])): + if idx >= len(e['yp']): + # for debugging purpose + break + context = shared['p'][rx[0]][rx[1]] + wordss = shared['x'][rx[0]][rx[1]] + yp_list = [e['yp'][idx] for e in e_list] + yp2_list = [e['yp2'][idx] for e in e_list] + answer = ensemble4(context, wordss, yp_list, yp2_list) + out[id_] = answer + + with open(args.out, 'w') as fh: + json.dump(out, fh) + + +def ensemble1(context, wordss, y1_list, y2_list): + """ + + :param context: Original context + :param wordss: tokenized words (nested 2D list) + :param y1_list: list of start index probs (each element corresponds to probs form single model) + :param y2_list: list of stop index probs + :return: + """ + sum_y1 = combine_y_list(y1_list) + sum_y2 = combine_y_list(y2_list) + span, score = get_best_span(sum_y1, sum_y2) + return get_phrase(context, wordss, span) + + +def ensemble2(context, wordss, y1_list, y2_list): + start_dict = defaultdict(float) + stop_dict = defaultdict(float) + for y1, y2 in zip(y1_list, y2_list): + span, score = get_best_span(y1, y2) + start_dict[span[0]] += y1[span[0][0]][span[0][1]] + stop_dict[span[1]] += y2[span[1][0]][span[1][1]] + start = max(start_dict.items(), key=lambda pair: pair[1])[0] + stop = max(stop_dict.items(), key=lambda pair: pair[1])[0] + best_span = (start, stop) + return get_phrase(context, wordss, best_span) + + +def ensemble3(context, wordss, y1_list, y2_list): + d = defaultdict(float) + for y1, y2 in zip(y1_list, y2_list): + span, score = get_best_span(y1, y2) + phrase = get_phrase(context, wordss, span) + d[phrase] += score + return max(d.items(), key=lambda pair: pair[1])[0] + + +def ensemble4(context, wordss, y1_list, y2_list): + d = defaultdict(lambda: 0.0) + for y1, y2 in zip(y1_list, y2_list): + for span, score in get_span_score_pairs(y1, y2): + d[span] += score + span = max(d.items(), key=lambda pair: pair[1])[0] + phrase = get_phrase(context, wordss, span) + return phrase + + +def combine_y_list(y_list, op='*'): + if op == '+': + func = sum + elif op == '*': + def func(l): return functools.reduce(mul, l) + else: + func = op + return [[func(yij_list) for yij_list in zip(*yi_list)] for yi_list in zip(*y_list)] + + +def main(): + args = get_args() + ensemble(args) + +if __name__ == "__main__": + main() + + diff --git a/tensorflow/SQuAD/basic/ensemble_fast.py b/tensorflow/SQuAD/basic/ensemble_fast.py new file mode 100644 index 0000000..f9a654c --- /dev/null +++ b/tensorflow/SQuAD/basic/ensemble_fast.py @@ -0,0 +1,39 @@ +import sys +import json +from collections import Counter, defaultdict +import re + +def key_func(pair): + return pair[1] + + +def get_func(vals, probs): + counter = Counter(vals) + # return max(zip(vals, probs), key=lambda pair: pair[1])[0] + # return max(zip(vals, probs), key=lambda pair: pair[1] * counter[pair[0]] / len(counter) - 999 * (len(pair[0]) == 0) )[0] + # return max(zip(vals, probs), key=lambda pair: pair[1] + 0.7 * counter[pair[0]] / len(counter) - 999 * (len(pair[0]) == 0) )[0] + d = defaultdict(float) + for val, prob in zip(vals, probs): + d[val] += prob + d[''] = 0 + return max(d.items(), key=lambda pair: pair[1])[0] + +third_path = sys.argv[1] +other_paths = sys.argv[2:] + +others = [json.load(open(path, 'r')) for path in other_paths] + + +c = {} + +assert min(map(len, others)) == max(map(len, others)), list(map(len, others)) + +for key in others[0].keys(): + if key == 'scores': + continue + probs = [other['scores'][key] for other in others] + vals = [other[key] for other in others] + largest_val = get_func(vals, probs) + c[key] = largest_val + +json.dump(c, open(third_path, 'w')) \ No newline at end of file diff --git a/tensorflow/SQuAD/basic/evaluator.py b/tensorflow/SQuAD/basic/evaluator.py new file mode 100644 index 0000000..b7b320f --- /dev/null +++ b/tensorflow/SQuAD/basic/evaluator.py @@ -0,0 +1,453 @@ +import numpy as np +import tensorflow as tf + +from basic.read_data import DataSet +from my.nltk_utils import span_f1 +from my.tensorflow import padded_reshape +from my.utils import argmax +from squad.utils import get_phrase, get_best_span, get_best_span_wy + + +class Evaluation(object): + def __init__(self, data_type, global_step, idxs, yp, tensor_dict=None): + self.data_type = data_type + self.global_step = global_step + self.idxs = idxs + self.yp = yp + self.num_examples = len(yp) + self.tensor_dict = None + self.dict = {'data_type': data_type, + 'global_step': global_step, + 'yp': yp, + 'idxs': idxs, + 'num_examples': self.num_examples} + if tensor_dict is not None: + self.tensor_dict = {key: val.tolist() for key, val in tensor_dict.items()} + for key, val in self.tensor_dict.items(): + self.dict[key] = val + self.summaries = None + + def __repr__(self): + return "{} step {}".format(self.data_type, self.global_step) + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_yp = self.yp + other.yp + new_idxs = self.idxs + other.idxs + new_tensor_dict = None + if self.tensor_dict is not None: + new_tensor_dict = {key: val + other.tensor_dict[key] for key, val in self.tensor_dict.items()} + return Evaluation(self.data_type, self.global_step, new_idxs, new_yp, tensor_dict=new_tensor_dict) + + def __radd__(self, other): + return self.__add__(other) + + +class LabeledEvaluation(Evaluation): + def __init__(self, data_type, global_step, idxs, yp, y, tensor_dict=None): + super(LabeledEvaluation, self).__init__(data_type, global_step, idxs, yp, tensor_dict=tensor_dict) + self.y = y + self.dict['y'] = y + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_yp = self.yp + other.yp + new_y = self.y + other.y + new_idxs = self.idxs + other.idxs + if self.tensor_dict is not None: + new_tensor_dict = {key: np.concatenate((val, other.tensor_dict[key]), axis=0) for key, val in self.tensor_dict.items()} + return LabeledEvaluation(self.data_type, self.global_step, new_idxs, new_yp, new_y, tensor_dict=new_tensor_dict) + + +class AccuracyEvaluation(LabeledEvaluation): + def __init__(self, data_type, global_step, idxs, yp, y, correct, loss, tensor_dict=None): + super(AccuracyEvaluation, self).__init__(data_type, global_step, idxs, yp, y, tensor_dict=tensor_dict) + self.loss = loss + self.correct = correct + self.acc = sum(correct) / len(correct) + self.dict['loss'] = loss + self.dict['correct'] = correct + self.dict['acc'] = self.acc + loss_summary = tf.Summary(value=[tf.Summary.Value(tag='{}/loss'.format(data_type), simple_value=self.loss)]) + acc_summary = tf.Summary(value=[tf.Summary.Value(tag='{}/acc'.format(data_type), simple_value=self.acc)]) + self.summaries = [loss_summary, acc_summary] + + def __repr__(self): + return "{} step {}: accuracy={}, loss={}".format(self.data_type, self.global_step, self.acc, self.loss) + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_idxs = self.idxs + other.idxs + new_yp = self.yp + other.yp + new_y = self.y + other.y + new_correct = self.correct + other.correct + new_loss = (self.loss * self.num_examples + other.loss * other.num_examples) / len(new_correct) + if self.tensor_dict is not None: + new_tensor_dict = {key: np.concatenate((val, other.tensor_dict[key]), axis=0) for key, val in self.tensor_dict.items()} + return AccuracyEvaluation(self.data_type, self.global_step, new_idxs, new_yp, new_y, new_correct, new_loss, tensor_dict=new_tensor_dict) + + +class Evaluator(object): + def __init__(self, config, model, tensor_dict=None): + self.config = config + self.model = model + self.global_step = model.global_step + self.yp = model.yp + self.tensor_dict = {} if tensor_dict is None else tensor_dict + + def get_evaluation(self, sess, batch): + idxs, data_set = batch + feed_dict = self.model.get_feed_dict(data_set, False, supervised=False) + global_step, yp, vals = sess.run([self.global_step, self.yp, list(self.tensor_dict.values())], feed_dict=feed_dict) + yp = yp[:data_set.num_examples] + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + e = Evaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), tensor_dict=tensor_dict) + return e + + def get_evaluation_from_batches(self, sess, batches): + e = sum(self.get_evaluation(sess, batch) for batch in batches) + return e + + +class LabeledEvaluator(Evaluator): + def __init__(self, config, model, tensor_dict=None): + super(LabeledEvaluator, self).__init__(config, model, tensor_dict=tensor_dict) + self.y = model.y + + def get_evaluation(self, sess, batch): + idxs, data_set = batch + feed_dict = self.model.get_feed_dict(data_set, False, supervised=False) + global_step, yp, vals = sess.run([self.global_step, self.yp, list(self.tensor_dict.values())], feed_dict=feed_dict) + yp = yp[:data_set.num_examples] + y = feed_dict[self.y] + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + e = LabeledEvaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), y.tolist(), tensor_dict=tensor_dict) + return e + + +class AccuracyEvaluator(LabeledEvaluator): + def __init__(self, config, model, tensor_dict=None): + super(AccuracyEvaluator, self).__init__(config, model, tensor_dict=tensor_dict) + self.loss = model.loss + + def get_evaluation(self, sess, batch): + idxs, data_set = batch + assert isinstance(data_set, DataSet) + feed_dict = self.model.get_feed_dict(data_set, False) + global_step, yp, loss, vals = sess.run([self.global_step, self.yp, self.loss, list(self.tensor_dict.values())], feed_dict=feed_dict) + y = data_set.data['y'] + yp = yp[:data_set.num_examples] + correct = [self.__class__.compare(yi, ypi) for yi, ypi in zip(y, yp)] + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + e = AccuracyEvaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), y, correct, float(loss), tensor_dict=tensor_dict) + return e + + @staticmethod + def compare(yi, ypi): + for start, stop in yi: + if start == int(np.argmax(ypi)): + return True + return False + + +class AccuracyEvaluator2(AccuracyEvaluator): + @staticmethod + def compare(yi, ypi): + for start, stop in yi: + para_start = int(np.argmax(np.max(ypi, 1))) + sent_start = int(np.argmax(ypi[para_start])) + if tuple(start) == (para_start, sent_start): + return True + return False + + +class ForwardEvaluation(Evaluation): + def __init__(self, data_type, global_step, idxs, yp, yp2, loss, id2answer_dict, tensor_dict=None): + super(ForwardEvaluation, self).__init__(data_type, global_step, idxs, yp, tensor_dict=tensor_dict) + self.yp2 = yp2 + self.loss = loss + self.dict['loss'] = loss + self.dict['yp2'] = yp2 + self.id2answer_dict = id2answer_dict + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_idxs = self.idxs + other.idxs + new_yp = self.yp + other.yp + new_yp2 = self.yp2 + other.yp2 + new_loss = (self.loss * self.num_examples + other.loss * other.num_examples) / len(new_yp) + new_id2answer_dict = dict(list(self.id2answer_dict.items()) + list(other.id2answer_dict.items())) + new_id2score_dict = dict(list(self.id2answer_dict['scores'].items()) + list(other.id2answer_dict['scores'].items())) + new_id2answer_dict['scores'] = new_id2score_dict + if self.tensor_dict is not None: + new_tensor_dict = {key: np.concatenate((val, other.tensor_dict[key]), axis=0) for key, val in self.tensor_dict.items()} + return ForwardEvaluation(self.data_type, self.global_step, new_idxs, new_yp, new_yp2, new_loss, new_id2answer_dict, tensor_dict=new_tensor_dict) + + def __repr__(self): + return "{} step {}: loss={:.4f}".format(self.data_type, self.global_step, self.loss) + + +class F1Evaluation(AccuracyEvaluation): + def __init__(self, data_type, global_step, idxs, yp, yp2, y, correct, loss, f1s, id2answer_dict, tensor_dict=None): + super(F1Evaluation, self).__init__(data_type, global_step, idxs, yp, y, correct, loss, tensor_dict=tensor_dict) + self.yp2 = yp2 + self.f1s = f1s + self.f1 = float(np.mean(f1s)) + self.dict['yp2'] = yp2 + self.dict['f1s'] = f1s + self.dict['f1'] = self.f1 + self.id2answer_dict = id2answer_dict + f1_summary = tf.Summary(value=[tf.Summary.Value(tag='{}/f1'.format(data_type), simple_value=self.f1)]) + self.summaries.append(f1_summary) + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_idxs = self.idxs + other.idxs + new_yp = self.yp + other.yp + new_yp2 = self.yp2 + other.yp2 + new_y = self.y + other.y + new_correct = self.correct + other.correct + new_f1s = self.f1s + other.f1s + new_loss = (self.loss * self.num_examples + other.loss * other.num_examples) / len(new_correct) + new_id2answer_dict = dict(list(self.id2answer_dict.items()) + list(other.id2answer_dict.items())) + new_id2score_dict = dict(list(self.id2answer_dict['scores'].items()) + list(other.id2answer_dict['scores'].items())) + new_id2answer_dict['scores'] = new_id2score_dict + if 'na' in self.id2answer_dict: + new_id2na_dict = dict(list(self.id2answer_dict['na'].items()) + list(other.id2answer_dict['na'].items())) + new_id2answer_dict['na'] = new_id2na_dict + e = F1Evaluation(self.data_type, self.global_step, new_idxs, new_yp, new_yp2, new_y, new_correct, new_loss, new_f1s, new_id2answer_dict) + if 'wyp' in self.dict: + new_wyp = self.dict['wyp'] + other.dict['wyp'] + e.dict['wyp'] = new_wyp + return e + + def __repr__(self): + return "{} step {}: accuracy={:.4f}, f1={:.4f}, loss={:.4f}".format(self.data_type, self.global_step, self.acc, self.f1, self.loss) + + +class F1Evaluator(LabeledEvaluator): + def __init__(self, config, model, tensor_dict=None): + super(F1Evaluator, self).__init__(config, model, tensor_dict=tensor_dict) + self.yp2 = model.yp2 + self.wyp = model.wyp + self.loss = model.loss + if config.na: + self.na = model.na_prob + + def get_evaluation(self, sess, batch): + idxs, data_set = self._split_batch(batch) + assert isinstance(data_set, DataSet) + feed_dict = self._get_feed_dict(batch) + if self.config.na: + global_step, yp, yp2, wyp, loss, na, vals = sess.run([self.global_step, self.yp, self.yp2, self.wyp, self.loss, self.na, list(self.tensor_dict.values())], feed_dict=feed_dict) + else: + global_step, yp, yp2, wyp, loss, vals = sess.run([self.global_step, self.yp, self.yp2, self.wyp, self.loss, list(self.tensor_dict.values())], feed_dict=feed_dict) + y = data_set.data['y'] + if self.config.squash: + new_y = [] + for xi, yi in zip(data_set.data['x'], y): + new_yi = [] + for start, stop in yi: + start_offset = sum(map(len, xi[:start[0]])) + stop_offset = sum(map(len, xi[:stop[0]])) + new_start = 0, start_offset + start[1] + new_stop = 0, stop_offset + stop[1] + new_yi.append((new_start, new_stop)) + new_y.append(new_yi) + y = new_y + if self.config.single: + new_y = [] + for yi in y: + new_yi = [] + for start, stop in yi: + new_start = 0, start[1] + new_stop = 0, stop[1] + new_yi.append((new_start, new_stop)) + new_y.append(new_yi) + y = new_y + + yp, yp2, wyp = yp[:data_set.num_examples], yp2[:data_set.num_examples], wyp[:data_set.num_examples] + if self.config.wy: + spans, scores = zip(*[get_best_span_wy(wypi, self.config.th) for wypi in wyp]) + else: + spans, scores = zip(*[get_best_span(ypi, yp2i) for ypi, yp2i in zip(yp, yp2)]) + + def _get(xi, span): + if len(xi) <= span[0][0]: + return [""] + if len(xi[span[0][0]]) <= span[1][1]: + return [""] + return xi[span[0][0]][span[0][1]:span[1][1]] + + def _get2(context, xi, span): + if len(xi) <= span[0][0]: + return "" + if len(xi[span[0][0]]) <= span[1][1]: + return "" + return get_phrase(context, xi, span) + + id2answer_dict = {id_: _get2(context, xi, span) + for id_, xi, span, context in zip(data_set.data['ids'], data_set.data['x'], spans, data_set.data['p'])} + id2score_dict = {id_: score for id_, score in zip(data_set.data['ids'], scores)} + id2answer_dict['scores'] = id2score_dict + if self.config.na: + id2na_dict = {id_: float(each) for id_, each in zip(data_set.data['ids'], na)} + id2answer_dict['na'] = id2na_dict + correct = [self.__class__.compare2(yi, span) for yi, span in zip(y, spans)] + f1s = [self.__class__.span_f1(yi, span) for yi, span in zip(y, spans)] + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + e = F1Evaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), yp2.tolist(), y, + correct, float(loss), f1s, id2answer_dict, tensor_dict=tensor_dict) + if self.config.wy: + e.dict['wyp'] = wyp.tolist() + return e + + def _split_batch(self, batch): + return batch + + def _get_feed_dict(self, batch): + return self.model.get_feed_dict(batch[1], False) + + @staticmethod + def compare(yi, ypi, yp2i): + for start, stop in yi: + aypi = argmax(ypi) + mask = np.zeros(yp2i.shape) + mask[aypi[0], aypi[1]:] = np.ones([yp2i.shape[1] - aypi[1]]) + if tuple(start) == aypi and (stop[0], stop[1]-1) == argmax(yp2i * mask): + return True + return False + + @staticmethod + def compare2(yi, span): + for start, stop in yi: + if tuple(start) == span[0] and tuple(stop) == span[1]: + return True + return False + + @staticmethod + def span_f1(yi, span): + max_f1 = 0 + for start, stop in yi: + if start[0] == span[0][0]: + true_span = start[1], stop[1] + pred_span = span[0][1], span[1][1] + f1 = span_f1(true_span, pred_span) + max_f1 = max(f1, max_f1) + return max_f1 + + +class MultiGPUF1Evaluator(F1Evaluator): + def __init__(self, config, models, tensor_dict=None): + super(MultiGPUF1Evaluator, self).__init__(config, models[0], tensor_dict=tensor_dict) + self.models = models + with tf.name_scope("eval_concat"): + N, M, JX = config.batch_size, config.max_num_sents, config.max_sent_size + self.yp = tf.concat(axis=0, values=[padded_reshape(model.yp, [N, M, JX]) for model in models]) + self.yp2 = tf.concat(axis=0, values=[padded_reshape(model.yp2, [N, M, JX]) for model in models]) + self.wy = tf.concat(axis=0, values=[padded_reshape(model.wy, [N, M, JX]) for model in models]) + self.loss = tf.add_n([model.loss for model in models])/len(models) + + def _split_batch(self, batches): + idxs_list, data_sets = zip(*batches) + idxs = sum(idxs_list, ()) + data_set = sum(data_sets, data_sets[0].get_empty()) + return idxs, data_set + + def _get_feed_dict(self, batches): + feed_dict = {} + for model, (_, data_set) in zip(self.models, batches): + feed_dict.update(model.get_feed_dict(data_set, False)) + return feed_dict + + +class ForwardEvaluator(Evaluator): + def __init__(self, config, model, tensor_dict=None): + super(ForwardEvaluator, self).__init__(config, model, tensor_dict=tensor_dict) + self.yp2 = model.yp2 + self.loss = model.loss + if config.na: + self.na = model.na_prob + + def get_evaluation(self, sess, batch): + idxs, data_set = batch + assert isinstance(data_set, DataSet) + feed_dict = self.model.get_feed_dict(data_set, False) + if self.config.na: + global_step, yp, yp2, loss, na, vals = sess.run([self.global_step, self.yp, self.yp2, self.loss, self.na, list(self.tensor_dict.values())], feed_dict=feed_dict) + else: + global_step, yp, yp2, loss, vals = sess.run([self.global_step, self.yp, self.yp2, self.loss, list(self.tensor_dict.values())], feed_dict=feed_dict) + + yp, yp2 = yp[:data_set.num_examples], yp2[:data_set.num_examples] + spans, scores = zip(*[get_best_span(ypi, yp2i) for ypi, yp2i in zip(yp, yp2)]) + + def _get(xi, span): + if len(xi) <= span[0][0]: + return [""] + if len(xi[span[0][0]]) <= span[1][1]: + return [""] + return xi[span[0][0]][span[0][1]:span[1][1]] + + def _get2(context, xi, span): + if len(xi) <= span[0][0]: + return "" + if len(xi[span[0][0]]) <= span[1][1]: + return "" + return get_phrase(context, xi, span) + + id2answer_dict = {id_: _get2(context, xi, span) + for id_, xi, span, context in zip(data_set.data['ids'], data_set.data['x'], spans, data_set.data['p'])} + id2score_dict = {id_: score for id_, score in zip(data_set.data['ids'], scores)} + id2answer_dict['scores'] = id2score_dict + if self.config.na: + id2na_dict = {id_: float(each) for id_, each in zip(data_set.data['ids'], na)} + id2answer_dict['na'] = id2na_dict + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + e = ForwardEvaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), yp2.tolist(), float(loss), id2answer_dict, tensor_dict=tensor_dict) + # TODO : wy support + return e + + @staticmethod + def compare(yi, ypi, yp2i): + for start, stop in yi: + aypi = argmax(ypi) + mask = np.zeros(yp2i.shape) + mask[aypi[0], aypi[1]:] = np.ones([yp2i.shape[1] - aypi[1]]) + if tuple(start) == aypi and (stop[0], stop[1]-1) == argmax(yp2i * mask): + return True + return False + + @staticmethod + def compare2(yi, span): + for start, stop in yi: + if tuple(start) == span[0] and tuple(stop) == span[1]: + return True + return False + + @staticmethod + def span_f1(yi, span): + max_f1 = 0 + for start, stop in yi: + if start[0] == span[0][0]: + true_span = start[1], stop[1] + pred_span = span[0][1], span[1][1] + f1 = span_f1(true_span, pred_span) + max_f1 = max(f1, max_f1) + return max_f1 + + diff --git a/tensorflow/SQuAD/basic/get_pr.py b/tensorflow/SQuAD/basic/get_pr.py new file mode 100644 index 0000000..be99be1 --- /dev/null +++ b/tensorflow/SQuAD/basic/get_pr.py @@ -0,0 +1,35 @@ +import json +import argparse + + +def get_args(): + parser = argparse.ArgumentParser() + parser.add_argument("path") + parser.add_argument("-t", "--th", type=float, default=0.5) + # TODO : put more args here + return parser.parse_args() + + +def get_pr(args): + with open(args.path, 'r') as fp: + answers = json.load(fp) + + na = answers['na'] + + tp = sum(int(not id_.startswith("neg") and score < args.th) for id_, score in na.items()) + fp = sum(int(id_.startswith("neg") and score < args.th) for id_, score in na.items()) + tn = sum(int(id_.startswith("neg") and score >= args.th) for id_, score in na.items()) + fn = sum(int(not id_.startswith("neg") and score >= args.th) for id_, score in na.items()) + + p = tp / (tp + fp) + r = tp / (tp + fn) + print("p={:.3f}, r={:.3f}".format(p, r)) + + +def main(): + args = get_args() + get_pr(args) + +if __name__ == "__main__": + main() + diff --git a/tensorflow/SQuAD/basic/graph_handler.py b/tensorflow/SQuAD/basic/graph_handler.py new file mode 100644 index 0000000..141e395 --- /dev/null +++ b/tensorflow/SQuAD/basic/graph_handler.py @@ -0,0 +1,79 @@ +import gzip +import json +from json import encoder +import os + +import tensorflow as tf + +from basic.evaluator import Evaluation, F1Evaluation +from my.utils import short_floats + +import pickle + + +class GraphHandler(object): + def __init__(self, config, model): + self.config = config + self.model = model + self.saver = tf.train.Saver(max_to_keep=config.max_to_keep) + self.writer = None + self.save_path = os.path.join(config.save_dir, config.model_name) + + def initialize(self, sess): + sess.run(tf.global_variables_initializer()) + if self.config.load: + self._load(sess) + + if self.config.mode == 'train': + self.writer = tf.summary.FileWriter(self.config.log_dir, graph=tf.get_default_graph()) + + def save(self, sess, global_step=None): + saver = tf.train.Saver(max_to_keep=self.config.max_to_keep) + saver.save(sess, self.save_path, global_step=global_step) + + def _load(self, sess): + config = self.config + vars_ = {var.name.split(":")[0]: var for var in tf.global_variables()} + if config.load_ema: + ema = self.model.var_ema + for var in tf.trainable_variables(): + del vars_[var.name.split(":")[0]] + vars_[ema.average_name(var)] = var + saver = tf.train.Saver(vars_, max_to_keep=config.max_to_keep) + + if config.load_path: + save_path = config.load_path + elif config.load_step > 0: + save_path = os.path.join(config.save_dir, "{}-{}".format(config.model_name, config.load_step)) + else: + save_dir = config.save_dir + checkpoint = tf.train.get_checkpoint_state(save_dir) + assert checkpoint is not None, "cannot load checkpoint at {}".format(save_dir) + save_path = checkpoint.model_checkpoint_path + print("Loading saved model from {}".format(save_path)) + saver.restore(sess, save_path) + + def add_summary(self, summary, global_step): + self.writer.add_summary(summary, global_step) + + def add_summaries(self, summaries, global_step): + for summary in summaries: + self.add_summary(summary, global_step) + + def dump_eval(self, e, precision=2, path=None): + assert isinstance(e, Evaluation) + if self.config.dump_pickle: + path = path or os.path.join(self.config.eval_dir, "{}-{}.pklz".format(e.data_type, str(e.global_step).zfill(6))) + with gzip.open(path, 'wb', compresslevel=3) as fh: + pickle.dump(e.dict, fh) + else: + path = path or os.path.join(self.config.eval_dir, "{}-{}.json".format(e.data_type, str(e.global_step).zfill(6))) + with open(path, 'w') as fh: + json.dump(short_floats(e.dict, precision), fh) + + def dump_answer(self, e, path=None): + assert isinstance(e, Evaluation) + path = path or os.path.join(self.config.answer_dir, "{}-{}.json".format(e.data_type, str(e.global_step).zfill(6))) + with open(path, 'w') as fh: + json.dump(e.id2answer_dict, fh) + diff --git a/tensorflow/SQuAD/basic/main.py b/tensorflow/SQuAD/basic/main.py new file mode 100644 index 0000000..70b77e4 --- /dev/null +++ b/tensorflow/SQuAD/basic/main.py @@ -0,0 +1,233 @@ +import argparse +import json +import math +import os +import shutil +from pprint import pprint + +import tensorflow as tf +from tqdm import tqdm +import numpy as np + +from basic.evaluator import ForwardEvaluator, MultiGPUF1Evaluator +from basic.graph_handler import GraphHandler +from basic.model import get_multi_gpu_models +from basic.trainer import MultiGPUTrainer +from basic.read_data import read_data, get_squad_data_filter, update_config +from my.tensorflow import get_num_params + + +def main(config): + set_dirs(config) + with tf.device(config.device): + if config.mode == 'train': + _train(config) + elif config.mode == 'test': + _test(config) + elif config.mode == 'forward': + _forward(config) + else: + raise ValueError("invalid value for 'mode': {}".format(config.mode)) + + +def set_dirs(config): + # create directories + assert config.load or config.mode == 'train', "config.load must be True if not training" + if not config.load and os.path.exists(config.out_dir): + shutil.rmtree(config.out_dir) + + config.save_dir = os.path.join(config.out_dir, "save") + config.log_dir = os.path.join(config.out_dir, "log") + config.eval_dir = os.path.join(config.out_dir, "eval") + config.answer_dir = os.path.join(config.out_dir, "answer") + if not os.path.exists(config.out_dir): + os.makedirs(config.out_dir) + if not os.path.exists(config.save_dir): + os.mkdir(config.save_dir) + if not os.path.exists(config.log_dir): + os.mkdir(config.log_dir) + if not os.path.exists(config.answer_dir): + os.mkdir(config.answer_dir) + if not os.path.exists(config.eval_dir): + os.mkdir(config.eval_dir) + + +def _config_debug(config): + if config.debug: + config.num_steps = 2 + config.eval_period = 1 + config.log_period = 1 + config.save_period = 1 + config.val_num_batches = 2 + config.test_num_batches = 2 + + +def _train(config): + data_filter = get_squad_data_filter(config) + train_data = read_data(config, 'train', config.load, data_filter=data_filter) + dev_data = read_data(config, 'dev', True, data_filter=data_filter) + update_config(config, [train_data, dev_data]) + + _config_debug(config) + + word2vec_dict = train_data.shared['lower_word2vec'] if config.lower_word else train_data.shared['word2vec'] + word2idx_dict = train_data.shared['word2idx'] + idx2vec_dict = {word2idx_dict[word]: vec for word, vec in word2vec_dict.items() if word in word2idx_dict} + emb_mat = np.array([idx2vec_dict[idx] if idx in idx2vec_dict + else np.random.multivariate_normal(np.zeros(config.word_emb_size), np.eye(config.word_emb_size)) + for idx in range(config.word_vocab_size)]) + config.emb_mat = emb_mat + + # construct model graph and variables (using default graph) + pprint(config.__flags, indent=2) + models = get_multi_gpu_models(config) + model = models[0] + print("num params: {}".format(get_num_params())) + trainer = MultiGPUTrainer(config, models) + evaluator = MultiGPUF1Evaluator(config, models, tensor_dict=model.tensor_dict if config.vis else None) + graph_handler = GraphHandler(config, model) # controls all tensors and variables in the graph, including loading /saving + + # Variables + sess = tf.Session(config=tf.ConfigProto(allow_soft_placement=True)) + graph_handler.initialize(sess) + + # Begin training + num_steps = config.num_steps or int(math.ceil(train_data.num_examples / (config.batch_size * config.num_gpus))) * config.num_epochs + global_step = 0 + for batches in tqdm(train_data.get_multi_batches(config.batch_size, config.num_gpus, + num_steps=num_steps, shuffle=True, cluster=config.cluster), total=num_steps): + global_step = sess.run(model.global_step) + 1 # +1 because all calculations are done after step + get_summary = global_step % config.log_period == 0 + loss, summary, train_op = trainer.step(sess, batches, get_summary=get_summary) + if get_summary: + graph_handler.add_summary(summary, global_step) + + # occasional saving + if global_step % config.save_period == 0: + graph_handler.save(sess, global_step=global_step) + + if not config.eval: + continue + # Occasional evaluation + if global_step % config.eval_period == 0: + num_steps = math.ceil(dev_data.num_examples / (config.batch_size * config.num_gpus)) + if 0 < config.val_num_batches < num_steps: + num_steps = config.val_num_batches + e_train = evaluator.get_evaluation_from_batches( + sess, tqdm(train_data.get_multi_batches(config.batch_size, config.num_gpus, num_steps=num_steps), total=num_steps) + ) + graph_handler.add_summaries(e_train.summaries, global_step) + e_dev = evaluator.get_evaluation_from_batches( + sess, tqdm(dev_data.get_multi_batches(config.batch_size, config.num_gpus, num_steps=num_steps), total=num_steps)) + graph_handler.add_summaries(e_dev.summaries, global_step) + + if config.dump_eval: + graph_handler.dump_eval(e_dev) + if config.dump_answer: + graph_handler.dump_answer(e_dev) + if global_step % config.save_period != 0: + graph_handler.save(sess, global_step=global_step) + + +def _test(config): + test_data = read_data(config, 'test', True) + update_config(config, [test_data]) + + _config_debug(config) + + if config.use_glove_for_unk: + word2vec_dict = test_data.shared['lower_word2vec'] if config.lower_word else test_data.shared['word2vec'] + new_word2idx_dict = test_data.shared['new_word2idx'] + idx2vec_dict = {idx: word2vec_dict[word] for word, idx in new_word2idx_dict.items()} + new_emb_mat = np.array([idx2vec_dict[idx] for idx in range(len(idx2vec_dict))], dtype='float32') + config.new_emb_mat = new_emb_mat + + pprint(config.__flags, indent=2) + models = get_multi_gpu_models(config) + model = models[0] + evaluator = MultiGPUF1Evaluator(config, models, tensor_dict=models[0].tensor_dict if config.vis else None) + graph_handler = GraphHandler(config, model) + + sess = tf.Session(config=tf.ConfigProto(allow_soft_placement=True)) + graph_handler.initialize(sess) + num_steps = math.ceil(test_data.num_examples / (config.batch_size * config.num_gpus)) + if 0 < config.test_num_batches < num_steps: + num_steps = config.test_num_batches + + e = None + for multi_batch in tqdm(test_data.get_multi_batches(config.batch_size, config.num_gpus, num_steps=num_steps, cluster=config.cluster), total=num_steps): + ei = evaluator.get_evaluation(sess, multi_batch) + e = ei if e is None else e + ei + if config.vis: + eval_subdir = os.path.join(config.eval_dir, "{}-{}".format(ei.data_type, str(ei.global_step).zfill(6))) + if not os.path.exists(eval_subdir): + os.mkdir(eval_subdir) + path = os.path.join(eval_subdir, str(ei.idxs[0]).zfill(8)) + graph_handler.dump_eval(ei, path=path) + + print(e) + if config.dump_answer: + print("dumping answer ...") + graph_handler.dump_answer(e) + if config.dump_eval: + print("dumping eval ...") + graph_handler.dump_eval(e) + + +def _forward(config): + assert config.load + test_data = read_data(config, config.forward_name, True) + update_config(config, [test_data]) + + _config_debug(config) + + if config.use_glove_for_unk: + word2vec_dict = test_data.shared['lower_word2vec'] if config.lower_word else test_data.shared['word2vec'] + new_word2idx_dict = test_data.shared['new_word2idx'] + idx2vec_dict = {idx: word2vec_dict[word] for word, idx in new_word2idx_dict.items()} + new_emb_mat = np.array([idx2vec_dict[idx] for idx in range(len(idx2vec_dict))], dtype='float32') + config.new_emb_mat = new_emb_mat + + pprint(config.__flags, indent=2) + models = get_multi_gpu_models(config) + model = models[0] + print("num params: {}".format(get_num_params())) + evaluator = ForwardEvaluator(config, model) + graph_handler = GraphHandler(config, model) # controls all tensors and variables in the graph, including loading /saving + + sess = tf.Session(config=tf.ConfigProto(allow_soft_placement=True)) + graph_handler.initialize(sess) + + num_batches = math.ceil(test_data.num_examples / config.batch_size) + if 0 < config.test_num_batches < num_batches: + num_batches = config.test_num_batches + e = evaluator.get_evaluation_from_batches(sess, tqdm(test_data.get_batches(config.batch_size, num_batches=num_batches), total=num_batches)) + print(e) + if config.dump_answer: + print("dumping answer ...") + graph_handler.dump_answer(e, path=config.answer_path) + if config.dump_eval: + print("dumping eval ...") + graph_handler.dump_eval(e, path=config.eval_path) + + +def _get_args(): + parser = argparse.ArgumentParser() + parser.add_argument("config_path") + return parser.parse_args() + + +class Config(object): + def __init__(self, **entries): + self.__dict__.update(entries) + + +def _run(): + args = _get_args() + with open(args.config_path, 'r') as fh: + config = Config(**json.load(fh)) + main(config) + + +if __name__ == "__main__": + _run() diff --git a/tensorflow/SQuAD/basic/model.py b/tensorflow/SQuAD/basic/model.py new file mode 100644 index 0000000..8f0dba0 --- /dev/null +++ b/tensorflow/SQuAD/basic/model.py @@ -0,0 +1,535 @@ +import random + +import itertools +import numpy as np +import tensorflow as tf +from tensorflow.contrib.rnn import BasicLSTMCell + +from basic.read_data import DataSet +from my.tensorflow import get_initializer +from my.tensorflow.nn import softsel, get_logits, highway_network, multi_conv1d +from my.tensorflow.rnn import bidirectional_dynamic_rnn +from my.tensorflow.rnn_cell import SwitchableDropoutWrapper, AttentionCell + + +def get_multi_gpu_models(config): + models = [] + with tf.variable_scope(tf.get_variable_scope()): + for gpu_idx in range(config.num_gpus): + with tf.name_scope("model_{}".format(gpu_idx)) as scope, tf.device("/{}:{}".format(config.device_type, gpu_idx)): + if gpu_idx > 0: + tf.get_variable_scope().reuse_variables() + model = Model(config, scope, rep=gpu_idx == 0) + models.append(model) + + # update the summary in a different scope to avoid reuse issue + with tf.variable_scope('loss_summary', reuse=False): + for gpu_idx in range(config.num_gpus): + with tf.name_scope("model_{}".format(gpu_idx)) as scope, tf.device("/{}:{}".format(config.device_type, gpu_idx)): + model = models[gpu_idx] + rep = gpu_idx == 0 + if rep: + model._build_var_ema() + if config.mode == 'train': + model._build_ema(); + model.summary = tf.summary.merge_all() + model.summary = tf.summary.merge(tf.get_collection("summaries", scope=model.scope)) + + return models + + +class Model(object): + def __init__(self, config, scope, rep=True): + self.scope = scope + self.config = config + self.global_step = tf.get_variable('global_step', shape=[], dtype='int32', + initializer=tf.constant_initializer(0), trainable=False) + + # Define forward inputs here + N, M, JX, JQ, VW, VC, W = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size, config.max_word_size + self.x = tf.placeholder('int32', [N, None, None], name='x') + self.cx = tf.placeholder('int32', [N, None, None, W], name='cx') + self.x_mask = tf.placeholder('bool', [N, None, None], name='x_mask') + self.q = tf.placeholder('int32', [N, None], name='q') + self.cq = tf.placeholder('int32', [N, None, W], name='cq') + self.q_mask = tf.placeholder('bool', [N, None], name='q_mask') + self.y = tf.placeholder('bool', [N, None, None], name='y') + self.y2 = tf.placeholder('bool', [N, None, None], name='y2') + self.wy = tf.placeholder('bool', [N, None, None], name='wy') + self.is_train = tf.placeholder('bool', [], name='is_train') + self.new_emb_mat = tf.placeholder('float', [None, config.word_emb_size], name='new_emb_mat') + self.na = tf.placeholder('bool', [N], name='na') + + # Define misc + self.tensor_dict = {} + + # Forward outputs / loss inputs + self.logits = None + self.yp = None + self.var_list = None + self.na_prob = None + + # Loss outputs + self.loss = None + + self._build_forward() + self._build_loss() + self.var_ema = None + # if rep: + # self._build_var_ema() + # if config.mode == 'train': + # self._build_ema() + + # self.summary = tf.summary.merge_all() + # self.summary = tf.summary.merge(tf.get_collection("summaries", scope=self.scope)) + + def _build_forward(self): + config = self.config + N, M, JX, JQ, VW, VC, d, W = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size, config.hidden_size, \ + config.max_word_size + JX = tf.shape(self.x)[2] + JQ = tf.shape(self.q)[1] + M = tf.shape(self.x)[1] + dc, dw, dco = config.char_emb_size, config.word_emb_size, config.char_out_size + + with tf.variable_scope("emb"): + if config.use_char_emb: + with tf.variable_scope("emb_var"), tf.device("/cpu:0"): + char_emb_mat = tf.get_variable("char_emb_mat", shape=[VC, dc], dtype='float') + + with tf.variable_scope("char"): + Acx = tf.nn.embedding_lookup(char_emb_mat, self.cx) # [N, M, JX, W, dc] + Acq = tf.nn.embedding_lookup(char_emb_mat, self.cq) # [N, JQ, W, dc] + Acx = tf.reshape(Acx, [-1, JX, W, dc]) + Acq = tf.reshape(Acq, [-1, JQ, W, dc]) + + filter_sizes = list(map(int, config.out_channel_dims.split(','))) + heights = list(map(int, config.filter_heights.split(','))) + assert sum(filter_sizes) == dco, (filter_sizes, dco) + with tf.variable_scope("conv"): + xx = multi_conv1d(Acx, filter_sizes, heights, "VALID", self.is_train, config.keep_prob, scope="xx") + if config.share_cnn_weights: + tf.get_variable_scope().reuse_variables() + qq = multi_conv1d(Acq, filter_sizes, heights, "VALID", self.is_train, config.keep_prob, scope="xx") + else: + qq = multi_conv1d(Acq, filter_sizes, heights, "VALID", self.is_train, config.keep_prob, scope="qq") + xx = tf.reshape(xx, [-1, M, JX, dco]) + qq = tf.reshape(qq, [-1, JQ, dco]) + + if config.use_word_emb: + with tf.variable_scope("emb_var"), tf.device("/cpu:0"): + if config.mode == 'train': + word_emb_mat = tf.get_variable("word_emb_mat", dtype='float', shape=[VW, dw], initializer=get_initializer(config.emb_mat)) + else: + word_emb_mat = tf.get_variable("word_emb_mat", shape=[VW, dw], dtype='float') + if config.use_glove_for_unk: + word_emb_mat = tf.concat(axis=0, values=[word_emb_mat, self.new_emb_mat]) + + with tf.name_scope("word"): + Ax = tf.nn.embedding_lookup(word_emb_mat, self.x) # [N, M, JX, d] + Aq = tf.nn.embedding_lookup(word_emb_mat, self.q) # [N, JQ, d] + self.tensor_dict['x'] = Ax + self.tensor_dict['q'] = Aq + if config.use_char_emb: + xx = tf.concat(axis=3, values=[xx, Ax]) # [N, M, JX, di] + qq = tf.concat(axis=2, values=[qq, Aq]) # [N, JQ, di] + else: + xx = Ax + qq = Aq + + # highway network + if config.highway: + with tf.variable_scope("highway"): + xx = highway_network(xx, config.highway_num_layers, True, wd=config.wd, is_train=self.is_train) + tf.get_variable_scope().reuse_variables() + qq = highway_network(qq, config.highway_num_layers, True, wd=config.wd, is_train=self.is_train) + + self.tensor_dict['xx'] = xx + self.tensor_dict['qq'] = qq + + cell_fw = BasicLSTMCell(d, state_is_tuple=True) + cell_bw = BasicLSTMCell(d, state_is_tuple=True) + d_cell_fw = SwitchableDropoutWrapper(cell_fw, self.is_train, input_keep_prob=config.input_keep_prob) + d_cell_bw = SwitchableDropoutWrapper(cell_bw, self.is_train, input_keep_prob=config.input_keep_prob) + cell2_fw = BasicLSTMCell(d, state_is_tuple=True) + cell2_bw = BasicLSTMCell(d, state_is_tuple=True) + d_cell2_fw = SwitchableDropoutWrapper(cell2_fw, self.is_train, input_keep_prob=config.input_keep_prob) + d_cell2_bw = SwitchableDropoutWrapper(cell2_bw, self.is_train, input_keep_prob=config.input_keep_prob) + cell3_fw = BasicLSTMCell(d, state_is_tuple=True) + cell3_bw = BasicLSTMCell(d, state_is_tuple=True) + d_cell3_fw = SwitchableDropoutWrapper(cell3_fw, self.is_train, input_keep_prob=config.input_keep_prob) + d_cell3_bw = SwitchableDropoutWrapper(cell3_bw, self.is_train, input_keep_prob=config.input_keep_prob) + cell4_fw = BasicLSTMCell(d, state_is_tuple=True) + cell4_bw = BasicLSTMCell(d, state_is_tuple=True) + d_cell4_fw = SwitchableDropoutWrapper(cell4_fw, self.is_train, input_keep_prob=config.input_keep_prob) + d_cell4_bw = SwitchableDropoutWrapper(cell4_bw, self.is_train, input_keep_prob=config.input_keep_prob) + x_len = tf.reduce_sum(tf.cast(self.x_mask, 'int32'), 2) # [N, M] + q_len = tf.reduce_sum(tf.cast(self.q_mask, 'int32'), 1) # [N] + + with tf.variable_scope("prepro"): + (fw_u, bw_u), ((_, fw_u_f), (_, bw_u_f)) = bidirectional_dynamic_rnn(d_cell_fw, d_cell_bw, qq, q_len, dtype='float', scope='u1') # [N, J, d], [N, d] + u = tf.concat(axis=2, values=[fw_u, bw_u]) + if config.share_lstm_weights: + tf.get_variable_scope().reuse_variables() + (fw_h, bw_h), _ = bidirectional_dynamic_rnn(cell_fw, cell_bw, xx, x_len, dtype='float', scope='u1') # [N, M, JX, 2d] + h = tf.concat(axis=3, values=[fw_h, bw_h]) # [N, M, JX, 2d] + else: + (fw_h, bw_h), _ = bidirectional_dynamic_rnn(cell_fw, cell_bw, xx, x_len, dtype='float', scope='h1') # [N, M, JX, 2d] + h = tf.concat(axis=3, values=[fw_h, bw_h]) # [N, M, JX, 2d] + self.tensor_dict['u'] = u + self.tensor_dict['h'] = h + + with tf.variable_scope("main"): + if config.dynamic_att: + p0 = h + u = tf.reshape(tf.tile(tf.expand_dims(u, 1), [1, M, 1, 1]), [N * M, JQ, 2 * d]) + q_mask = tf.reshape(tf.tile(tf.expand_dims(self.q_mask, 1), [1, M, 1]), [N * M, JQ]) + first_cell_fw = AttentionCell(cell2_fw, u, mask=q_mask, mapper='sim', + input_keep_prob=self.config.input_keep_prob, is_train=self.is_train) + first_cell_bw = AttentionCell(cell2_bw, u, mask=q_mask, mapper='sim', + input_keep_prob=self.config.input_keep_prob, is_train=self.is_train) + second_cell_fw = AttentionCell(cell3_fw, u, mask=q_mask, mapper='sim', + input_keep_prob=self.config.input_keep_prob, is_train=self.is_train) + second_cell_bw = AttentionCell(cell3_bw, u, mask=q_mask, mapper='sim', + input_keep_prob=self.config.input_keep_prob, is_train=self.is_train) + else: + p0 = attention_layer(config, self.is_train, h, u, h_mask=self.x_mask, u_mask=self.q_mask, scope="p0", tensor_dict=self.tensor_dict) + first_cell_fw = d_cell2_fw + second_cell_fw = d_cell3_fw + first_cell_bw = d_cell2_bw + second_cell_bw = d_cell3_bw + + (fw_g0, bw_g0), _ = bidirectional_dynamic_rnn(first_cell_fw, first_cell_bw, p0, x_len, dtype='float', scope='g0') # [N, M, JX, 2d] + g0 = tf.concat(axis=3, values=[fw_g0, bw_g0]) + (fw_g1, bw_g1), _ = bidirectional_dynamic_rnn(second_cell_fw, second_cell_bw, g0, x_len, dtype='float', scope='g1') # [N, M, JX, 2d] + g1 = tf.concat(axis=3, values=[fw_g1, bw_g1]) + + logits = get_logits([g1, p0], d, True, wd=config.wd, input_keep_prob=config.input_keep_prob, + mask=self.x_mask, is_train=self.is_train, func=config.answer_func, scope='logits1') + a1i = softsel(tf.reshape(g1, [N, M * JX, 2 * d]), tf.reshape(logits, [N, M * JX])) + a1i = tf.tile(tf.expand_dims(tf.expand_dims(a1i, 1), 1), [1, M, JX, 1]) + + (fw_g2, bw_g2), _ = bidirectional_dynamic_rnn(d_cell4_fw, d_cell4_bw, tf.concat(axis=3, values=[p0, g1, a1i, g1 * a1i]), + x_len, dtype='float', scope='g2') # [N, M, JX, 2d] + g2 = tf.concat(axis=3, values=[fw_g2, bw_g2]) + logits2 = get_logits([g2, p0], d, True, wd=config.wd, input_keep_prob=config.input_keep_prob, + mask=self.x_mask, + is_train=self.is_train, func=config.answer_func, scope='logits2') + + flat_logits = tf.reshape(logits, [-1, M * JX]) + flat_yp = tf.nn.softmax(flat_logits) # [-1, M*JX] + flat_logits2 = tf.reshape(logits2, [-1, M * JX]) + flat_yp2 = tf.nn.softmax(flat_logits2) + + if config.na: + na_bias = tf.get_variable("na_bias", shape=[], dtype='float') + na_bias_tiled = tf.tile(tf.reshape(na_bias, [1, 1]), [N, 1]) # [N, 1] + concat_flat_logits = tf.concat(axis=1, values=[na_bias_tiled, flat_logits]) + concat_flat_yp = tf.nn.softmax(concat_flat_logits) + na_prob = tf.squeeze(tf.slice(concat_flat_yp, [0, 0], [-1, 1]), [1]) + flat_yp = tf.slice(concat_flat_yp, [0, 1], [-1, -1]) + + concat_flat_logits2 = tf.concat(axis=1, values=[na_bias_tiled, flat_logits2]) + concat_flat_yp2 = tf.nn.softmax(concat_flat_logits2) + na_prob2 = tf.squeeze(tf.slice(concat_flat_yp2, [0, 0], [-1, 1]), [1]) # [N] + flat_yp2 = tf.slice(concat_flat_yp2, [0, 1], [-1, -1]) + + self.concat_logits = concat_flat_logits + self.concat_logits2 = concat_flat_logits2 + self.na_prob = na_prob * na_prob2 + + yp = tf.reshape(flat_yp, [-1, M, JX]) + yp2 = tf.reshape(flat_yp2, [-1, M, JX]) + wyp = tf.nn.sigmoid(logits2) + + self.tensor_dict['g1'] = g1 + self.tensor_dict['g2'] = g2 + + self.logits = flat_logits + self.logits2 = flat_logits2 + self.yp = yp + self.yp2 = yp2 + self.wyp = wyp + + def _build_loss(self): + config = self.config + JX = tf.shape(self.x)[2] + M = tf.shape(self.x)[1] + JQ = tf.shape(self.q)[1] + + loss_mask = tf.reduce_max(tf.cast(self.q_mask, 'float'), 1) + if config.wy: + losses = tf.nn.sigmoid_cross_entropy_with_logits( + logits=tf.reshape(self.logits2, [-1, M, JX]), labels=tf.cast(self.wy, 'float')) # [N, M, JX] + num_pos = tf.reduce_sum(tf.cast(self.wy, 'float')) + num_neg = tf.reduce_sum(tf.cast(self.x_mask, 'float')) - num_pos + damp_ratio = num_pos / num_neg + dampened_losses = losses * ( + (tf.cast(self.x_mask, 'float') - tf.cast(self.wy, 'float')) * damp_ratio + tf.cast(self.wy, 'float')) + new_losses = tf.reduce_sum(dampened_losses, [1, 2]) + ce_loss = tf.reduce_mean(loss_mask * new_losses) + """ + if config.na: + na = tf.reshape(self.na, [-1, 1]) + concat_y = tf.concat(1, [na, tf.reshape(self.wy, [-1, M * JX])]) + losses = tf.nn.softmax_cross_entropy_with_logits( + self.concat_logits, tf.cast(concat_y, 'float') / tf.reduce_sum(tf.cast(self.wy, 'float'))) + else: + losses = tf.nn.softmax_cross_entropy_with_logits( + self.logits2, tf.cast(tf.reshape(self.wy, [-1, M * JX]), 'float') / tf.reduce_sum(tf.cast(self.wy, 'float'))) + ce_loss = tf.reduce_mean(loss_mask * losses) + """ + tf.add_to_collection('losses', ce_loss) + + else: + if config.na: + na = tf.reshape(self.na, [-1, 1]) + concat_y = tf.concat(axis=1, values=[na, tf.reshape(self.y, [-1, M * JX])]) + losses = tf.nn.softmax_cross_entropy_with_logits(logits=self.concat_logits, labels=tf.cast(concat_y, 'float')) + concat_y2 = tf.concat(axis=1, values=[na, tf.reshape(self.y2, [-1, M * JX])]) + losses2 = tf.nn.softmax_cross_entropy_with_logits(logits=self.concat_logits2, labels=tf.cast(concat_y2, 'float')) + else: + losses = tf.nn.softmax_cross_entropy_with_logits( + logits=self.logits, labels=tf.cast(tf.reshape(self.y, [-1, M * JX]), 'float')) + losses2 = tf.nn.softmax_cross_entropy_with_logits( + logits=self.logits2, labels=tf.cast(tf.reshape(self.y2, [-1, M * JX]), 'float')) + ce_loss = tf.reduce_mean(loss_mask * losses) + ce_loss2 = tf.reduce_mean(loss_mask * losses2) + tf.add_to_collection('losses', ce_loss) + tf.add_to_collection("losses", ce_loss2) + + self.loss = tf.add_n(tf.get_collection('losses', scope=self.scope), name='loss') + tf.summary.scalar(self.loss.op.name, self.loss) + tf.add_to_collection('ema/scalar', self.loss) + + def _build_ema(self): + self.ema = tf.train.ExponentialMovingAverage(self.config.decay) + ema = self.ema + tensors = tf.get_collection("ema/scalar", scope=self.scope) + tf.get_collection("ema/vector", scope=self.scope) + ema_op = ema.apply(tensors) + for var in tf.get_collection("ema/scalar", scope=self.scope): + ema_var = ema.average(var) + tf.summary.scalar(ema_var.op.name, ema_var) + for var in tf.get_collection("ema/vector", scope=self.scope): + ema_var = ema.average(var) + tf.summary.histogram(ema_var.op.name, ema_var) + + with tf.control_dependencies([ema_op]): + self.loss = tf.identity(self.loss) + + def _build_var_ema(self): + self.var_ema = tf.train.ExponentialMovingAverage(self.config.var_decay) + ema = self.var_ema + ema_op = ema.apply(tf.trainable_variables()) + with tf.control_dependencies([ema_op]): + self.loss = tf.identity(self.loss) + + def get_loss(self): + return self.loss + + def get_global_step(self): + return self.global_step + + def get_var_list(self): + return self.var_list + + def get_feed_dict(self, batch, is_train, supervised=True): + assert isinstance(batch, DataSet) + config = self.config + N, M, JX, JQ, VW, VC, d, W = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size, config.hidden_size, config.max_word_size + feed_dict = {} + + if config.len_opt: + """ + Note that this optimization results in variable GPU RAM usage (i.e. can cause OOM in the middle of training.) + First test without len_opt and make sure no OOM, and use len_opt + """ + if sum(len(sent) for para in batch.data['x'] for sent in para) == 0: + new_JX = 1 + else: + new_JX = max(len(sent) for para in batch.data['x'] for sent in para) + JX = min(JX, new_JX) + + if sum(len(ques) for ques in batch.data['q']) == 0: + new_JQ = 1 + else: + new_JQ = max(len(ques) for ques in batch.data['q']) + JQ = min(JQ, new_JQ) + + if config.cpu_opt: + if sum(len(para) for para in batch.data['x']) == 0: + new_M = 1 + else: + new_M = max(len(para) for para in batch.data['x']) + M = min(M, new_M) + + x = np.zeros([N, M, JX], dtype='int32') + cx = np.zeros([N, M, JX, W], dtype='int32') + x_mask = np.zeros([N, M, JX], dtype='bool') + q = np.zeros([N, JQ], dtype='int32') + cq = np.zeros([N, JQ, W], dtype='int32') + q_mask = np.zeros([N, JQ], dtype='bool') + + feed_dict[self.x] = x + feed_dict[self.x_mask] = x_mask + feed_dict[self.cx] = cx + feed_dict[self.q] = q + feed_dict[self.cq] = cq + feed_dict[self.q_mask] = q_mask + feed_dict[self.is_train] = is_train + if config.use_glove_for_unk: + feed_dict[self.new_emb_mat] = batch.shared['new_emb_mat'] + + X = batch.data['x'] + CX = batch.data['cx'] + + if supervised: + y = np.zeros([N, M, JX], dtype='bool') + y2 = np.zeros([N, M, JX], dtype='bool') + wy = np.zeros([N, M, JX], dtype='bool') + na = np.zeros([N], dtype='bool') + feed_dict[self.y] = y + feed_dict[self.y2] = y2 + feed_dict[self.wy] = wy + feed_dict[self.na] = na + + for i, (xi, cxi, yi, nai) in enumerate(zip(X, CX, batch.data['y'], batch.data['na'])): + if nai: + na[i] = nai + continue + start_idx, stop_idx = random.choice(yi) + j, k = start_idx + j2, k2 = stop_idx + if config.single: + X[i] = [xi[j]] + CX[i] = [cxi[j]] + j, j2 = 0, 0 + if config.squash: + offset = sum(map(len, xi[:j])) + j, k = 0, k + offset + offset = sum(map(len, xi[:j2])) + j2, k2 = 0, k2 + offset + y[i, j, k] = True + y2[i, j2, k2-1] = True + if j == j2: + wy[i, j, k:k2] = True + else: + wy[i, j, k:len(batch.data['x'][i][j])] = True + wy[i, j2, :k2] = True + + def _get_word(word): + d = batch.shared['word2idx'] + for each in (word, word.lower(), word.capitalize(), word.upper()): + if each in d: + return d[each] + if config.use_glove_for_unk: + d2 = batch.shared['new_word2idx'] + for each in (word, word.lower(), word.capitalize(), word.upper()): + if each in d2: + return d2[each] + len(d) + return 1 + + def _get_char(char): + d = batch.shared['char2idx'] + if char in d: + return d[char] + return 1 + + for i, xi in enumerate(X): + if self.config.squash: + xi = [list(itertools.chain(*xi))] + for j, xij in enumerate(xi): + if j == config.max_num_sents: + break + for k, xijk in enumerate(xij): + if k == config.max_sent_size: + break + each = _get_word(xijk) + assert isinstance(each, int), each + x[i, j, k] = each + x_mask[i, j, k] = True + + for i, cxi in enumerate(CX): + if self.config.squash: + cxi = [list(itertools.chain(*cxi))] + for j, cxij in enumerate(cxi): + if j == config.max_num_sents: + break + for k, cxijk in enumerate(cxij): + if k == config.max_sent_size: + break + for l, cxijkl in enumerate(cxijk): + if l == config.max_word_size: + break + cx[i, j, k, l] = _get_char(cxijkl) + + for i, qi in enumerate(batch.data['q']): + for j, qij in enumerate(qi): + q[i, j] = _get_word(qij) + q_mask[i, j] = True + + for i, cqi in enumerate(batch.data['cq']): + for j, cqij in enumerate(cqi): + for k, cqijk in enumerate(cqij): + cq[i, j, k] = _get_char(cqijk) + if k + 1 == config.max_word_size: + break + + if supervised: + assert np.sum(~(x_mask | ~wy)) == 0 + + return feed_dict + + +def bi_attention(config, is_train, h, u, h_mask=None, u_mask=None, scope=None, tensor_dict=None): + with tf.variable_scope(scope or "bi_attention"): + JX = tf.shape(h)[2] + M = tf.shape(h)[1] + JQ = tf.shape(u)[1] + h_aug = tf.tile(tf.expand_dims(h, 3), [1, 1, 1, JQ, 1]) + u_aug = tf.tile(tf.expand_dims(tf.expand_dims(u, 1), 1), [1, M, JX, 1, 1]) + if h_mask is None: + hu_mask = None + else: + h_mask_aug = tf.tile(tf.expand_dims(h_mask, 3), [1, 1, 1, JQ]) + u_mask_aug = tf.tile(tf.expand_dims(tf.expand_dims(u_mask, 1), 1), [1, M, JX, 1]) + hu_mask = h_mask_aug & u_mask_aug + + u_logits = get_logits([h_aug, u_aug], None, True, wd=config.wd, mask=hu_mask, + is_train=is_train, func=config.logit_func, scope='u_logits') # [N, M, JX, JQ] + u_a = softsel(u_aug, u_logits) # [N, M, JX, d] + h_a = softsel(h, tf.reduce_max(u_logits, 3)) # [N, M, d] + h_a = tf.tile(tf.expand_dims(h_a, 2), [1, 1, JX, 1]) + + if tensor_dict is not None: + a_u = tf.nn.softmax(u_logits) # [N, M, JX, JQ] + a_h = tf.nn.softmax(tf.reduce_max(u_logits, 3)) + tensor_dict['a_u'] = a_u + tensor_dict['a_h'] = a_h + variables = tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES, scope=tf.get_variable_scope().name) + for var in variables: + tensor_dict[var.name] = var + + return u_a, h_a + + +def attention_layer(config, is_train, h, u, h_mask=None, u_mask=None, scope=None, tensor_dict=None): + with tf.variable_scope(scope or "attention_layer"): + JX = tf.shape(h)[2] + M = tf.shape(h)[1] + JQ = tf.shape(u)[1] + if config.q2c_att or config.c2q_att: + u_a, h_a = bi_attention(config, is_train, h, u, h_mask=h_mask, u_mask=u_mask, tensor_dict=tensor_dict) + if not config.c2q_att: + u_a = tf.tile(tf.expand_dims(tf.expand_dims(tf.reduce_mean(u, 1), 1), 1), [1, M, JX, 1]) + if config.q2c_att: + p0 = tf.concat(axis=3, values=[h, u_a, h * u_a, h * h_a]) + else: + p0 = tf.concat(axis=3, values=[h, u_a, h * u_a]) + return p0 diff --git a/tensorflow/SQuAD/basic/read_data.py b/tensorflow/SQuAD/basic/read_data.py new file mode 100644 index 0000000..9f4f2c1 --- /dev/null +++ b/tensorflow/SQuAD/basic/read_data.py @@ -0,0 +1,316 @@ +import json +import os +import random +import itertools +import math +from collections import defaultdict + +import numpy as np + +from my.tensorflow import grouper +from my.utils import index + + +class Data(object): + def get_size(self): + raise NotImplementedError() + + def get_by_idxs(self, idxs): + """ + Efficient way to obtain a batch of items from filesystem + :param idxs: + :return dict: {'X': [,], 'Y', } + """ + data = defaultdict(list) + for idx in idxs: + each_data = self.get_one(idx) + for key, val in each_data.items(): + data[key].append(val) + return data + + def get_one(self, idx): + raise NotImplementedError() + + def get_empty(self): + raise NotImplementedError() + + def __add__(self, other): + raise NotImplementedError() + + +class DataSet(object): + def __init__(self, data, data_type, shared=None, valid_idxs=None): + self.data = data # e.g. {'X': [0, 1, 2], 'Y': [2, 3, 4]} + self.data_type = data_type + self.shared = shared + total_num_examples = self.get_data_size() + self.valid_idxs = range(total_num_examples) if valid_idxs is None else valid_idxs + self.num_examples = len(self.valid_idxs) + + def _sort_key(self, idx): + rx = self.data['*x'][idx] + x = self.shared['x'][rx[0]][rx[1]] + return max(map(len, x)) + + def get_data_size(self): + if isinstance(self.data, dict): + return len(next(iter(self.data.values()))) + elif isinstance(self.data, Data): + return self.data.get_size() + raise Exception() + + def get_by_idxs(self, idxs): + if isinstance(self.data, dict): + out = defaultdict(list) + for key, val in self.data.items(): + out[key].extend(val[idx] for idx in idxs) + return out + elif isinstance(self.data, Data): + return self.data.get_by_idxs(idxs) + raise Exception() + + def get_batches(self, batch_size, num_batches=None, shuffle=False, cluster=False): + """ + + :param batch_size: + :param num_batches: + :param shuffle: + :param cluster: cluster examples by their lengths; this might give performance boost (i.e. faster training). + :return: + """ + num_batches_per_epoch = int(math.ceil(self.num_examples / batch_size)) + if num_batches is None: + num_batches = num_batches_per_epoch + num_epochs = int(math.ceil(num_batches / num_batches_per_epoch)) + + if shuffle: + random_idxs = random.sample(self.valid_idxs, len(self.valid_idxs)) + if cluster: + sorted_idxs = sorted(random_idxs, key=self._sort_key) + sorted_grouped = lambda: list(grouper(sorted_idxs, batch_size)) + grouped = lambda: random.sample(sorted_grouped(), num_batches_per_epoch) + else: + random_grouped = lambda: list(grouper(random_idxs, batch_size)) + grouped = random_grouped + else: + raw_grouped = lambda: list(grouper(self.valid_idxs, batch_size)) + grouped = raw_grouped + + batch_idx_tuples = itertools.chain.from_iterable(grouped() for _ in range(num_epochs)) + for _ in range(num_batches): + batch_idxs = tuple(i for i in next(batch_idx_tuples) if i is not None) + batch_data = self.get_by_idxs(batch_idxs) + shared_batch_data = {} + for key, val in batch_data.items(): + if key.startswith('*'): + assert self.shared is not None + shared_key = key[1:] + shared_batch_data[shared_key] = [index(self.shared[shared_key], each) for each in val] + batch_data.update(shared_batch_data) + + batch_ds = DataSet(batch_data, self.data_type, shared=self.shared) + yield batch_idxs, batch_ds + + def get_multi_batches(self, batch_size, num_batches_per_step, num_steps=None, shuffle=False, cluster=False): + batch_size_per_step = batch_size * num_batches_per_step + batches = self.get_batches(batch_size_per_step, num_batches=num_steps, shuffle=shuffle, cluster=cluster) + multi_batches = (tuple(zip(grouper(idxs, batch_size, shorten=True, num_groups=num_batches_per_step), + data_set.divide(num_batches_per_step))) for idxs, data_set in batches) + return multi_batches + + def get_empty(self): + if isinstance(self.data, dict): + data = {key: [] for key in self.data} + elif isinstance(self.data, Data): + data = self.data.get_empty() + else: + raise Exception() + return DataSet(data, self.data_type, shared=self.shared) + + def __add__(self, other): + if isinstance(self.data, dict): + data = {key: val + other.data[key] for key, val in self.data.items()} + elif isinstance(self.data, Data): + data = self.data + other.data + else: + raise Exception() + + valid_idxs = list(self.valid_idxs) + [valid_idx + self.num_examples for valid_idx in other.valid_idxs] + return DataSet(data, self.data_type, shared=self.shared, valid_idxs=valid_idxs) + + def divide(self, integer): + batch_size = int(math.ceil(self.num_examples / integer)) + idxs_gen = grouper(self.valid_idxs, batch_size, shorten=True, num_groups=integer) + data_gen = (self.get_by_idxs(idxs) for idxs in idxs_gen) + ds_tuple = tuple(DataSet(data, self.data_type, shared=self.shared) for data in data_gen) + return ds_tuple + + +def load_metadata(config, data_type): + metadata_path = os.path.join(config.data_dir, "metadata_{}.json".format(data_type)) + with open(metadata_path, 'r') as fh: + metadata = json.load(fh) + for key, val in metadata.items(): + config.__setattr__(key, val) + return metadata + + +def read_data(config, data_type, ref, data_filter=None): + data_path = os.path.join(config.data_dir, "data_{}.json".format(data_type)) + shared_path = os.path.join(config.data_dir, "shared_{}.json".format(data_type)) + with open(data_path, 'r') as fh: + data = json.load(fh) + with open(shared_path, 'r') as fh: + shared = json.load(fh) + + num_examples = len(next(iter(data.values()))) + if data_filter is None: + valid_idxs = range(num_examples) + else: + mask = [] + keys = data.keys() + values = data.values() + for vals in zip(*values): + each = {key: val for key, val in zip(keys, vals)} + mask.append(data_filter(each, shared)) + valid_idxs = [idx for idx in range(len(mask)) if mask[idx]] + + print("Loaded {}/{} examples from {}".format(len(valid_idxs), num_examples, data_type)) + + shared_path = config.shared_path or os.path.join(config.out_dir, "shared.json") + if not ref: + word2vec_dict = shared['lower_word2vec'] if config.lower_word else shared['word2vec'] + word_counter = shared['lower_word_counter'] if config.lower_word else shared['word_counter'] + char_counter = shared['char_counter'] + if config.finetune: + shared['word2idx'] = {word: idx + 2 for idx, word in + enumerate(word for word, count in word_counter.items() + if count > config.word_count_th or (config.known_if_glove and word in word2vec_dict))} + else: + assert config.known_if_glove + assert config.use_glove_for_unk + shared['word2idx'] = {word: idx + 2 for idx, word in + enumerate(word for word, count in word_counter.items() + if count > config.word_count_th and word not in word2vec_dict)} + shared['char2idx'] = {char: idx + 2 for idx, char in + enumerate(char for char, count in char_counter.items() + if count > config.char_count_th)} + NULL = "-NULL-" + UNK = "-UNK-" + shared['word2idx'][NULL] = 0 + shared['word2idx'][UNK] = 1 + shared['char2idx'][NULL] = 0 + shared['char2idx'][UNK] = 1 + json.dump({'word2idx': shared['word2idx'], 'char2idx': shared['char2idx']}, open(shared_path, 'w')) + else: + new_shared = json.load(open(shared_path, 'r')) + for key, val in new_shared.items(): + shared[key] = val + + if config.use_glove_for_unk: + # create new word2idx and word2vec + word2vec_dict = shared['lower_word2vec'] if config.lower_word else shared['word2vec'] + new_word2idx_dict = {word: idx for idx, word in enumerate(word for word in word2vec_dict.keys() if word not in shared['word2idx'])} + shared['new_word2idx'] = new_word2idx_dict + offset = len(shared['word2idx']) + word2vec_dict = shared['lower_word2vec'] if config.lower_word else shared['word2vec'] + new_word2idx_dict = shared['new_word2idx'] + idx2vec_dict = {idx: word2vec_dict[word] for word, idx in new_word2idx_dict.items()} + # print("{}/{} unique words have corresponding glove vectors.".format(len(idx2vec_dict), len(word2idx_dict))) + new_emb_mat = np.array([idx2vec_dict[idx] for idx in range(len(idx2vec_dict))], dtype='float32') + shared['new_emb_mat'] = new_emb_mat + + data_set = DataSet(data, data_type, shared=shared, valid_idxs=valid_idxs) + return data_set + + +def get_squad_data_filter(config): + def data_filter(data_point, shared): + assert shared is not None + rx, rcx, q, cq, y = (data_point[key] for key in ('*x', '*cx', 'q', 'cq', 'y')) + x, cx = shared['x'], shared['cx'] + if len(q) > config.ques_size_th: + return False + + # x filter + xi = x[rx[0]][rx[1]] + if config.squash: + for start, stop in y: + stop_offset = sum(map(len, xi[:stop[0]])) + if stop_offset + stop[1] > config.para_size_th: + return False + return True + + if config.single: + for start, stop in y: + if start[0] != stop[0]: + return False + + if config.data_filter == 'max': + for start, stop in y: + if stop[0] >= config.num_sents_th: + return False + if start[0] != stop[0]: + return False + if stop[1] >= config.sent_size_th: + return False + elif config.data_filter == 'valid': + if len(xi) > config.num_sents_th: + return False + if any(len(xij) > config.sent_size_th for xij in xi): + return False + elif config.data_filter == 'semi': + """ + Only answer sentence needs to be valid. + """ + for start, stop in y: + if stop[0] >= config.num_sents_th: + return False + if start[0] != start[0]: + return False + if len(xi[start[0]]) > config.sent_size_th: + return False + else: + raise Exception() + + return True + return data_filter + + +def update_config(config, data_sets): + config.max_num_sents = 0 + config.max_sent_size = 0 + config.max_ques_size = 0 + config.max_word_size = 0 + config.max_para_size = 0 + for data_set in data_sets: + data = data_set.data + shared = data_set.shared + for idx in data_set.valid_idxs: + rx = data['*x'][idx] + q = data['q'][idx] + sents = shared['x'][rx[0]][rx[1]] + config.max_para_size = max(config.max_para_size, sum(map(len, sents))) + config.max_num_sents = max(config.max_num_sents, len(sents)) + config.max_sent_size = max(config.max_sent_size, max(map(len, sents))) + config.max_word_size = max(config.max_word_size, max(len(word) for sent in sents for word in sent)) + if len(q) > 0: + config.max_ques_size = max(config.max_ques_size, len(q)) + config.max_word_size = max(config.max_word_size, max(len(word) for word in q)) + + if config.mode == 'train': + config.max_num_sents = min(config.max_num_sents, config.num_sents_th) + config.max_sent_size = min(config.max_sent_size, config.sent_size_th) + config.max_para_size = min(config.max_para_size, config.para_size_th) + + config.max_word_size = min(config.max_word_size, config.word_size_th) + + config.char_vocab_size = len(data_sets[0].shared['char2idx']) + config.word_emb_size = len(next(iter(data_sets[0].shared['word2vec'].values()))) + config.word_vocab_size = len(data_sets[0].shared['word2idx']) + + if config.single: + config.max_num_sents = 1 + if config.squash: + config.max_sent_size = config.max_para_size + config.max_num_sents = 1 diff --git a/tensorflow/SQuAD/basic/run_ensemble.sh b/tensorflow/SQuAD/basic/run_ensemble.sh new file mode 100755 index 0000000..622d4f1 --- /dev/null +++ b/tensorflow/SQuAD/basic/run_ensemble.sh @@ -0,0 +1,29 @@ +#!/usr/bin/env bash +source_path=$1 +target_path=$2 +inter_dir="inter_ensemble" +root_dir="save" + +parg="" +marg="" +if [ "$3" = "debug" ] +then + parg="-d" + marg="--debug" +fi + +# Preprocess data +python3 -m squad.prepro --mode single --single_path $source_path $parg --target_dir $inter_dir --glove_dir . + +eargs="" +for num in 31 33 34 35 36 37 40 41 43 44 45 46; do + load_path="$root_dir/$num/save" + shared_path="$root_dir/$num/shared.json" + eval_path="$inter_dir/eval-$num.pklz" + eargs="$eargs $eval_path" + python3 -m basic.cli --data_dir $inter_dir --eval_path $eval_path --nodump_answer --load_path $load_path --shared_path $shared_path $marg --eval_num_batches 0 --mode forward --batch_size 1 --len_opt --cluster --cpu_opt --load_ema & +done +wait + +# Ensemble +python3 -m basic.ensemble --data_path $inter_dir/data_single.json --shared_path $inter_dir/shared_single.json -o $target_path $eargs diff --git a/tensorflow/SQuAD/basic/run_single.sh b/tensorflow/SQuAD/basic/run_single.sh new file mode 100755 index 0000000..06f7b7b --- /dev/null +++ b/tensorflow/SQuAD/basic/run_single.sh @@ -0,0 +1,27 @@ +#!/usr/bin/env bash +source_path=$1 +target_path=$2 +inter_dir="inter_single" +root_dir="save" + +parg="" +marg="" +if [ "$3" = "debug" ] +then + parg="-d" + marg="--debug" +fi + +# Preprocess data +python3 -m squad.prepro --mode single --single_path $source_path $parg --target_dir $inter_dir --glove_dir . + +num=37 +load_path="$root_dir/$num/save" +shared_path="$root_dir/$num/shared.json" +eval_path="$inter_dir/eval.pklz" +python3 -m basic.cli --data_dir $inter_dir --eval_path $eval_path --nodump_answer --load_path $load_path --shared_path $shared_path $marg --eval_num_batches 0 --mode forward --batch_size 1 --len_opt --cluster --cpu_opt --load_ema + +# Ensemble (for single run, just one input) +python3 -m basic.ensemble --data_path $inter_dir/data_single.json --shared_path $inter_dir/shared_single.json -o $target_path $eval_path + + diff --git a/tensorflow/SQuAD/basic/templates/visualizer.html b/tensorflow/SQuAD/basic/templates/visualizer.html new file mode 100644 index 0000000..8a4602d --- /dev/null +++ b/tensorflow/SQuAD/basic/templates/visualizer.html @@ -0,0 +1,76 @@ + + + + + {{ title }} + + + + + + +

{{ title }}

+ + + + + + + + + + {% for row in rows %} + + + + + + + + + {% endfor %} +
IDQuestionAnswersPredictedScoreParagraph
{{ row.id }} + {% for qj in row.ques %} + {{ qj }} + {% endfor %} + + {% for aa in row.a %} +
  • {{ aa }}
    • + {% endfor %} +     		{{ row.ap }}{{ row.score }} + + {% for xj, ypj, yp2j in zip(row.para, row.yp, row.yp2) %} + + {% set rowloop = loop %} + {% for xjk, ypjk in zip(xj, ypj) %} + + {% endfor %} + + + {% for xjk, yp2jk in zip(xj, yp2j) %} + + {% endfor %} + + {% endfor %} +
    + {% if row.y[0][0] == rowloop.index0 and row.y[0][1] <= loop.index0 <= row.y[1][1] %} + {{ xjk }} + {% else %} + {{ xjk }} + {% endif %} +
    -
    +
    + + \ No newline at end of file diff --git a/tensorflow/SQuAD/basic/trainer.py b/tensorflow/SQuAD/basic/trainer.py new file mode 100644 index 0000000..78d5e81 --- /dev/null +++ b/tensorflow/SQuAD/basic/trainer.py @@ -0,0 +1,73 @@ +import tensorflow as tf + +from basic.model import Model +from my.tensorflow import average_gradients + + +class Trainer(object): + def __init__(self, config, model): + assert isinstance(model, Model) + self.config = config + self.model = model + self.opt = tf.train.AdamOptimizer(config.init_lr) + self.loss = model.get_loss() + self.var_list = model.get_var_list() + self.global_step = model.get_global_step() + self.summary = model.summary + self.grads = self.opt.compute_gradients(self.loss, var_list=self.var_list) + self.train_op = self.opt.apply_gradients(self.grads, global_step=self.global_step) + + def get_train_op(self): + return self.train_op + + def step(self, sess, batch, get_summary=False): + assert isinstance(sess, tf.Session) + _, ds = batch + feed_dict = self.model.get_feed_dict(ds, True) + if get_summary: + loss, summary, train_op = \ + sess.run([self.loss, self.summary, self.train_op], feed_dict=feed_dict) + else: + loss, train_op = sess.run([self.loss, self.train_op], feed_dict=feed_dict) + summary = None + return loss, summary, train_op + + +class MultiGPUTrainer(object): + def __init__(self, config, models): + model = models[0] + assert isinstance(model, Model) + self.config = config + self.model = model + self.opt = tf.train.AdamOptimizer(config.init_lr) + self.var_list = model.get_var_list() + self.global_step = model.get_global_step() + self.summary = model.summary + self.models = models + losses = [] + grads_list = [] + for gpu_idx, model in enumerate(models): + with tf.name_scope("grads_{}".format(gpu_idx)), tf.device("/{}:{}".format(config.device_type, gpu_idx)): + loss = model.get_loss() + grads = self.opt.compute_gradients(loss, var_list=self.var_list) + losses.append(loss) + grads_list.append(grads) + + self.loss = tf.add_n(losses)/len(losses) + self.grads = average_gradients(grads_list) + self.train_op = self.opt.apply_gradients(self.grads, global_step=self.global_step) + + def step(self, sess, batches, get_summary=False): + assert isinstance(sess, tf.Session) + feed_dict = {} + for batch, model in zip(batches, self.models): + _, ds = batch + feed_dict.update(model.get_feed_dict(ds, True)) + + if get_summary: + loss, summary, train_op = \ + sess.run([self.loss, self.summary, self.train_op], feed_dict=feed_dict) + else: + loss, train_op = sess.run([self.loss, self.train_op], feed_dict=feed_dict) + summary = None + return loss, summary, train_op diff --git a/tensorflow/SQuAD/basic/visualizer.py b/tensorflow/SQuAD/basic/visualizer.py new file mode 100644 index 0000000..6ca4181 --- /dev/null +++ b/tensorflow/SQuAD/basic/visualizer.py @@ -0,0 +1,140 @@ +import shutil +from collections import OrderedDict +import http.server +import socketserver +import argparse +import json +import os +import numpy as np +from tqdm import tqdm +import pickle +import gzip + +from jinja2 import Environment, FileSystemLoader + +from squad.utils import get_best_span, get_best_span_wy + + +def bool_(string): + if string == 'True': + return True + elif string == 'False': + return False + else: + raise Exception() + +def get_args(): + parser = argparse.ArgumentParser() + parser.add_argument("--model_name", type=str, default='basic') + parser.add_argument("--data_type", type=str, default='dev') + parser.add_argument("--step", type=int, default=5000) + parser.add_argument("--template_name", type=str, default="visualizer.html") + parser.add_argument("--num_per_page", type=int, default=100) + parser.add_argument("--data_dir", type=str, default="data/squad") + parser.add_argument("--port", type=int, default=8000) + parser.add_argument("--host", type=str, default="0.0.0.0") + parser.add_argument("--open", type=str, default='False') + parser.add_argument("--run_id", type=str, default="0") + parser.add_argument("-w", "--wy", action='store_true') + + args = parser.parse_args() + return args + + +def _decode(decoder, sent): + return " ".join(decoder[idx] for idx in sent) + + +def accuracy2_visualizer(args): + model_name = args.model_name + data_type = args.data_type + num_per_page = args.num_per_page + data_dir = args.data_dir + run_id = args.run_id.zfill(2) + step = args.step + + eval_path =os.path.join("out", model_name, run_id, "eval", "{}-{}.pklz".format(data_type, str(step).zfill(6))) + print("loading {}".format(eval_path)) + eval_ = pickle.load(gzip.open(eval_path, 'r')) + + _id = 0 + html_dir = "/tmp/list_results%d" % _id + while os.path.exists(html_dir): + _id += 1 + html_dir = "/tmp/list_results%d" % _id + + if os.path.exists(html_dir): + shutil.rmtree(html_dir) + os.mkdir(html_dir) + + cur_dir = os.path.dirname(os.path.realpath(__file__)) + templates_dir = os.path.join(cur_dir, 'templates') + env = Environment(loader=FileSystemLoader(templates_dir)) + env.globals.update(zip=zip, reversed=reversed) + template = env.get_template(args.template_name) + + data_path = os.path.join(data_dir, "data_{}.json".format(data_type)) + shared_path = os.path.join(data_dir, "shared_{}.json".format(data_type)) + print("loading {}".format(data_path)) + data = json.load(open(data_path, 'r')) + print("loading {}".format(shared_path)) + shared = json.load(open(shared_path, 'r')) + + rows = [] + for i, (idx, yi, ypi, yp2i, wypi) in tqdm(enumerate(zip(*[eval_[key] for key in ('idxs', 'y', 'yp', 'yp2', 'wyp')])), total=len(eval_['idxs'])): + id_, q, rx, answers = (data[key][idx] for key in ('ids', 'q', '*x', 'answerss')) + x = shared['x'][rx[0]][rx[1]] + ques = [" ".join(q)] + para = [[word for word in sent] for sent in x] + span, score = get_best_span_wy(wypi, 0.5) if args.wy else get_best_span(ypi, yp2i) + ap = get_segment(para, span) + # score = "{:.3f}".format(ypi[span[0][0]][span[0][1]] * yp2i[span[1][0]][span[1][1]-1]) + + row = { + 'id': id_, + 'title': "Hello world!", + 'ques': ques, + 'para': para, + 'y': yi[0][0], + 'y2': yi[0][1], + 'yp': wypi if args.wy else ypi, + 'yp2': wypi if args.wy else yp2i, + 'a': answers, + 'ap': ap, + 'score': score + } + rows.append(row) + + if i % num_per_page == 0: + html_path = os.path.join(html_dir, "%s.html" % str(i).zfill(8)) + + if (i + 1) % num_per_page == 0 or (i + 1) == len(eval_['y']): + var_dict = {'title': "Accuracy Visualization", + 'rows': rows + } + with open(html_path, "wb") as f: + f.write(template.render(**var_dict).encode('UTF-8')) + rows = [] + + os.chdir(html_dir) + port = args.port + host = args.host + # Overriding to suppress log message + class MyHandler(http.server.SimpleHTTPRequestHandler): + def log_message(self, format, *args): + pass + handler = MyHandler + httpd = socketserver.TCPServer((host, port), handler) + if args.open == 'True': + os.system("open http://%s:%d" % (args.host, args.port)) + print("serving at %s:%d" % (host, port)) + httpd.serve_forever() + + +def get_segment(para, span): + return " ".join(para[span[0][0]][span[0][1]:span[1][1]]) + + +if __name__ == "__main__": + ARGS = get_args() + accuracy2_visualizer(ARGS) \ No newline at end of file diff --git a/tensorflow/SQuAD/basic_cnn/__init__.py b/tensorflow/SQuAD/basic_cnn/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/tensorflow/SQuAD/basic_cnn/cli.py b/tensorflow/SQuAD/basic_cnn/cli.py new file mode 100644 index 0000000..f3d78d1 --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/cli.py @@ -0,0 +1,103 @@ +import os + +import tensorflow as tf + +from basic_cnn.main import main as m + +flags = tf.app.flags + +flags.DEFINE_string("model_name", "basic_cnn", "Model name [basic]") +flags.DEFINE_string("data_dir", "data/cnn", "Data dir [data/cnn]") +flags.DEFINE_string("root_dir", "/Users/minjoons/data/cnn/questions", "root dir [~/data/cnn/questions]") +flags.DEFINE_string("run_id", "0", "Run ID [0]") +flags.DEFINE_string("out_base_dir", "out", "out base dir [out]") + +flags.DEFINE_integer("batch_size", 60, "Batch size [60]") +flags.DEFINE_float("init_lr", 0.5, "Initial learning rate [0.5]") +flags.DEFINE_integer("num_epochs", 50, "Total number of epochs for training [50]") +flags.DEFINE_integer("num_steps", 20000, "Number of steps [20000]") +flags.DEFINE_integer("eval_num_batches", 100, "eval num batches [100]") +flags.DEFINE_integer("load_step", 0, "load step [0]") +flags.DEFINE_integer("early_stop", 4, "early stop [4]") + +flags.DEFINE_string("mode", "test", "train | dev | test | forward [test]") +flags.DEFINE_boolean("load", True, "load saved data? [True]") +flags.DEFINE_boolean("progress", True, "Show progress? [True]") +flags.DEFINE_integer("log_period", 100, "Log period [100]") +flags.DEFINE_integer("eval_period", 1000, "Eval period [1000]") +flags.DEFINE_integer("save_period", 1000, "Save Period [1000]") +flags.DEFINE_float("decay", 0.9, "Exponential moving average decay [0.9]") + +flags.DEFINE_boolean("draft", False, "Draft for quick testing? [False]") + +flags.DEFINE_integer("hidden_size", 100, "Hidden size [100]") +flags.DEFINE_integer("char_out_size", 100, "Char out size [100]") +flags.DEFINE_float("input_keep_prob", 0.8, "Input keep prob [0.8]") +flags.DEFINE_integer("char_emb_size", 8, "Char emb size [8]") +flags.DEFINE_integer("char_filter_height", 5, "Char filter height [5]") +flags.DEFINE_float("wd", 0.0, "Weight decay [0.0]") +flags.DEFINE_bool("lower_word", True, "lower word [True]") +flags.DEFINE_bool("dump_eval", False, "dump eval? [True]") +flags.DEFINE_bool("dump_answer", True, "dump answer? [True]") +flags.DEFINE_string("model", "2", "config 1 |2 [2]") +flags.DEFINE_bool("squash", False, "squash the sentences into one? [False]") +flags.DEFINE_bool("single", False, "supervise only the answer sentence? [False]") + +flags.DEFINE_integer("word_count_th", 10, "word count th [100]") +flags.DEFINE_integer("char_count_th", 50, "char count th [500]") +flags.DEFINE_integer("sent_size_th", 60, "sent size th [64]") +flags.DEFINE_integer("num_sents_th", 200, "num sents th [8]") +flags.DEFINE_integer("ques_size_th", 30, "ques size th [32]") +flags.DEFINE_integer("word_size_th", 16, "word size th [16]") +flags.DEFINE_integer("para_size_th", 256, "para size th [256]") + +flags.DEFINE_bool("swap_memory", True, "swap memory? [True]") +flags.DEFINE_string("data_filter", "max", "max | valid | semi [max]") +flags.DEFINE_bool("finetune", False, "finetune? [False]") +flags.DEFINE_bool("feed_gt", False, "feed gt prev token during training [False]") +flags.DEFINE_bool("feed_hard", False, "feed hard argmax prev token during testing [False]") +flags.DEFINE_bool("use_glove_for_unk", True, "use glove for unk [False]") +flags.DEFINE_bool("known_if_glove", True, "consider as known if present in glove [False]") +flags.DEFINE_bool("eval", True, "eval? [True]") +flags.DEFINE_integer("highway_num_layers", 2, "highway num layers [2]") +flags.DEFINE_bool("use_word_emb", True, "use word embedding? [True]") + +flags.DEFINE_string("forward_name", "single", "Forward name [single]") +flags.DEFINE_string("answer_path", "", "Answer path []") +flags.DEFINE_string("load_path", "", "Load path []") +flags.DEFINE_string("shared_path", "", "Shared path []") +flags.DEFINE_string("device", "/cpu:0", "default device [/cpu:0]") +flags.DEFINE_integer("num_gpus", 1, "num of gpus [1]") + +flags.DEFINE_string("out_channel_dims", "100", "Out channel dims, separated by commas [100]") +flags.DEFINE_string("filter_heights", "5", "Filter heights, separated by commas [5]") + +flags.DEFINE_bool("share_cnn_weights", True, "Share CNN weights [False]") +flags.DEFINE_bool("share_lstm_weights", True, "Share LSTM weights [True]") +flags.DEFINE_bool("two_prepro_layers", False, "Use two layers for preprocessing? [False]") +flags.DEFINE_bool("aug_att", False, "Augment attention layers with more features? [False]") +flags.DEFINE_integer("max_to_keep", 20, "Max recent saves to keep [20]") +flags.DEFINE_bool("vis", False, "output visualization numbers? [False]") +flags.DEFINE_bool("dump_pickle", True, "Dump pickle instead of json? [True]") +flags.DEFINE_float("keep_prob", 1.0, "keep prob [1.0]") +flags.DEFINE_string("prev_mode", "a", "prev mode gy | y | a [a]") +flags.DEFINE_string("logit_func", "tri_linear", "logit func [tri_linear]") +flags.DEFINE_bool("sh", False, "use superhighway [False]") +flags.DEFINE_string("answer_func", "linear", "answer logit func [linear]") +flags.DEFINE_bool("cluster", False, "Cluster data for faster training [False]") +flags.DEFINE_bool("len_opt", False, "Length optimization? [False]") +flags.DEFINE_string("sh_logit_func", "tri_linear", "sh logit func [tri_linear]") +flags.DEFINE_float("filter_ratio", 1.0, "filter ratio [1.0]") +flags.DEFINE_bool("bi", False, "bi-directional attention? [False]") +flags.DEFINE_integer("width", 5, "width around entity [5]") + + +def main(_): + config = flags.FLAGS + + config.out_dir = os.path.join(config.out_base_dir, config.model_name, str(config.run_id).zfill(2)) + + m(config) + +if __name__ == "__main__": + tf.app.run() diff --git a/tensorflow/SQuAD/basic_cnn/evaluator.py b/tensorflow/SQuAD/basic_cnn/evaluator.py new file mode 100644 index 0000000..c978b68 --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/evaluator.py @@ -0,0 +1,494 @@ +import itertools +from collections import defaultdict + +import numpy as np +import tensorflow as tf +import os + +from basic_cnn.read_data import DataSet +from my.nltk_utils import span_f1 +from my.tensorflow import padded_reshape +from my.utils import argmax + + +class Evaluation(object): + def __init__(self, data_type, global_step, idxs, yp, tensor_dict=None): + self.data_type = data_type + self.global_step = global_step + self.idxs = idxs + self.yp = yp + self.num_examples = len(yp) + self.tensor_dict = None + self.dict = {'data_type': data_type, + 'global_step': global_step, + 'yp': yp, + 'idxs': idxs, + 'num_examples': self.num_examples} + if tensor_dict is not None: + self.tensor_dict = {key: val.tolist() for key, val in tensor_dict.items()} + for key, val in self.tensor_dict.items(): + self.dict[key] = val + self.summaries = None + + def __repr__(self): + return "{} step {}".format(self.data_type, self.global_step) + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_yp = self.yp + other.yp + new_idxs = self.idxs + other.idxs + new_tensor_dict = None + if self.tensor_dict is not None: + new_tensor_dict = {key: val + other.tensor_dict[key] for key, val in self.tensor_dict.items()} + return Evaluation(self.data_type, self.global_step, new_idxs, new_yp, tensor_dict=new_tensor_dict) + + def __radd__(self, other): + return self.__add__(other) + + +class LabeledEvaluation(Evaluation): + def __init__(self, data_type, global_step, idxs, yp, y, id2answer_dict, tensor_dict=None): + super(LabeledEvaluation, self).__init__(data_type, global_step, idxs, yp, tensor_dict=tensor_dict) + self.y = y + self.dict['y'] = y + self.id2answer_dict = id2answer_dict + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_yp = self.yp + other.yp + new_y = self.y + other.y + new_idxs = self.idxs + other.idxs + new_id2answer_dict = dict(list(self.id2answer_dict.items()) + list(other.id2answer_dict.items())) + new_id2score_dict = dict(list(self.id2answer_dict['scores'].items()) + list(other.id2answer_dict['scores'].items())) + new_id2answer_dict['scores'] = new_id2score_dict + if self.tensor_dict is not None: + new_tensor_dict = {key: np.concatenate((val, other.tensor_dict[key]), axis=0) for key, val in self.tensor_dict.items()} + return LabeledEvaluation(self.data_type, self.global_step, new_idxs, new_yp, new_y, new_id2answer_dict, tensor_dict=new_tensor_dict) + + +class AccuracyEvaluation(LabeledEvaluation): + def __init__(self, data_type, global_step, idxs, yp, y, id2answer_dict, correct, loss, tensor_dict=None): + super(AccuracyEvaluation, self).__init__(data_type, global_step, idxs, yp, y, id2answer_dict, tensor_dict=tensor_dict) + self.loss = loss + self.correct = correct + self.id2answer_dict = id2answer_dict + self.acc = sum(correct) / len(correct) + self.dict['loss'] = loss + self.dict['correct'] = correct + self.dict['acc'] = self.acc + loss_summary = tf.Summary(value=[tf.Summary.Value(tag='{}/loss'.format(data_type), simple_value=self.loss)]) + acc_summary = tf.Summary(value=[tf.Summary.Value(tag='{}/acc'.format(data_type), simple_value=self.acc)]) + self.summaries = [loss_summary, acc_summary] + + def __repr__(self): + return "{} step {}: accuracy={}={}/{}, loss={}".format(self.data_type, self.global_step, self.acc, + sum(self.correct), self.num_examples, self.loss) + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_idxs = self.idxs + other.idxs + new_yp = self.yp + other.yp + new_y = self.y + other.y + new_correct = self.correct + other.correct + new_loss = (self.loss * self.num_examples + other.loss * other.num_examples) / len(new_correct) + new_id2answer_dict = dict(list(self.id2answer_dict.items()) + list(other.id2answer_dict.items())) + new_id2score_dict = dict(list(self.id2answer_dict['scores'].items()) + list(other.id2answer_dict['scores'].items())) + new_id2answer_dict['scores'] = new_id2score_dict + new_tensor_dict = None + if self.tensor_dict is not None: + new_tensor_dict = {key: np.concatenate((val, other.tensor_dict[key]), axis=0) for key, val in self.tensor_dict.items()} + return AccuracyEvaluation(self.data_type, self.global_step, new_idxs, new_yp, new_y, new_id2answer_dict, new_correct, new_loss, tensor_dict=new_tensor_dict) + + +class Evaluator(object): + def __init__(self, config, model, tensor_dict=None): + self.config = config + self.model = model + self.global_step = model.global_step + self.yp = model.yp + self.tensor_dict = {} if tensor_dict is None else tensor_dict + + def get_evaluation(self, sess, batch): + idxs, data_set = batch + feed_dict = self.model.get_feed_dict(data_set, False, supervised=False) + global_step, yp, vals = sess.run([self.global_step, self.yp, list(self.tensor_dict.values())], feed_dict=feed_dict) + yp = yp[:data_set.num_examples] + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + e = Evaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), tensor_dict=tensor_dict) + return e + + def get_evaluation_from_batches(self, sess, batches): + e = sum(self.get_evaluation(sess, batch) for batch in batches) + return e + + +class LabeledEvaluator(Evaluator): + def __init__(self, config, model, tensor_dict=None): + super(LabeledEvaluator, self).__init__(config, model, tensor_dict=tensor_dict) + self.y = model.y + + def get_evaluation(self, sess, batch): + idxs, data_set = batch + feed_dict = self.model.get_feed_dict(data_set, False, supervised=False) + global_step, yp, vals = sess.run([self.global_step, self.yp, list(self.tensor_dict.values())], feed_dict=feed_dict) + yp = yp[:data_set.num_examples] + y = feed_dict[self.y] + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + e = LabeledEvaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), y.tolist(), tensor_dict=tensor_dict) + return e + + +class AccuracyEvaluator(LabeledEvaluator): + def __init__(self, config, model, tensor_dict=None): + super(AccuracyEvaluator, self).__init__(config, model, tensor_dict=tensor_dict) + self.loss = model.loss + + def get_evaluation(self, sess, batch): + idxs, data_set = self._split_batch(batch) + assert isinstance(data_set, DataSet) + feed_dict = self._get_feed_dict(batch) + y = data_set.data['y'] + global_step, yp, loss, vals = sess.run([self.global_step, self.yp, self.loss, list(self.tensor_dict.values())], feed_dict=feed_dict) + yp = yp[:data_set.num_examples] + correct, probs, preds = zip(*[self.__class__.compare(data_set.get_one(idx), ypi) for idx, ypi in zip(data_set.valid_idxs, yp)]) + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + ids = data_set.data['ids'] + id2score_dict = {id_: prob for id_, prob in zip(ids, probs)} + id2answer_dict = {id_: pred for id_, pred in zip(ids, preds)} + id2answer_dict['scores'] = id2score_dict + e = AccuracyEvaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), y, id2answer_dict, correct, float(loss), tensor_dict=tensor_dict) + return e + + @staticmethod + def compare(data, ypi): + prob = float(np.max(ypi)) + yi = data['y'] + for start, stop in yi: + if start == int(np.argmax(ypi)): + return True, prob, " " + return False, prob, " " + + def _split_batch(self, batch): + return batch + + def _get_feed_dict(self, batch): + return self.model.get_feed_dict(batch[1], False) + + +class CNNAccuracyEvaluator(AccuracyEvaluator): + @staticmethod + def compare(data, ypi): + # ypi: [N, M, JX] numbers + yi = data['y'][0] # entity + xi = data['x'][0] # [N, M, JX] words + dist = defaultdict(int) + for ypij, xij in zip(ypi, xi): + for ypijk, xijk in zip(ypij, xij): + if xijk.startswith("@"): + dist[xijk] += ypijk + pred, prob = max(dist.items(), key=lambda item: item[1]) + assert pred.startswith("@") + assert yi.startswith("@") + return pred == yi, prob, pred + + +class AccuracyEvaluator2(AccuracyEvaluator): + @staticmethod + def compare(yi, ypi): + for start, stop in yi: + para_start = int(np.argmax(np.max(ypi, 1))) + sent_start = int(np.argmax(ypi[para_start])) + if tuple(start) == (para_start, sent_start): + return True + return False + + +class ForwardEvaluation(Evaluation): + def __init__(self, data_type, global_step, idxs, yp, yp2, loss, id2answer_dict, tensor_dict=None): + super(ForwardEvaluation, self).__init__(data_type, global_step, idxs, yp, tensor_dict=tensor_dict) + self.yp2 = yp2 + self.loss = loss + self.dict['loss'] = loss + self.dict['yp2'] = yp2 + self.id2answer_dict = id2answer_dict + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_idxs = self.idxs + other.idxs + new_yp = self.yp + other.yp + new_yp2 = self.yp2 + other.yp2 + new_loss = (self.loss * self.num_examples + other.loss * other.num_examples) / len(new_yp) + new_id2answer_dict = dict(list(self.id2answer_dict.items()) + list(other.id2answer_dict.items())) + if self.tensor_dict is not None: + new_tensor_dict = {key: np.concatenate((val, other.tensor_dict[key]), axis=0) for key, val in self.tensor_dict.items()} + return ForwardEvaluation(self.data_type, self.global_step, new_idxs, new_yp, new_yp2, new_loss, new_id2answer_dict, tensor_dict=new_tensor_dict) + + def __repr__(self): + return "{} step {}: loss={:.4f}".format(self.data_type, self.global_step, self.loss) + + +class F1Evaluation(AccuracyEvaluation): + def __init__(self, data_type, global_step, idxs, yp, yp2, y, correct, loss, f1s, id2answer_dict, tensor_dict=None): + super(F1Evaluation, self).__init__(data_type, global_step, idxs, yp, y, correct, loss, tensor_dict=tensor_dict) + self.yp2 = yp2 + self.f1s = f1s + self.f1 = float(np.mean(f1s)) + self.dict['yp2'] = yp2 + self.dict['f1s'] = f1s + self.dict['f1'] = self.f1 + self.id2answer_dict = id2answer_dict + f1_summary = tf.Summary(value=[tf.Summary.Value(tag='{}/f1'.format(data_type), simple_value=self.f1)]) + self.summaries.append(f1_summary) + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_idxs = self.idxs + other.idxs + new_yp = self.yp + other.yp + new_yp2 = self.yp2 + other.yp2 + new_y = self.y + other.y + new_correct = self.correct + other.correct + new_f1s = self.f1s + other.f1s + new_loss = (self.loss * self.num_examples + other.loss * other.num_examples) / len(new_correct) + new_id2answer_dict = dict(list(self.id2answer_dict.items()) + list(other.id2answer_dict.items())) + return F1Evaluation(self.data_type, self.global_step, new_idxs, new_yp, new_yp2, new_y, new_correct, new_loss, new_f1s, new_id2answer_dict) + + def __repr__(self): + return "{} step {}: accuracy={:.4f}, f1={:.4f}, loss={:.4f}".format(self.data_type, self.global_step, self.acc, self.f1, self.loss) + + +class F1Evaluator(LabeledEvaluator): + def __init__(self, config, model, tensor_dict=None): + super(F1Evaluator, self).__init__(config, model, tensor_dict=tensor_dict) + self.yp2 = model.yp2 + self.loss = model.loss + + def get_evaluation(self, sess, batch): + idxs, data_set = self._split_batch(batch) + assert isinstance(data_set, DataSet) + feed_dict = self._get_feed_dict(batch) + global_step, yp, yp2, loss, vals = sess.run([self.global_step, self.yp, self.yp2, self.loss, list(self.tensor_dict.values())], feed_dict=feed_dict) + y = data_set.data['y'] + if self.config.squash: + new_y = [] + for xi, yi in zip(data_set.data['x'], y): + new_yi = [] + for start, stop in yi: + start_offset = sum(map(len, xi[:start[0]])) + stop_offset = sum(map(len, xi[:stop[0]])) + new_start = 0, start_offset + start[1] + new_stop = 0, stop_offset + stop[1] + new_yi.append((new_start, new_stop)) + new_y.append(new_yi) + y = new_y + if self.config.single: + new_y = [] + for yi in y: + new_yi = [] + for start, stop in yi: + new_start = 0, start[1] + new_stop = 0, stop[1] + new_yi.append((new_start, new_stop)) + new_y.append(new_yi) + y = new_y + + yp, yp2 = yp[:data_set.num_examples], yp2[:data_set.num_examples] + spans = [get_best_span(ypi, yp2i) for ypi, yp2i in zip(yp, yp2)] + + def _get(xi, span): + if len(xi) <= span[0][0]: + return [""] + if len(xi[span[0][0]]) <= span[1][1]: + return [""] + return xi[span[0][0]][span[0][1]:span[1][1]] + + id2answer_dict = {id_: " ".join(_get(xi, span)) + for id_, xi, span in zip(data_set.data['ids'], data_set.data['x'], spans)} + correct = [self.__class__.compare2(yi, span) for yi, span in zip(y, spans)] + f1s = [self.__class__.span_f1(yi, span) for yi, span in zip(y, spans)] + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + e = F1Evaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), yp2.tolist(), y, + correct, float(loss), f1s, id2answer_dict, tensor_dict=tensor_dict) + return e + + def _split_batch(self, batch): + return batch + + def _get_feed_dict(self, batch): + return self.model.get_feed_dict(batch[1], False) + + @staticmethod + def compare(yi, ypi, yp2i): + for start, stop in yi: + aypi = argmax(ypi) + mask = np.zeros(yp2i.shape) + mask[aypi[0], aypi[1]:] = np.ones([yp2i.shape[1] - aypi[1]]) + if tuple(start) == aypi and (stop[0], stop[1]-1) == argmax(yp2i * mask): + return True + return False + + @staticmethod + def compare2(yi, span): + for start, stop in yi: + if tuple(start) == span[0] and tuple(stop) == span[1]: + return True + return False + + @staticmethod + def span_f1(yi, span): + max_f1 = 0 + for start, stop in yi: + if start[0] == span[0][0]: + true_span = start[1], stop[1] + pred_span = span[0][1], span[1][1] + f1 = span_f1(true_span, pred_span) + max_f1 = max(f1, max_f1) + return max_f1 + + +class MultiGPUF1Evaluator(F1Evaluator): + def __init__(self, config, models, tensor_dict=None): + super(MultiGPUF1Evaluator, self).__init__(config, models[0], tensor_dict=tensor_dict) + self.models = models + with tf.name_scope("eval_concat"): + N, M, JX = config.batch_size, config.max_num_sents, config.max_sent_size + self.yp = tf.concat(axis=0, values=[padded_reshape(model.yp, [N, M, JX]) for model in models]) + self.yp2 = tf.concat(axis=0, values=[padded_reshape(model.yp2, [N, M, JX]) for model in models]) + self.loss = tf.add_n([model.loss for model in models])/len(models) + + def _split_batch(self, batches): + idxs_list, data_sets = zip(*batches) + idxs = sum(idxs_list, ()) + data_set = sum(data_sets, data_sets[0].get_empty()) + return idxs, data_set + + def _get_feed_dict(self, batches): + feed_dict = {} + for model, (_, data_set) in zip(self.models, batches): + feed_dict.update(model.get_feed_dict(data_set, False)) + return feed_dict + + +class MultiGPUCNNAccuracyEvaluator(CNNAccuracyEvaluator): + def __init__(self, config, models, tensor_dict=None): + super(MultiGPUCNNAccuracyEvaluator, self).__init__(config, models[0], tensor_dict=tensor_dict) + self.models = models + with tf.name_scope("eval_concat"): + N, M, JX = config.batch_size, config.max_num_sents, config.max_sent_size + self.yp = tf.concat(axis=0, values=[padded_reshape(model.yp, [N, M, JX]) for model in models]) + self.loss = tf.add_n([model.loss for model in models])/len(models) + + def _split_batch(self, batches): + idxs_list, data_sets = zip(*batches) + idxs = sum(idxs_list, ()) + data_set = sum(data_sets, data_sets[0].get_empty()) + return idxs, data_set + + def _get_feed_dict(self, batches): + feed_dict = {} + for model, (_, data_set) in zip(self.models, batches): + feed_dict.update(model.get_feed_dict(data_set, False)) + return feed_dict + + +class ForwardEvaluator(Evaluator): + def __init__(self, config, model, tensor_dict=None): + super(ForwardEvaluator, self).__init__(config, model, tensor_dict=tensor_dict) + self.yp2 = model.yp2 + self.loss = model.loss + + def get_evaluation(self, sess, batch): + idxs, data_set = batch + assert isinstance(data_set, DataSet) + feed_dict = self.model.get_feed_dict(data_set, False) + global_step, yp, yp2, loss, vals = sess.run([self.global_step, self.yp, self.yp2, self.loss, list(self.tensor_dict.values())], feed_dict=feed_dict) + + yp, yp2 = yp[:data_set.num_examples], yp2[:data_set.num_examples] + spans = [get_best_span(ypi, yp2i) for ypi, yp2i in zip(yp, yp2)] + + def _get(xi, span): + if len(xi) <= span[0][0]: + return [""] + if len(xi[span[0][0]]) <= span[1][1]: + return [""] + return xi[span[0][0]][span[0][1]:span[1][1]] + + id2answer_dict = {id_: " ".join(_get(xi, span)) + for id_, xi, span in zip(data_set.data['ids'], data_set.data['x'], spans)} + tensor_dict = dict(zip(self.tensor_dict.keys(), vals)) + e = ForwardEvaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), yp2.tolist(), float(loss), id2answer_dict, tensor_dict=tensor_dict) + return e + + @staticmethod + def compare(yi, ypi, yp2i): + for start, stop in yi: + aypi = argmax(ypi) + mask = np.zeros(yp2i.shape) + mask[aypi[0], aypi[1]:] = np.ones([yp2i.shape[1] - aypi[1]]) + if tuple(start) == aypi and (stop[0], stop[1]-1) == argmax(yp2i * mask): + return True + return False + + @staticmethod + def compare2(yi, span): + for start, stop in yi: + if tuple(start) == span[0] and tuple(stop) == span[1]: + return True + return False + + @staticmethod + def span_f1(yi, span): + max_f1 = 0 + for start, stop in yi: + if start[0] == span[0][0]: + true_span = start[1], stop[1] + pred_span = span[0][1], span[1][1] + f1 = span_f1(true_span, pred_span) + max_f1 = max(f1, max_f1) + return max_f1 + + +def get_best_span(ypi, yp2i): + + max_val = 0 + best_word_span = (0, 1) + best_sent_idx = 0 + for f, (ypif, yp2if) in enumerate(zip(ypi, yp2i)): + argmax_j1 = 0 + for j in range(len(ypif)): + val1 = ypif[argmax_j1] + if val1 < ypif[j]: + val1 = ypif[j] + argmax_j1 = j + + val2 = yp2if[j] + if val1 * val2 > max_val: + best_word_span = (argmax_j1, j) + best_sent_idx = f + max_val = val1 * val2 + return (best_sent_idx, best_word_span[0]), (best_sent_idx, best_word_span[1] + 1) + + +def get_span_score_pairs(ypi, yp2i): + span_score_pairs = [] + for f, (ypif, yp2if) in enumerate(zip(ypi, yp2i)): + for j in range(len(ypif)): + for k in range(j, len(yp2if)): + span = ((f, j), (f, k+1)) + score = ypif[j] * yp2if[k] + span_score_pairs.append((span, score)) + return span_score_pairs diff --git a/tensorflow/SQuAD/basic_cnn/graph_handler.py b/tensorflow/SQuAD/basic_cnn/graph_handler.py new file mode 100644 index 0000000..8bd1098 --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/graph_handler.py @@ -0,0 +1,70 @@ +import gzip +import json +from json import encoder +import os + +import tensorflow as tf + +from basic_cnn.evaluator import Evaluation, F1Evaluation +from my.utils import short_floats + +import pickle + + +class GraphHandler(object): + def __init__(self, config): + self.config = config + self.saver = tf.train.Saver(max_to_keep=config.max_to_keep) + self.writer = None + self.save_path = os.path.join(config.save_dir, config.model_name) + + def initialize(self, sess): + if self.config.load: + self._load(sess) + else: + sess.run(tf.global_variables_initializer()) + + if self.config.mode == 'train': + self.writer = tf.summary.FileWriter(self.config.log_dir, graph=tf.get_default_graph()) + + def save(self, sess, global_step=None): + self.saver.save(sess, self.save_path, global_step=global_step) + + def _load(self, sess): + config = self.config + if config.load_path: + save_path = config.load_path + elif config.load_step > 0: + save_path = os.path.join(config.save_dir, "{}-{}".format(config.model_name, config.load_step)) + else: + save_dir = config.save_dir + checkpoint = tf.train.get_checkpoint_state(save_dir) + assert checkpoint is not None, "cannot load checkpoint at {}".format(save_dir) + save_path = checkpoint.model_checkpoint_path + print("Loading saved model from {}".format(save_path)) + self.saver.restore(sess, save_path) + + def add_summary(self, summary, global_step): + self.writer.add_summary(summary, global_step) + + def add_summaries(self, summaries, global_step): + for summary in summaries: + self.add_summary(summary, global_step) + + def dump_eval(self, e, precision=2, path=None): + assert isinstance(e, Evaluation) + if self.config.dump_pickle: + path = path or os.path.join(self.config.eval_dir, "{}-{}.pklz".format(e.data_type, str(e.global_step).zfill(6))) + with gzip.open(path, 'wb', compresslevel=3) as fh: + pickle.dump(e.dict, fh) + else: + path = path or os.path.join(self.config.eval_dir, "{}-{}.json".format(e.data_type, str(e.global_step).zfill(6))) + with open(path, 'w') as fh: + json.dump(short_floats(e.dict, precision), fh) + + def dump_answer(self, e, path=None): + assert isinstance(e, Evaluation) + path = path or os.path.join(self.config.answer_dir, "{}-{}.json".format(e.data_type, str(e.global_step).zfill(6))) + with open(path, 'w') as fh: + json.dump(e.id2answer_dict, fh) + diff --git a/tensorflow/SQuAD/basic_cnn/main.py b/tensorflow/SQuAD/basic_cnn/main.py new file mode 100644 index 0000000..7d05492 --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/main.py @@ -0,0 +1,238 @@ +import argparse +import json +import math +import os +import shutil +from pprint import pprint + +import tensorflow as tf +from tqdm import tqdm +import numpy as np + +from basic_cnn.evaluator import F1Evaluator, Evaluator, ForwardEvaluator, MultiGPUF1Evaluator, CNNAccuracyEvaluator, \ + MultiGPUCNNAccuracyEvaluator +from basic_cnn.graph_handler import GraphHandler +from basic_cnn.model import Model, get_multi_gpu_models +from basic_cnn.trainer import Trainer, MultiGPUTrainer + +from basic_cnn.read_data import read_data, get_cnn_data_filter, update_config + + +def main(config): + set_dirs(config) + with tf.device(config.device): + if config.mode == 'train': + _train(config) + elif config.mode == 'test' or config.mode == 'dev': + _test(config) + elif config.mode == 'forward': + _forward(config) + else: + raise ValueError("invalid value for 'mode': {}".format(config.mode)) + + +def _config_draft(config): + if config.draft: + config.num_steps = 2 + config.eval_period = 1 + config.log_period = 1 + config.save_period = 1 + config.eval_num_batches = 1 + + +def _train(config): + # load_metadata(config, 'train') # this updates the config file according to metadata file + + data_filter = get_cnn_data_filter(config) + train_data = read_data(config, 'train', config.load, data_filter=data_filter) + dev_data = read_data(config, 'dev', True, data_filter=data_filter) + # test_data = read_data(config, 'test', True, data_filter=data_filter) + update_config(config, [train_data, dev_data]) + + _config_draft(config) + + word2vec_dict = train_data.shared['lower_word2vec'] if config.lower_word else train_data.shared['word2vec'] + word2idx_dict = train_data.shared['word2idx'] + idx2vec_dict = {word2idx_dict[word]: vec for word, vec in word2vec_dict.items() if word in word2idx_dict} + print("{}/{} unique words have corresponding glove vectors.".format(len(idx2vec_dict), len(word2idx_dict))) + emb_mat = np.array([idx2vec_dict[idx] if idx in idx2vec_dict + else np.random.multivariate_normal(np.zeros(config.word_emb_size), np.eye(config.word_emb_size)) + for idx in range(config.word_vocab_size)]) + config.emb_mat = emb_mat + + # construct model graph and variables (using default graph) + pprint(config.__flags, indent=2) + # model = Model(config) + models = get_multi_gpu_models(config) + model = models[0] + trainer = MultiGPUTrainer(config, models) + evaluator = MultiGPUCNNAccuracyEvaluator(config, models, tensor_dict=model.tensor_dict if config.vis else None) + graph_handler = GraphHandler(config) # controls all tensors and variables in the graph, including loading /saving + + # Variables + sess = tf.Session(config=tf.ConfigProto(allow_soft_placement=True)) + graph_handler.initialize(sess) + + # begin training + print(train_data.num_examples) + num_steps = config.num_steps or int(math.ceil(train_data.num_examples / (config.batch_size * config.num_gpus))) * config.num_epochs + global_step = 0 + for batches in tqdm(train_data.get_multi_batches(config.batch_size, config.num_gpus, + num_steps=num_steps, shuffle=True, cluster=config.cluster), total=num_steps): + global_step = sess.run(model.global_step) + 1 # +1 because all calculations are done after step + get_summary = global_step % config.log_period == 0 + loss, summary, train_op = trainer.step(sess, batches, get_summary=get_summary) + if get_summary: + graph_handler.add_summary(summary, global_step) + + # occasional saving + if global_step % config.save_period == 0: + graph_handler.save(sess, global_step=global_step) + + if not config.eval: + continue + # Occasional evaluation + if global_step % config.eval_period == 0: + num_steps = math.ceil(dev_data.num_examples / (config.batch_size * config.num_gpus)) + if 0 < config.eval_num_batches < num_steps: + num_steps = config.eval_num_batches + e_train = evaluator.get_evaluation_from_batches( + sess, tqdm(train_data.get_multi_batches(config.batch_size, config.num_gpus, num_steps=num_steps), total=num_steps) + ) + graph_handler.add_summaries(e_train.summaries, global_step) + e_dev = evaluator.get_evaluation_from_batches( + sess, tqdm(dev_data.get_multi_batches(config.batch_size, config.num_gpus, num_steps=num_steps), total=num_steps)) + graph_handler.add_summaries(e_dev.summaries, global_step) + + if config.dump_eval: + graph_handler.dump_eval(e_dev) + if config.dump_answer: + graph_handler.dump_answer(e_dev) + if global_step % config.save_period != 0: + graph_handler.save(sess, global_step=global_step) + + +def _test(config): + assert config.load + test_data = read_data(config, config.mode, True) + update_config(config, [test_data]) + + _config_draft(config) + + if config.use_glove_for_unk: + word2vec_dict = test_data.shared['lower_word2vec'] if config.lower_word else test_data.shared['word2vec'] + new_word2idx_dict = test_data.shared['new_word2idx'] + idx2vec_dict = {idx: word2vec_dict[word] for word, idx in new_word2idx_dict.items()} + # print("{}/{} unique words have corresponding glove vectors.".format(len(idx2vec_dict), len(word2idx_dict))) + new_emb_mat = np.array([idx2vec_dict[idx] for idx in range(len(idx2vec_dict))], dtype='float32') + config.new_emb_mat = new_emb_mat + + pprint(config.__flags, indent=2) + models = get_multi_gpu_models(config) + evaluator = MultiGPUCNNAccuracyEvaluator(config, models, tensor_dict=models[0].tensor_dict if config.vis else None) + graph_handler = GraphHandler(config) # controls all tensors and variables in the graph, including loading /saving + + sess = tf.Session(config=tf.ConfigProto(allow_soft_placement=True)) + graph_handler.initialize(sess) + num_steps = math.ceil(test_data.num_examples / (config.batch_size * config.num_gpus)) + if 0 < config.eval_num_batches < num_steps: + num_steps = config.eval_num_batches + + e = None + for multi_batch in tqdm(test_data.get_multi_batches(config.batch_size, config.num_gpus, num_steps=num_steps, cluster=config.cluster), total=num_steps): + ei = evaluator.get_evaluation(sess, multi_batch) + e = ei if e is None else e + ei + if config.vis: + eval_subdir = os.path.join(config.eval_dir, "{}-{}".format(ei.data_type, str(ei.global_step).zfill(6))) + if not os.path.exists(eval_subdir): + os.mkdir(eval_subdir) + path = os.path.join(eval_subdir, str(ei.idxs[0]).zfill(8)) + graph_handler.dump_eval(ei, path=path) + + print(e) + if config.dump_answer: + print("dumping answer ...") + graph_handler.dump_answer(e) + if config.dump_eval: + print("dumping eval ...") + graph_handler.dump_eval(e) + + +def _forward(config): + assert config.load + test_data = read_data(config, config.forward_name, True) + update_config(config, [test_data]) + + _config_draft(config) + + if config.use_glove_for_unk: + word2vec_dict = test_data.shared['lower_word2vec'] if config.lower_word else test_data.shared['word2vec'] + new_word2idx_dict = test_data.shared['new_word2idx'] + idx2vec_dict = {idx: word2vec_dict[word] for word, idx in new_word2idx_dict.items()} + # print("{}/{} unique words have corresponding glove vectors.".format(len(idx2vec_dict), len(word2idx_dict))) + new_emb_mat = np.array([idx2vec_dict[idx] for idx in range(len(idx2vec_dict))], dtype='float32') + config.new_emb_mat = new_emb_mat + + pprint(config.__flags, indent=2) + models = get_multi_gpu_models(config) + model = models[0] + evaluator = ForwardEvaluator(config, model) + graph_handler = GraphHandler(config) # controls all tensors and variables in the graph, including loading /saving + + sess = tf.Session(config=tf.ConfigProto(allow_soft_placement=True)) + graph_handler.initialize(sess) + + num_batches = math.ceil(test_data.num_examples / config.batch_size) + if 0 < config.eval_num_batches < num_batches: + num_batches = config.eval_num_batches + e = evaluator.get_evaluation_from_batches(sess, tqdm(test_data.get_batches(config.batch_size, num_batches=num_batches), total=num_batches)) + print(e) + if config.dump_answer: + print("dumping answer ...") + graph_handler.dump_answer(e, path=config.answer_path) + if config.dump_eval: + print("dumping eval ...") + graph_handler.dump_eval(e) + + +def set_dirs(config): + # create directories + if not config.load and os.path.exists(config.out_dir): + shutil.rmtree(config.out_dir) + + config.save_dir = os.path.join(config.out_dir, "save") + config.log_dir = os.path.join(config.out_dir, "log") + config.eval_dir = os.path.join(config.out_dir, "eval") + config.answer_dir = os.path.join(config.out_dir, "answer") + if not os.path.exists(config.out_dir): + os.makedirs(config.out_dir) + if not os.path.exists(config.save_dir): + os.mkdir(config.save_dir) + if not os.path.exists(config.log_dir): + os.mkdir(config.log_dir) + if not os.path.exists(config.answer_dir): + os.mkdir(config.answer_dir) + if not os.path.exists(config.eval_dir): + os.mkdir(config.eval_dir) + + +def _get_args(): + parser = argparse.ArgumentParser() + parser.add_argument("config_path") + return parser.parse_args() + + +class Config(object): + def __init__(self, **entries): + self.__dict__.update(entries) + + +def _run(): + args = _get_args() + with open(args.config_path, 'r') as fh: + config = Config(**json.load(fh)) + main(config) + + +if __name__ == "__main__": + _run() diff --git a/tensorflow/SQuAD/basic_cnn/model.py b/tensorflow/SQuAD/basic_cnn/model.py new file mode 100644 index 0000000..5985c80 --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/model.py @@ -0,0 +1,375 @@ +import random + +import itertools +import numpy as np +import tensorflow as tf +from tensorflow.python.ops.rnn_cell import BasicLSTMCell, GRUCell + +from basic_cnn.read_data import DataSet +from basic_cnn.superhighway import SHCell +from my.tensorflow import exp_mask, get_initializer, VERY_SMALL_NUMBER +from my.tensorflow.nn import linear, double_linear_logits, linear_logits, softsel, dropout, get_logits, softmax, \ + highway_network, multi_conv1d +from my.tensorflow.rnn import bidirectional_dynamic_rnn, dynamic_rnn +from my.tensorflow.rnn_cell import SwitchableDropoutWrapper, AttentionCell + + +def bi_attention(config, is_train, h, u, h_mask=None, u_mask=None, scope=None, tensor_dict=None): + """ + h_a: + all u attending on h + choosing an element of h that max-matches u + First creates confusion matrix between h and u + Then take max of the attention weights over u row + Finally softmax over + + u_a: + each h attending on u + + :param h: [N, M, JX, d] + :param u: [N, JQ, d] + :param h_mask: [N, M, JX] + :param u_mask: [N, B] + :param scope: + :return: [N, M, d], [N, M, JX, d] + """ + with tf.variable_scope(scope or "bi_attention"): + N, M, JX, JQ, d = config.batch_size, config.max_num_sents, config.max_sent_size, config.max_ques_size, config.hidden_size + JX = tf.shape(h)[2] + h_aug = tf.tile(tf.expand_dims(h, 3), [1, 1, 1, JQ, 1]) + u_aug = tf.tile(tf.expand_dims(tf.expand_dims(u, 1), 1), [1, M, JX, 1, 1]) + if h_mask is None: + and_mask = None + else: + h_mask_aug = tf.tile(tf.expand_dims(h_mask, 3), [1, 1, 1, JQ]) + u_mask_aug = tf.tile(tf.expand_dims(tf.expand_dims(u_mask, 1), 1), [1, M, JX, 1]) + and_mask = h_mask_aug & u_mask_aug + + u_logits = get_logits([h_aug, u_aug], None, True, wd=config.wd, mask=and_mask, + is_train=is_train, func=config.logit_func, scope='u_logits') # [N, M, JX, JQ] + u_a = softsel(u_aug, u_logits) # [N, M, JX, d] + if tensor_dict is not None: + # a_h = tf.nn.softmax(h_logits) # [N, M, JX] + a_u = tf.nn.softmax(u_logits) # [N, M, JX, JQ] + # tensor_dict['a_h'] = a_h + tensor_dict['a_u'] = a_u + if config.bi: + h_a = softsel(h, tf.reduce_max(u_logits, 3)) # [N, M, d] + h_a = tf.tile(tf.expand_dims(h_a, 2), [1, 1, JX, 1]) + else: + h_a = None + return u_a, h_a + + +def attention_layer(config, is_train, h, u, h_mask=None, u_mask=None, scope=None, tensor_dict=None): + with tf.variable_scope(scope or "attention_layer"): + u_a, h_a = bi_attention(config, is_train, h, u, h_mask=h_mask, u_mask=u_mask, tensor_dict=tensor_dict) + if config.bi: + p0 = tf.concat(axis=3, values=[h , u_a, h * u_a, h * h_a]) + else: + p0 = tf.concat(axis=3, values=[h , u_a, h * u_a]) + return p0 + + +class Model(object): + def __init__(self, config, scope): + self.scope = scope + self.config = config + self.global_step = tf.get_variable('global_step', shape=[], dtype='int32', + initializer=tf.constant_initializer(0), trainable=False) + + # Define forward inputs here + N, M, JX, JQ, VW, VC, W = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size, config.max_word_size + self.x = tf.placeholder('int32', [N, M, None], name='x') + self.cx = tf.placeholder('int32', [N, M, None, W], name='cx') + self.x_mask = tf.placeholder('bool', [N, M, None], name='x_mask') + self.q = tf.placeholder('int32', [N, JQ], name='q') + self.cq = tf.placeholder('int32', [N, JQ, W], name='cq') + self.q_mask = tf.placeholder('bool', [N, JQ], name='q_mask') + self.y = tf.placeholder('bool', [N, M, JX], name='y') + self.is_train = tf.placeholder('bool', [], name='is_train') + self.new_emb_mat = tf.placeholder('float', [None, config.word_emb_size], name='new_emb_mat') + + # Define misc + self.tensor_dict = {} + + # Forward outputs / loss inputs + self.logits = None + self.yp = None + self.var_list = None + + # Loss outputs + self.loss = None + + self._build_forward() + self._build_loss() + if config.mode == 'train': + self._build_ema() + + self.summary = tf.summary.merge_all() + self.summary = tf.summary.merge(tf.get_collection("summaries", scope=self.scope)) + + def _build_forward(self): + config = self.config + N, M, JX, JQ, VW, VC, d, W = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size, config.hidden_size, \ + config.max_word_size + JX = tf.shape(self.x)[2] + dc, dw, dco = config.char_emb_size, config.word_emb_size, config.char_out_size + + with tf.variable_scope("emb"): + with tf.variable_scope("emb_var"), tf.device("/cpu:0"): + char_emb_mat = tf.get_variable("char_emb_mat", shape=[VC, dc], dtype='float') + + with tf.variable_scope("char"): + Acx = tf.nn.embedding_lookup(char_emb_mat, self.cx) # [N, M, JX, W, dc] + Acq = tf.nn.embedding_lookup(char_emb_mat, self.cq) # [N, JQ, W, dc] + Acx = tf.reshape(Acx, [-1, JX, W, dc]) + Acq = tf.reshape(Acq, [-1, JQ, W, dc]) + + filter_sizes = list(map(int, config.out_channel_dims.split(','))) + heights = list(map(int, config.filter_heights.split(','))) + assert sum(filter_sizes) == dco + with tf.variable_scope("conv"): + xx = multi_conv1d(Acx, filter_sizes, heights, "VALID", self.is_train, config.keep_prob, scope="xx") + if config.share_cnn_weights: + tf.get_variable_scope().reuse_variables() + qq = multi_conv1d(Acq, filter_sizes, heights, "VALID", self.is_train, config.keep_prob, scope="xx") + else: + qq = multi_conv1d(Acq, filter_sizes, heights, "VALID", self.is_train, config.keep_prob, scope="qq") + xx = tf.reshape(xx, [-1, M, JX, dco]) + qq = tf.reshape(qq, [-1, JQ, dco]) + + if config.use_word_emb: + with tf.variable_scope("emb_var"), tf.device("/cpu:0"): + if config.mode == 'train': + word_emb_mat = tf.get_variable("word_emb_mat", dtype='float', shape=[VW, dw], initializer=get_initializer(config.emb_mat)) + else: + word_emb_mat = tf.get_variable("word_emb_mat", shape=[VW, dw], dtype='float') + if config.use_glove_for_unk: + word_emb_mat = tf.concat(axis=0, values=[word_emb_mat, self.new_emb_mat]) + + with tf.name_scope("word"): + Ax = tf.nn.embedding_lookup(word_emb_mat, self.x) # [N, M, JX, d] + Aq = tf.nn.embedding_lookup(word_emb_mat, self.q) # [N, JQ, d] + self.tensor_dict['x'] = Ax + self.tensor_dict['q'] = Aq + xx = tf.concat(axis=3, values=[xx, Ax]) # [N, M, JX, di] + qq = tf.concat(axis=2, values=[qq, Aq]) # [N, JQ, di] + + # highway network + with tf.variable_scope("highway"): + xx = highway_network(xx, config.highway_num_layers, True, wd=config.wd, is_train=self.is_train) + tf.get_variable_scope().reuse_variables() + qq = highway_network(qq, config.highway_num_layers, True, wd=config.wd, is_train=self.is_train) + self.tensor_dict['xx'] = xx + self.tensor_dict['qq'] = qq + + cell = BasicLSTMCell(d, state_is_tuple=True) + d_cell = SwitchableDropoutWrapper(cell, self.is_train, input_keep_prob=config.input_keep_prob) + x_len = tf.reduce_sum(tf.cast(self.x_mask, 'int32'), 2) # [N, M] + q_len = tf.reduce_sum(tf.cast(self.q_mask, 'int32'), 1) # [N] + + with tf.variable_scope("prepro"): + (fw_u, bw_u), ((_, fw_u_f), (_, bw_u_f)) = bidirectional_dynamic_rnn(d_cell, d_cell, qq, q_len, dtype='float', scope='u1') # [N, J, d], [N, d] + u = tf.concat(axis=2, values=[fw_u, bw_u]) + if config.two_prepro_layers: + (fw_u, bw_u), ((_, fw_u_f), (_, bw_u_f)) = bidirectional_dynamic_rnn(d_cell, d_cell, u, q_len, dtype='float', scope='u2') # [N, J, d], [N, d] + u = tf.concat(axis=2, values=[fw_u, bw_u]) + if config.share_lstm_weights: + tf.get_variable_scope().reuse_variables() + (fw_h, bw_h), _ = bidirectional_dynamic_rnn(cell, cell, xx, x_len, dtype='float', scope='u1') # [N, M, JX, 2d] + h = tf.concat(axis=3, values=[fw_h, bw_h]) # [N, M, JX, 2d] + if config.two_prepro_layers: + (fw_h, bw_h), _ = bidirectional_dynamic_rnn(cell, cell, h, x_len, dtype='float', scope='u2') # [N, M, JX, 2d] + h = tf.concat(axis=3, values=[fw_h, bw_h]) # [N, M, JX, 2d] + + else: + (fw_h, bw_h), _ = bidirectional_dynamic_rnn(cell, cell, xx, x_len, dtype='float', scope='h1') # [N, M, JX, 2d] + h = tf.concat(axis=3, values=[fw_h, bw_h]) # [N, M, JX, 2d] + if config.two_prepro_layers: + (fw_h, bw_h), _ = bidirectional_dynamic_rnn(cell, cell, h, x_len, dtype='float', scope='h2') # [N, M, JX, 2d] + h = tf.concat(axis=3, values=[fw_h, bw_h]) # [N, M, JX, 2d] + self.tensor_dict['u'] = u + self.tensor_dict['h'] = h + + with tf.variable_scope("main"): + p0 = attention_layer(config, self.is_train, h, u, h_mask=self.x_mask, u_mask=self.q_mask, scope="p0", tensor_dict=self.tensor_dict) + (fw_g0, bw_g0), _ = bidirectional_dynamic_rnn(d_cell, d_cell, p0, x_len, dtype='float', scope='g0') # [N, M, JX, 2d] + g0 = tf.concat(axis=3, values=[fw_g0, bw_g0]) + # p1 = attention_layer(config, self.is_train, g0, u, h_mask=self.x_mask, u_mask=self.q_mask, scope="p1") + (fw_g1, bw_g1), _ = bidirectional_dynamic_rnn(d_cell, d_cell, g0, x_len, dtype='float', scope='g1') # [N, M, JX, 2d] + g1 = tf.concat(axis=3, values=[fw_g1, bw_g1]) + # logits = u_logits(config, self.is_train, g1, u, h_mask=self.x_mask, u_mask=self.q_mask, scope="logits") + # [N, M, JX] + logits = get_logits([g1, p0], d, True, wd=config.wd, input_keep_prob=config.input_keep_prob, mask=self.x_mask, is_train=self.is_train, func=config.answer_func, scope='logits1') + a1i = softsel(tf.reshape(g1, [N, M*JX, 2*d]), tf.reshape(logits, [N, M*JX])) + + if config.feed_gt: + logy = tf.log(tf.cast(self.y, 'float') + VERY_SMALL_NUMBER) + logits = tf.cond(self.is_train, lambda: logy, lambda: logits) + if config.feed_hard: + hard_yp = tf.argmax(tf.reshape(logits, [N, M*JX]), 1) + hard_logits = tf.reshape(tf.one_hot(hard_yp, M*JX), [N, M, JX]) # [N, M, JX] + logits = tf.cond(self.is_train, lambda: logits, lambda: hard_logits) + + flat_logits = tf.reshape(logits, [-1, M * JX]) + flat_yp = tf.nn.softmax(flat_logits) # [-1, M*JX] + yp = tf.reshape(flat_yp, [-1, M, JX]) + + self.tensor_dict['g1'] = g1 + + self.logits = flat_logits + self.yp = yp + + def _build_loss(self): + config = self.config + N, M, JX, JQ, VW, VC = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size + JX = tf.shape(self.x)[2] + loss_mask = tf.reduce_max(tf.cast(self.q_mask, 'float'), 1) + losses = -tf.log(tf.reduce_sum(self.yp * tf.cast(self.y, 'float'), [1, 2]) + VERY_SMALL_NUMBER) + ce_loss = tf.reduce_mean(loss_mask * losses) + tf.add_to_collection('losses', ce_loss) + + self.loss = tf.add_n(tf.get_collection('losses', scope=self.scope), name='loss') + tf.summary.scalar(self.loss.op.name, self.loss) + tf.add_to_collection('ema/scalar', self.loss) + + def _build_ema(self): + ema = tf.train.ExponentialMovingAverage(self.config.decay) + ema_op = ema.apply(tf.get_collection("ema/scalar", scope=self.scope) + tf.get_collection("ema/histogram", scope=self.scope)) + for var in tf.get_collection("ema/scalar", scope=self.scope): + ema_var = ema.average(var) + tf.summary.scalar(ema_var.op.name, ema_var) + for var in tf.get_collection("ema/histogram", scope=self.scope): + ema_var = ema.average(var) + tf.summary.histogram(ema_var.op.name, ema_var) + + with tf.control_dependencies([ema_op]): + self.loss = tf.identity(self.loss) + + def get_loss(self): + return self.loss + + def get_global_step(self): + return self.global_step + + def get_var_list(self): + return self.var_list + + def get_feed_dict(self, batch, is_train, supervised=True): + assert isinstance(batch, DataSet) + config = self.config + N, M, JX, JQ, VW, VC, d, W = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size, config.hidden_size, config.max_word_size + feed_dict = {} + + if config.len_opt: + """ + Note that this optimization results in variable GPU RAM usage (i.e. can cause OOM in the middle of training.) + First test without len_opt and make sure no OOM, and use len_opt + """ + if sum(len(para) for para in batch.data['x']) == 0: + new_JX = 1 + else: + new_JX = max(len(para) for para in batch.data['x']) + JX = min(JX, new_JX) + # print(JX) + + x = np.zeros([N, M, JX], dtype='int32') + cx = np.zeros([N, M, JX, W], dtype='int32') + x_mask = np.zeros([N, M, JX], dtype='bool') + q = np.zeros([N, JQ], dtype='int32') + cq = np.zeros([N, JQ, W], dtype='int32') + q_mask = np.zeros([N, JQ], dtype='bool') + + feed_dict[self.x] = x + feed_dict[self.x_mask] = x_mask + feed_dict[self.cx] = cx + feed_dict[self.q] = q + feed_dict[self.cq] = cq + feed_dict[self.q_mask] = q_mask + feed_dict[self.is_train] = is_train + if config.use_glove_for_unk: + feed_dict[self.new_emb_mat] = batch.shared['new_emb_mat'] + + X = batch.data['x'] + CX = batch.data['cx'] + + def _get_word(word): + if word.startswith("@"): + return 2 + d = batch.shared['word2idx'] + for each in (word, word.lower(), word.capitalize(), word.upper()): + if each in d: + return d[each] + if config.use_glove_for_unk: + d2 = batch.shared['new_word2idx'] + for each in (word, word.lower(), word.capitalize(), word.upper()): + if each in d2: + return d2[each] + len(d) + return 1 + + def _get_char(char): + d = batch.shared['char2idx'] + if char in d: + return d[char] + return 1 + + if supervised: + y = np.zeros([N, M, JX], dtype='int32') + feed_dict[self.y] = y + + for i, (xi, yi) in enumerate(zip(batch.data['x'], batch.data['y'])): + count = 0 + for j, xij in enumerate(xi): + for k, xijk in enumerate(xij): + if xijk == yi: + y[i, j, k] = True + count += 1 + assert count > 0 + + for i, xi in enumerate(X): + for j, xij in enumerate(xi): + for k, xijk in enumerate(xij): + each = _get_word(xijk) + x[i, j, k] = each + x_mask[i, j, k] = True + + for i, cxi in enumerate(CX): + for j, cxij in enumerate(cxi): + for k, cxijk in enumerate(cxij): + for l, cxijkl in enumerate(cxijk): + cx[i, j, k, l] = _get_char(cxijkl) + if l + 1 == config.max_word_size: + break + + for i, qi in enumerate(batch.data['q']): + for j, qij in enumerate(qi): + q[i, j] = _get_word(qij) + q_mask[i, j] = True + + for i, cqi in enumerate(batch.data['cq']): + for j, cqij in enumerate(cqi): + for k, cqijk in enumerate(cqij): + cq[i, j, k] = _get_char(cqijk) + if k + 1 == config.max_word_size: + break + + return feed_dict + + +def get_multi_gpu_models(config): + models = [] + for gpu_idx in range(config.num_gpus): + with tf.name_scope("model_{}".format(gpu_idx)) as scope, tf.device("/gpu:{}".format(gpu_idx)): + model = Model(config, scope) + tf.get_variable_scope().reuse_variables() + models.append(model) + return models diff --git a/tensorflow/SQuAD/basic_cnn/read_data.py b/tensorflow/SQuAD/basic_cnn/read_data.py new file mode 100644 index 0000000..853ef36 --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/read_data.py @@ -0,0 +1,294 @@ +import json +import os +import random +import itertools +import math +from collections import defaultdict + +import numpy as np + +from cnn_dm.prepro import para2sents +from my.tensorflow import grouper +from my.utils import index + + +class Data(object): + def get_size(self): + raise NotImplementedError() + + def get_by_idxs(self, idxs): + """ + Efficient way to obtain a batch of items from filesystem + :param idxs: + :return dict: {'X': [,], 'Y', } + """ + data = defaultdict(list) + for idx in idxs: + each_data = self.get_one(idx) + for key, val in each_data.items(): + data[key].append(val) + return data + + def get_one(self, idx): + raise NotImplementedError() + + def get_empty(self): + raise NotImplementedError() + + def __add__(self, other): + raise NotImplementedError() + +class MyData(Data): + def __init__(self, config, root_dir, file_names): + self.root_dir = root_dir + self.file_names = file_names + self.config = config + + def get_one(self, idx): + file_name = self.file_names[idx] + with open(os.path.join(self.root_dir, file_name), 'r') as fh: + url = fh.readline().strip() + _ = fh.readline() + para = fh.readline().strip() + _ = fh.readline() + ques = fh.readline().strip() + _ = fh.readline() + answer = fh.readline().strip() + _ = fh.readline() + cands = list(line.strip() for line in fh) + cand_ents = list(cand.split(":")[0] for cand in cands) + wordss = para2sents(para, self.config.width) + ques_words = ques.split(" ") + + x = wordss + cx = [[list(word) for word in words] for words in wordss] + q = ques_words + cq = [list(word) for word in ques_words] + y = answer + c = cand_ents + + data = {'x': x, 'cx': cx, 'q': q, 'cq': cq, 'y': y, 'c': c, 'ids': file_name} + return data + + def get_empty(self): + return MyData(self.config, self.root_dir, []) + + def __add__(self, other): + file_names = self.file_names + other.file_names + return MyData(self.config, self.root_dir, file_names) + + def get_size(self): + return len(self.file_names) + + +class DataSet(object): + def __init__(self, data, data_type, shared=None, valid_idxs=None): + self.data = data # e.g. {'X': [0, 1, 2], 'Y': [2, 3, 4]} + self.data_type = data_type + self.shared = shared + total_num_examples = self.get_data_size() + self.valid_idxs = range(total_num_examples) if valid_idxs is None else valid_idxs + self.num_examples = total_num_examples + + def _sort_key(self, idx): + rx = self.data['*x'][idx] + x = self.shared['x'][rx[0]][rx[1]] + return max(map(len, x)) + + def get_data_size(self): + if isinstance(self.data, dict): + return len(next(iter(self.data.values()))) + elif isinstance(self.data, Data): + return self.data.get_size() + raise Exception() + + def get_by_idxs(self, idxs): + if isinstance(self.data, dict): + out = defaultdict(list) + for key, val in self.data.items(): + out[key].extend(val[idx] for idx in idxs) + return out + elif isinstance(self.data, Data): + return self.data.get_by_idxs(idxs) + raise Exception() + + def get_one(self, idx): + if isinstance(self.data, dict): + out = {key: [val[idx]] for key, val in self.data.items()} + return out + elif isinstance(self.data, Data): + return self.data.get_one(idx) + + def get_batches(self, batch_size, num_batches=None, shuffle=False, cluster=False): + """ + + :param batch_size: + :param num_batches: + :param shuffle: + :param cluster: cluster examples by their lengths; this might give performance boost (i.e. faster training). + :return: + """ + num_batches_per_epoch = int(math.ceil(self.num_examples / batch_size)) + if num_batches is None: + num_batches = num_batches_per_epoch + num_epochs = int(math.ceil(num_batches / num_batches_per_epoch)) + + if shuffle: + random_idxs = random.sample(self.valid_idxs, len(self.valid_idxs)) + if cluster: + sorted_idxs = sorted(random_idxs, key=self._sort_key) + sorted_grouped = lambda: list(grouper(sorted_idxs, batch_size)) + grouped = lambda: random.sample(sorted_grouped(), num_batches_per_epoch) + else: + random_grouped = lambda: list(grouper(random_idxs, batch_size)) + grouped = random_grouped + else: + raw_grouped = lambda: list(grouper(self.valid_idxs, batch_size)) + grouped = raw_grouped + + batch_idx_tuples = itertools.chain.from_iterable(grouped() for _ in range(num_epochs)) + for _ in range(num_batches): + batch_idxs = tuple(i for i in next(batch_idx_tuples) if i is not None) + batch_data = self.get_by_idxs(batch_idxs) + shared_batch_data = {} + for key, val in batch_data.items(): + if key.startswith('*'): + assert self.shared is not None + shared_key = key[1:] + shared_batch_data[shared_key] = [index(self.shared[shared_key], each) for each in val] + batch_data.update(shared_batch_data) + + batch_ds = DataSet(batch_data, self.data_type, shared=self.shared) + yield batch_idxs, batch_ds + + def get_multi_batches(self, batch_size, num_batches_per_step, num_steps=None, shuffle=False, cluster=False): + batch_size_per_step = batch_size * num_batches_per_step + batches = self.get_batches(batch_size_per_step, num_batches=num_steps, shuffle=shuffle, cluster=cluster) + multi_batches = (tuple(zip(grouper(idxs, batch_size, shorten=True, num_groups=num_batches_per_step), + data_set.divide(num_batches_per_step))) for idxs, data_set in batches) + return multi_batches + + def get_empty(self): + if isinstance(self.data, dict): + data = {key: [] for key in self.data} + elif isinstance(self.data, Data): + data = self.data.get_empty() + else: + raise Exception() + return DataSet(data, self.data_type, shared=self.shared) + + def __add__(self, other): + if isinstance(self.data, dict): + data = {key: val + other.data[key] for key, val in self.data.items()} + elif isinstance(self.data, Data): + data = self.data + other.data + else: + raise Exception() + + valid_idxs = list(self.valid_idxs) + [valid_idx + self.num_examples for valid_idx in other.valid_idxs] + return DataSet(data, self.data_type, shared=self.shared, valid_idxs=valid_idxs) + + def divide(self, integer): + batch_size = int(math.ceil(self.num_examples / integer)) + idxs_gen = grouper(self.valid_idxs, batch_size, shorten=True, num_groups=integer) + data_gen = (self.get_by_idxs(idxs) for idxs in idxs_gen) + ds_tuple = tuple(DataSet(data, self.data_type, shared=self.shared) for data in data_gen) + return ds_tuple + + +class MyDataSet(DataSet): + def __init__(self, data, data_type, shared=None, valid_idxs=None): + super(MyDataSet, self).__init__(data, data_type, shared=shared, valid_idxs=valid_idxs) + shared['max_num_sents'] = len(self.get_one(self.num_examples-1)['x']) + + def _sort_key(self, idx): + return idx + + +def read_data(config, data_type, ref, data_filter=None): + shared_path = os.path.join(config.data_dir, "shared_{}.json".format(data_type)) + with open(shared_path, 'r') as fh: + shared = json.load(fh) + + paths = shared['sorted'] + if config.filter_ratio < 1.0: + stop = int(round(len(paths) * config.filter_ratio)) + paths = paths[:stop] + num_examples = len(paths) + valid_idxs = range(num_examples) + + print("Loaded {}/{} examples from {}".format(len(valid_idxs), num_examples, data_type)) + + shared_path = config.shared_path or os.path.join(config.out_dir, "shared.json") + if not ref: + word2vec_dict = shared['lower_word2vec'] if config.lower_word else shared['word2vec'] + word_counter = shared['lower_word_counter'] if config.lower_word else shared['word_counter'] + char_counter = shared['char_counter'] + if config.finetune: + shared['word2idx'] = {word: idx + 3 for idx, word in + enumerate(word for word, count in word_counter.items() + if count > config.word_count_th or (config.known_if_glove and word in word2vec_dict))} + else: + assert config.known_if_glove + assert config.use_glove_for_unk + shared['word2idx'] = {word: idx + 3 for idx, word in + enumerate(word for word, count in word_counter.items() + if count > config.word_count_th and word not in word2vec_dict)} + shared['char2idx'] = {char: idx + 2 for idx, char in + enumerate(char for char, count in char_counter.items() + if count > config.char_count_th)} + NULL = "-NULL-" + UNK = "-UNK-" + ENT = "-ENT-" + shared['word2idx'][NULL] = 0 + shared['word2idx'][UNK] = 1 + shared['word2idx'][ENT] = 2 + shared['char2idx'][NULL] = 0 + shared['char2idx'][UNK] = 1 + + json.dump({'word2idx': shared['word2idx'], 'char2idx': shared['char2idx']}, open(shared_path, 'w')) + else: + new_shared = json.load(open(shared_path, 'r')) + for key, val in new_shared.items(): + shared[key] = val + + if config.use_glove_for_unk: + # create new word2idx and word2vec + word2vec_dict = shared['lower_word2vec'] if config.lower_word else shared['word2vec'] + new_word2idx_dict = {word: idx for idx, word in enumerate(word for word in word2vec_dict.keys() if word not in shared['word2idx'])} + shared['new_word2idx'] = new_word2idx_dict + offset = len(shared['word2idx']) + word2vec_dict = shared['lower_word2vec'] if config.lower_word else shared['word2vec'] + new_word2idx_dict = shared['new_word2idx'] + idx2vec_dict = {idx: word2vec_dict[word] for word, idx in new_word2idx_dict.items()} + # print("{}/{} unique words have corresponding glove vectors.".format(len(idx2vec_dict), len(word2idx_dict))) + new_emb_mat = np.array([idx2vec_dict[idx] for idx in range(len(idx2vec_dict))], dtype='float32') + shared['new_emb_mat'] = new_emb_mat + + data = MyData(config, os.path.join(config.root_dir, data_type), paths) + data_set = MyDataSet(data, data_type, shared=shared, valid_idxs=valid_idxs) + return data_set + + +def get_cnn_data_filter(config): + return True + + +def update_config(config, data_sets): + config.max_num_sents = 0 + config.max_sent_size = 0 + config.max_ques_size = 0 + config.max_word_size = 0 + for data_set in data_sets: + shared = data_set.shared + config.max_sent_size = max(config.max_sent_size, shared['max_sent_size']) + config.max_ques_size = max(config.max_ques_size, shared['max_ques_size']) + config.max_word_size = max(config.max_word_size, shared['max_word_size']) + config.max_num_sents = max(config.max_num_sents, shared['max_num_sents']) + + config.max_word_size = min(config.max_word_size, config.word_size_th) + + config.char_vocab_size = len(data_sets[0].shared['char2idx']) + config.word_emb_size = len(next(iter(data_sets[0].shared['word2vec'].values()))) + config.word_vocab_size = len(data_sets[0].shared['word2idx']) + diff --git a/tensorflow/SQuAD/basic_cnn/superhighway.py b/tensorflow/SQuAD/basic_cnn/superhighway.py new file mode 100644 index 0000000..059a01e --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/superhighway.py @@ -0,0 +1,47 @@ +import tensorflow as tf +from tensorflow.python.ops.rnn_cell import RNNCell + +from my.tensorflow.nn import linear + + +class SHCell(RNNCell): + """ + Super-Highway Cell + """ + def __init__(self, input_size, logit_func='tri_linear', scalar=False): + self._state_size = input_size + self._output_size = input_size + self._logit_func = logit_func + self._scalar = scalar + + @property + def state_size(self): + return self._state_size + + @property + def output_size(self): + return self._output_size + + def __call__(self, inputs, state, scope=None): + with tf.variable_scope(scope or "SHCell"): + a_size = 1 if self._scalar else self._state_size + h, u = tf.split(axis=1, num_or_size_splits=2, value=inputs) + if self._logit_func == 'mul_linear': + args = [h * u, state * u] + a = tf.nn.sigmoid(linear(args, a_size, True)) + elif self._logit_func == 'linear': + args = [h, u, state] + a = tf.nn.sigmoid(linear(args, a_size, True)) + elif self._logit_func == 'tri_linear': + args = [h, u, state, h * u, state * u] + a = tf.nn.sigmoid(linear(args, a_size, True)) + elif self._logit_func == 'double': + args = [h, u, state] + a = tf.nn.sigmoid(linear(tf.tanh(linear(args, a_size, True)), self._state_size, True)) + + else: + raise Exception() + new_state = a * state + (1 - a) * h + outputs = state + return outputs, new_state + diff --git a/tensorflow/SQuAD/basic_cnn/templates/visualizer.html b/tensorflow/SQuAD/basic_cnn/templates/visualizer.html new file mode 100644 index 0000000..8a4602d --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/templates/visualizer.html @@ -0,0 +1,76 @@ + + + + + {{ title }} + + + + + + +

    {{ title }}

    + + + + + + + + + + {% for row in rows %} + + + + + + + + + {% endfor %} +
    IDQuestionAnswersPredictedScoreParagraph
    {{ row.id }} + {% for qj in row.ques %} + {{ qj }} + {% endfor %} + + {% for aa in row.a %} +
  • {{ aa }}
    • + {% endfor %} +     		{{ row.ap }}{{ row.score }} + + {% for xj, ypj, yp2j in zip(row.para, row.yp, row.yp2) %} + + {% set rowloop = loop %} + {% for xjk, ypjk in zip(xj, ypj) %} + + {% endfor %} + + + {% for xjk, yp2jk in zip(xj, yp2j) %} + + {% endfor %} + + {% endfor %} +
    + {% if row.y[0][0] == rowloop.index0 and row.y[0][1] <= loop.index0 <= row.y[1][1] %} + {{ xjk }} + {% else %} + {{ xjk }} + {% endif %} +
    -
    +
    + + \ No newline at end of file diff --git a/tensorflow/SQuAD/basic_cnn/trainer.py b/tensorflow/SQuAD/basic_cnn/trainer.py new file mode 100644 index 0000000..9f4548f --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/trainer.py @@ -0,0 +1,73 @@ +import tensorflow as tf + +from basic_cnn.model import Model +from my.tensorflow import average_gradients + + +class Trainer(object): + def __init__(self, config, model): + assert isinstance(model, Model) + self.config = config + self.model = model + self.opt = tf.train.AdadeltaOptimizer(config.init_lr) + self.loss = model.get_loss() + self.var_list = model.get_var_list() + self.global_step = model.get_global_step() + self.summary = model.summary + self.grads = self.opt.compute_gradients(self.loss, var_list=self.var_list) + self.train_op = self.opt.apply_gradients(self.grads, global_step=self.global_step) + + def get_train_op(self): + return self.train_op + + def step(self, sess, batch, get_summary=False): + assert isinstance(sess, tf.Session) + _, ds = batch + feed_dict = self.model.get_feed_dict(ds, True) + if get_summary: + loss, summary, train_op = \ + sess.run([self.loss, self.summary, self.train_op], feed_dict=feed_dict) + else: + loss, train_op = sess.run([self.loss, self.train_op], feed_dict=feed_dict) + summary = None + return loss, summary, train_op + + +class MultiGPUTrainer(object): + def __init__(self, config, models): + model = models[0] + assert isinstance(model, Model) + self.config = config + self.model = model + self.opt = tf.train.AdadeltaOptimizer(config.init_lr) + self.var_list = model.get_var_list() + self.global_step = model.get_global_step() + self.summary = model.summary + self.models = models + losses = [] + grads_list = [] + for gpu_idx, model in enumerate(models): + with tf.name_scope("grads_{}".format(gpu_idx)), tf.device("/gpu:{}".format(gpu_idx)): + loss = model.get_loss() + grads = self.opt.compute_gradients(loss, var_list=self.var_list) + losses.append(loss) + grads_list.append(grads) + + self.loss = tf.add_n(losses)/len(losses) + self.grads = average_gradients(grads_list) + self.train_op = self.opt.apply_gradients(self.grads, global_step=self.global_step) + + def step(self, sess, batches, get_summary=False): + assert isinstance(sess, tf.Session) + feed_dict = {} + for batch, model in zip(batches, self.models): + _, ds = batch + feed_dict.update(model.get_feed_dict(ds, True)) + + if get_summary: + loss, summary, train_op = \ + sess.run([self.loss, self.summary, self.train_op], feed_dict=feed_dict) + else: + loss, train_op = sess.run([self.loss, self.train_op], feed_dict=feed_dict) + summary = None + return loss, summary, train_op diff --git a/tensorflow/SQuAD/basic_cnn/visualizer.py b/tensorflow/SQuAD/basic_cnn/visualizer.py new file mode 100644 index 0000000..18b0798 --- /dev/null +++ b/tensorflow/SQuAD/basic_cnn/visualizer.py @@ -0,0 +1,137 @@ +import shutil +from collections import OrderedDict +import http.server +import socketserver +import argparse +import json +import os +import numpy as np +from tqdm import tqdm + +from jinja2 import Environment, FileSystemLoader + +from basic_cnn.evaluator import get_span_score_pairs, get_best_span + + +def bool_(string): + if string == 'True': + return True + elif string == 'False': + return False + else: + raise Exception() + +def get_args(): + parser = argparse.ArgumentParser() + parser.add_argument("--model_name", type=str, default='basic') + parser.add_argument("--data_type", type=str, default='dev') + parser.add_argument("--step", type=int, default=5000) + parser.add_argument("--template_name", type=str, default="visualizer.html") + parser.add_argument("--num_per_page", type=int, default=100) + parser.add_argument("--data_dir", type=str, default="data/squad") + parser.add_argument("--port", type=int, default=8000) + parser.add_argument("--host", type=str, default="0.0.0.0") + parser.add_argument("--open", type=str, default='False') + parser.add_argument("--run_id", type=str, default="0") + + args = parser.parse_args() + return args + + +def _decode(decoder, sent): + return " ".join(decoder[idx] for idx in sent) + + +def accuracy2_visualizer(args): + model_name = args.model_name + data_type = args.data_type + num_per_page = args.num_per_page + data_dir = args.data_dir + run_id = args.run_id.zfill(2) + step = args.step + + eval_path =os.path.join("out", model_name, run_id, "eval", "{}-{}.json".format(data_type, str(step).zfill(6))) + print("loading {}".format(eval_path)) + eval_ = json.load(open(eval_path, 'r')) + + _id = 0 + html_dir = "/tmp/list_results%d" % _id + while os.path.exists(html_dir): + _id += 1 + html_dir = "/tmp/list_results%d" % _id + + if os.path.exists(html_dir): + shutil.rmtree(html_dir) + os.mkdir(html_dir) + + cur_dir = os.path.dirname(os.path.realpath(__file__)) + templates_dir = os.path.join(cur_dir, 'templates') + env = Environment(loader=FileSystemLoader(templates_dir)) + env.globals.update(zip=zip, reversed=reversed) + template = env.get_template(args.template_name) + + data_path = os.path.join(data_dir, "data_{}.json".format(data_type)) + shared_path = os.path.join(data_dir, "shared_{}.json".format(data_type)) + print("loading {}".format(data_path)) + data = json.load(open(data_path, 'r')) + print("loading {}".format(shared_path)) + shared = json.load(open(shared_path, 'r')) + + rows = [] + for i, (idx, yi, ypi, yp2i) in tqdm(enumerate(zip(*[eval_[key] for key in ('idxs', 'y', 'yp', 'yp2')])), total=len(eval_['idxs'])): + id_, q, rx, answers = (data[key][idx] for key in ('ids', 'q', '*x', 'answerss')) + x = shared['x'][rx[0]][rx[1]] + ques = [" ".join(q)] + para = [[word for word in sent] for sent in x] + span = get_best_span(ypi, yp2i) + ap = get_segment(para, span) + score = "{:.3f}".format(ypi[span[0][0]][span[0][1]] * yp2i[span[1][0]][span[1][1]-1]) + + row = { + 'id': id_, + 'title': "Hello world!", + 'ques': ques, + 'para': para, + 'y': yi[0][0], + 'y2': yi[0][1], + 'yp': ypi, + 'yp2': yp2i, + 'a': answers, + 'ap': ap, + 'score': score + } + rows.append(row) + + if i % num_per_page == 0: + html_path = os.path.join(html_dir, "%s.html" % str(i).zfill(8)) + + if (i + 1) % num_per_page == 0 or (i + 1) == len(eval_['y']): + var_dict = {'title': "Accuracy Visualization", + 'rows': rows + } + with open(html_path, "wb") as f: + f.write(template.render(**var_dict).encode('UTF-8')) + rows = [] + + os.chdir(html_dir) + port = args.port + host = args.host + # Overriding to suppress log message + class MyHandler(http.server.SimpleHTTPRequestHandler): + def log_message(self, format, *args): + pass + handler = MyHandler + httpd = socketserver.TCPServer((host, port), handler) + if args.open == 'True': + os.system("open http://%s:%d" % (args.host, args.port)) + print("serving at %s:%d" % (host, port)) + httpd.serve_forever() + + +def get_segment(para, span): + return " ".join(para[span[0][0]][span[0][1]:span[1][1]]) + + +if __name__ == "__main__": + ARGS = get_args() + accuracy2_visualizer(ARGS) \ No newline at end of file diff --git a/tensorflow/SQuAD/cnn_dm/__init__.py b/tensorflow/SQuAD/cnn_dm/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/tensorflow/SQuAD/cnn_dm/eda.ipynb b/tensorflow/SQuAD/cnn_dm/eda.ipynb new file mode 100644 index 0000000..e17a978 --- /dev/null +++ b/tensorflow/SQuAD/cnn_dm/eda.ipynb @@ -0,0 +1,359 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 22, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "100%|██████████| 3198/3198 [00:12<00:00, 264.62it/s]\n" + ] + } + ], + "source": [ + "import os\n", + "import nltk\n", + "import re\n", + "\n", + "from collections import Counter\n", + "\n", + "from tqdm import tqdm\n", + "\n", + "root_dir = \"/Users/minjoons/data/cnn/questions\"\n", + "data_dir = os.path.join(root_dir, \"test\")\n", + "\n", + "char_counter = Counter()\n", + "word_counter = Counter()\n", + "ent_counter = Counter()\n", + "max_num_words = 0\n", + "max_num_ques_words = 0\n", + "max_num_sents = 0\n", + "max_num_words_per_sent = 0\n", + "max_num_chars = 0\n", + "\n", + "nums_words = []\n", + "nums_ques_words = []\n", + "nums_sents = []\n", + "nums_words_per_sent = []\n", + "nums_chars = []\n", + "nums_entities = []\n", + "\n", + "sent_tokenize = lambda x: re.split(\"[.!?]\", x)\n", + "sent_tokenize = nltk.sent_tokenize\n", + "\n", + "num_ques = len(list(os.listdir(data_dir)))\n", + "\n", + "cand_set= set()\n", + "\n", + "for path in tqdm(os.listdir(data_dir), total=num_ques):\n", + " if path.endswith(\".question\"):\n", + " with open(os.path.join(data_dir, path), 'r') as fh:\n", + " url = fh.readline().strip()\n", + " _ = fh.readline()\n", + " para = fh.readline().strip()\n", + " _ = fh.readline()\n", + " ques = fh.readline().strip()\n", + " _ = fh.readline()\n", + " answer = fh.readline().strip()\n", + " _ = fh.readline()\n", + " cands = list(line.strip() for line in fh)\n", + " cand_ents, cand_names = zip(*[cand.split(\":\") for cand in cands])\n", + " cand_set = cand_set | set(cand_names)\n", + " words = para.split(\" \")\n", + " sents = sent_tokenize(para)\n", + " wordss = list(sent.split(\" \") for sent in sents)\n", + " ques_words = ques.split(\" \")\n", + " \n", + " ents = [word for word in words if word.startswith(\"@\")]\n", + " num_ents = len(ents)\n", + " \n", + " nums_entities.append(num_ents)\n", + " nums_words.append(len(words))\n", + " nums_ques_words.append(len(ques_words))\n", + " nums_sents.append(len(sents))\n", + " nums_words_per_sent.extend(map(len, wordss))\n", + " nums_chars.extend(map(len, words))\n", + " \n", + " for word in ques_words:\n", + " if word.startswith(\"@\"):\n", + " ent_counter[word] += 1\n", + " else:\n", + " word_counter[word] += 1\n", + " for c in word:\n", + " char_counter[c] += 1\n", + " \n", + " for word in words:\n", + " if word.startswith(\"@\"):\n", + " ent_counter[word] += 1\n", + " else:\n", + " word_counter[word] += 1\n", + " for c in word:\n", + " char_counter[c] += 1" + ] + }, + { + "cell_type": "code", + "execution_count": 23, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "text/plain": [ + "(22747, 465, 77, 12465)" + ] + }, + "execution_count": 23, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "len(word_counter), len(ent_counter), len(char_counter), len(cand_set)" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "text/plain": [ + "(1989, 37, 122, 443, 24)" + ] + }, + "execution_count": 7, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "max_num_words, max_num_ques_words, max_num_sents, max_num_words_per_sent, max_num_chars" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "text/plain": [ + "['hello', ' Wow', ' Hmm', '']" + ] + }, + "execution_count": 4, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "import re\n", + "re.split(\"[.!?]\", \"hello. Wow! Hmm?\")\n" + ] + }, + { + "cell_type": "code", + "execution_count": 22, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niRAlfRGRCFHSFxGJkP8HWh63bv3V45sAAAAASUVORK5CYII=\n", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "%matplotlib inline\n", + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "counter = Counter(nums_words)\n", + "values = list(counter.values())\n", + "plt.plot(list(counter.keys()), np.cumsum(values)/sum(values))\n", + "plt.show()\n", + "# plt.hist(nums_words)\n", + "# plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAEACAYAAACznAEdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKgXY9oUcA04AAAAASUVORK5CYII=\n", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "\n", + "counter = Counter(nums_words_per_sent)\n", + "plt.plot(list(counter.keys()), list(counter.values()))\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAEACAYAAACznAEdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plt.hist(nums_sents)\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJqLQFxFpIgp9EZEm8v8BFfxqFwpqM18AAAAASUVORK5CYII=\n", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "import matplotlib.pyplot as plt\n", + "%matplotlib inline\n", + "counter = Counter(nums_entities)\n", + "plt.plot(list(counter.keys()), list(counter.values()))\n", + "plt.show()\n" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAEACAYAAAC9Gb03AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nLxlQngAAAABJRU5ErkJggg==\n", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "%matplotlib inline\n", + "counter = Counter(nums_entities)\n", + "keys = sorted(counter.keys())\n", + "values = [counter[key] for key in keys]\n", + "plt.plot(keys, np.cumsum(values)/sum(values))\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEACAYAAABI5zaHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "%matplotlib inline\n", + "counter = Counter(nums_ques_words)\n", + "keys = sorted(counter.keys())\n", + "values = [counter[key] for key in keys]\n", + "plt.plot(keys, np.cumsum(values)/sum(values))\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.5.1" + } + }, + "nbformat": 4, + "nbformat_minor": 0 +} diff --git a/tensorflow/SQuAD/cnn_dm/evaluate.py b/tensorflow/SQuAD/cnn_dm/evaluate.py new file mode 100644 index 0000000..33d9936 --- /dev/null +++ b/tensorflow/SQuAD/cnn_dm/evaluate.py @@ -0,0 +1,38 @@ +import json +import os +import sys + +root_dir = sys.argv[1] +answer_path = sys.argv[2] +file_names = os.listdir(root_dir) + +num_correct = 0 +num_wrong = 0 + +with open(answer_path, 'r') as fh: + id2answer_dict = json.load(fh) + +for file_name in file_names: + if not file_name.endswith(".question"): + continue + with open(os.path.join(root_dir, file_name), 'r') as fh: + url = fh.readline().strip() + _ = fh.readline() + para = fh.readline().strip() + _ = fh.readline() + ques = fh.readline().strip() + _ = fh.readline() + answer = fh.readline().strip() + _ = fh.readline() + if file_name in id2answer_dict: + pred = id2answer_dict[file_name] + if pred == answer: + num_correct += 1 + else: + num_wrong += 1 + else: + num_wrong += 1 + +total = num_correct + num_wrong +acc = float(num_correct) / total +print("{} = {} / {}".format(acc, num_correct, total)) \ No newline at end of file diff --git a/tensorflow/SQuAD/cnn_dm/prepro.py b/tensorflow/SQuAD/cnn_dm/prepro.py new file mode 100644 index 0000000..6bd9b81 --- /dev/null +++ b/tensorflow/SQuAD/cnn_dm/prepro.py @@ -0,0 +1,185 @@ +import argparse +import json +import os +# data: q, cq, (dq), (pq), y, *x, *cx +# shared: x, cx, (dx), (px), word_counter, char_counter, word2vec +# no metadata +from collections import Counter + +from tqdm import tqdm + +from my.utils import process_tokens +from squad.utils import get_word_span, process_tokens + + +def bool_(arg): + if arg == 'True': + return True + elif arg == 'False': + return False + raise Exception(arg) + + +def main(): + args = get_args() + prepro(args) + + +def get_args(): + parser = argparse.ArgumentParser() + home = os.path.expanduser("~") + source_dir = os.path.join(home, "data", "cnn", 'questions') + target_dir = "data/cnn" + glove_dir = os.path.join(home, "data", "glove") + parser.add_argument("--source_dir", default=source_dir) + parser.add_argument("--target_dir", default=target_dir) + parser.add_argument("--glove_dir", default=glove_dir) + parser.add_argument("--glove_corpus", default='6B') + parser.add_argument("--glove_vec_size", default=100, type=int) + parser.add_argument("--debug", default=False, type=bool_) + parser.add_argument("--num_sents_th", default=200, type=int) + parser.add_argument("--ques_size_th", default=30, type=int) + parser.add_argument("--width", default=5, type=int) + # TODO : put more args here + return parser.parse_args() + + +def prepro(args): + prepro_each(args, 'train') + prepro_each(args, 'dev') + prepro_each(args, 'test') + + +def para2sents(para, width): + """ + Turn para into double array of words (wordss) + Where each sentence is up to 5 word neighbors of each entity + :param para: + :return: + """ + words = para.split(" ") + sents = [] + for i, word in enumerate(words): + if word.startswith("@"): + start = max(i - width, 0) + stop = min(i + width + 1, len(words)) + sent = words[start:stop] + sents.append(sent) + return sents + + +def get_word2vec(args, word_counter): + glove_path = os.path.join(args.glove_dir, "glove.{}.{}d.txt".format(args.glove_corpus, args.glove_vec_size)) + sizes = {'6B': int(4e5), '42B': int(1.9e6), '840B': int(2.2e6), '2B': int(1.2e6)} + total = sizes[args.glove_corpus] + word2vec_dict = {} + with open(glove_path, 'r', encoding='utf-8') as fh: + for line in tqdm(fh, total=total): + array = line.lstrip().rstrip().split(" ") + word = array[0] + vector = list(map(float, array[1:])) + if word in word_counter: + word2vec_dict[word] = vector + elif word.capitalize() in word_counter: + word2vec_dict[word.capitalize()] = vector + elif word.lower() in word_counter: + word2vec_dict[word.lower()] = vector + elif word.upper() in word_counter: + word2vec_dict[word.upper()] = vector + + print("{}/{} of word vocab have corresponding vectors in {}".format(len(word2vec_dict), len(word_counter), glove_path)) + return word2vec_dict + + +def prepro_each(args, mode): + source_dir = os.path.join(args.source_dir, mode) + word_counter = Counter() + lower_word_counter = Counter() + ent_counter = Counter() + char_counter = Counter() + max_sent_size = 0 + max_word_size = 0 + max_ques_size = 0 + max_num_sents = 0 + + file_names = list(os.listdir(source_dir)) + if args.debug: + file_names = file_names[:1000] + lens = [] + + out_file_names = [] + for file_name in tqdm(file_names, total=len(file_names)): + if file_name.endswith(".question"): + with open(os.path.join(source_dir, file_name), 'r') as fh: + url = fh.readline().strip() + _ = fh.readline() + para = fh.readline().strip() + _ = fh.readline() + ques = fh.readline().strip() + _ = fh.readline() + answer = fh.readline().strip() + _ = fh.readline() + cands = list(line.strip() for line in fh) + cand_ents = list(cand.split(":")[0] for cand in cands) + sents = para2sents(para, args.width) + ques_words = ques.split(" ") + + # Filtering + if len(sents) > args.num_sents_th or len(ques_words) > args.ques_size_th: + continue + + max_sent_size = max(max(map(len, sents)), max_sent_size) + max_ques_size = max(len(ques_words), max_ques_size) + max_word_size = max(max(len(word) for sent in sents for word in sent), max_word_size) + max_num_sents = max(len(sents), max_num_sents) + + for word in ques_words: + if word.startswith("@"): + ent_counter[word] += 1 + word_counter[word] += 1 + else: + word_counter[word] += 1 + lower_word_counter[word.lower()] += 1 + for c in word: + char_counter[c] += 1 + for sent in sents: + for word in sent: + if word.startswith("@"): + ent_counter[word] += 1 + word_counter[word] += 1 + else: + word_counter[word] += 1 + lower_word_counter[word.lower()] += 1 + for c in word: + char_counter[c] += 1 + + out_file_names.append(file_name) + lens.append(len(sents)) + num_examples = len(out_file_names) + + assert len(out_file_names) == len(lens) + sorted_file_names, lens = zip(*sorted(zip(out_file_names, lens), key=lambda each: each[1])) + assert lens[-1] == max_num_sents + + word2vec_dict = get_word2vec(args, word_counter) + lower_word2vec_dit = get_word2vec(args, lower_word_counter) + + shared = {'word_counter': word_counter, 'ent_counter': ent_counter, 'char_counter': char_counter, + 'lower_word_counter': lower_word_counter, + 'max_num_sents': max_num_sents, 'max_sent_size': max_sent_size, 'max_word_size': max_word_size, + 'max_ques_size': max_ques_size, + 'word2vec': word2vec_dict, 'lower_word2vec': lower_word2vec_dit, 'sorted': sorted_file_names, + 'num_examples': num_examples} + + print("max num sents: {}".format(max_num_sents)) + print("max ques size: {}".format(max_ques_size)) + + if not os.path.exists(args.target_dir): + os.makedirs(args.target_dir) + shared_path = os.path.join(args.target_dir, "shared_{}.json".format(mode)) + with open(shared_path, 'w') as fh: + json.dump(shared, fh) + + +if __name__ == "__main__": + main() diff --git a/tensorflow/SQuAD/download.sh b/tensorflow/SQuAD/download.sh new file mode 100755 index 0000000..5588d4f --- /dev/null +++ b/tensorflow/SQuAD/download.sh @@ -0,0 +1,25 @@ +#!/usr/bin/env bash + +DATA_DIR=$HOME/data +mkdir $DATA_DIR + +# Download SQuAD +SQUAD_DIR=$DATA_DIR/squad +mkdir $SQUAD_DIR +wget https://rajpurkar.github.io/SQuAD-explorer/dataset/train-v1.1.json -O $SQUAD_DIR/train-v1.1.json +wget https://rajpurkar.github.io/SQuAD-explorer/dataset/dev-v1.1.json -O $SQUAD_DIR/dev-v1.1.json + + +# Download CNN and DailyMail +# Download at: http://cs.nyu.edu/~kcho/DMQA/ + + +# Download GloVe +GLOVE_DIR=$DATA_DIR/glove +mkdir $GLOVE_DIR +wget http://nlp.stanford.edu/data/glove.6B.zip -O $GLOVE_DIR/glove.6B.zip +unzip $GLOVE_DIR/glove.6B.zip -d $GLOVE_DIR + +# Download NLTK (for tokenizer) +# Make sure that nltk is installed! +python3 -m nltk.downloader -d $HOME/nltk_data punkt diff --git a/tensorflow/SQuAD/my/__init__.py b/tensorflow/SQuAD/my/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/tensorflow/SQuAD/my/corenlp_interface.py b/tensorflow/SQuAD/my/corenlp_interface.py new file mode 100644 index 0000000..474c8cd --- /dev/null +++ b/tensorflow/SQuAD/my/corenlp_interface.py @@ -0,0 +1,55 @@ +import logging + +import requests +import nltk +import json +import networkx as nx +import time + + +class CoreNLPInterface(object): + def __init__(self, url, port): + self._url = url + self._port = port + + def get(self, type_, in_, num_max_requests=100): + in_ = in_.encode("utf-8") + url = "http://{}:{}/{}".format(self._url, self._port, type_) + out = None + for _ in range(num_max_requests): + try: + r = requests.post(url, data=in_) + out = r.content.decode('utf-8') + if out == 'error': + out = None + break + except: + time.sleep(1) + return out + + def split_doc(self, doc): + out = self.get("doc", doc) + return out if out is None else json.loads(out) + + def split_sent(self, sent): + out = self.get("sent", sent) + return out if out is None else json.loads(out) + + def get_dep(self, sent): + out = self.get("dep", sent) + return out if out is None else json.loads(out) + + def get_const(self, sent): + out = self.get("const", sent) + return out + + def get_const_tree(self, sent): + out = self.get_const(sent) + return out if out is None else nltk.tree.Tree.fromstring(out) + + @staticmethod + def dep2tree(dep): + tree = nx.DiGraph() + for dep, i, gov, j, label in dep: + tree.add_edge(gov, dep, label=label) + return tree diff --git a/tensorflow/SQuAD/my/nltk_utils.py b/tensorflow/SQuAD/my/nltk_utils.py new file mode 100644 index 0000000..202f8e9 --- /dev/null +++ b/tensorflow/SQuAD/my/nltk_utils.py @@ -0,0 +1,129 @@ +import nltk +import numpy as np + + +def _set_span(t, i): + if isinstance(t[0], str): + t.span = (i, i+len(t)) + else: + first = True + for c in t: + cur_span = _set_span(c, i) + i = cur_span[1] + if first: + min_ = cur_span[0] + first = False + max_ = cur_span[1] + t.span = (min_, max_) + return t.span + + +def set_span(t): + assert isinstance(t, nltk.tree.Tree) + try: + return _set_span(t, 0) + except: + print(t) + exit() + + +def tree_contains_span(tree, span): + """ + Assumes that tree span has been set with set_span + Returns true if any subtree of t has exact span as the given span + :param t: + :param span: + :return bool: + """ + return span in set(t.span for t in tree.subtrees()) + + +def span_len(span): + return span[1] - span[0] + + +def span_overlap(s1, s2): + start = max(s1[0], s2[0]) + stop = min(s1[1], s2[1]) + if stop > start: + return start, stop + return None + + +def span_prec(true_span, pred_span): + overlap = span_overlap(true_span, pred_span) + if overlap is None: + return 0 + return span_len(overlap) / span_len(pred_span) + + +def span_recall(true_span, pred_span): + overlap = span_overlap(true_span, pred_span) + if overlap is None: + return 0 + return span_len(overlap) / span_len(true_span) + + +def span_f1(true_span, pred_span): + p = span_prec(true_span, pred_span) + r = span_recall(true_span, pred_span) + if p == 0 or r == 0: + return 0.0 + return 2 * p * r / (p + r) + + +def find_max_f1_span(tree, span): + return find_max_f1_subtree(tree, span).span + + +def find_max_f1_subtree(tree, span): + return max(((t, span_f1(span, t.span)) for t in tree.subtrees()), key=lambda p: p[1])[0] + + +def tree2matrix(tree, node2num, row_size=None, col_size=None, dtype='int32'): + set_span(tree) + D = tree.height() - 1 + B = len(tree.leaves()) + row_size = row_size or D + col_size = col_size or B + matrix = np.zeros([row_size, col_size], dtype=dtype) + mask = np.zeros([row_size, col_size, col_size], dtype='bool') + + for subtree in tree.subtrees(): + row = subtree.height() - 2 + col = subtree.span[0] + matrix[row, col] = node2num(subtree) + for subsub in subtree.subtrees(): + if isinstance(subsub, nltk.tree.Tree): + mask[row, col, subsub.span[0]] = True + if not isinstance(subsub[0], nltk.tree.Tree): + c = subsub.span[0] + for r in range(row): + mask[r, c, c] = True + else: + mask[row, col, col] = True + + return matrix, mask + + +def load_compressed_tree(s): + + def compress_tree(tree): + assert not isinstance(tree, str) + if len(tree) == 1: + if isinstance(tree[0], nltk.tree.Tree): + return compress_tree(tree[0]) + else: + return tree + else: + for i, t in enumerate(tree): + if isinstance(t, nltk.tree.Tree): + tree[i] = compress_tree(t) + else: + tree[i] = t + return tree + + return compress_tree(nltk.tree.Tree.fromstring(s)) + + + diff --git a/tensorflow/SQuAD/my/tensorflow/__init__.py b/tensorflow/SQuAD/my/tensorflow/__init__.py new file mode 100644 index 0000000..d8a36dc --- /dev/null +++ b/tensorflow/SQuAD/my/tensorflow/__init__.py @@ -0,0 +1 @@ +from my.tensorflow.general import * \ No newline at end of file diff --git a/tensorflow/SQuAD/my/tensorflow/general.py b/tensorflow/SQuAD/my/tensorflow/general.py new file mode 100644 index 0000000..e14b4f0 --- /dev/null +++ b/tensorflow/SQuAD/my/tensorflow/general.py @@ -0,0 +1,177 @@ +from itertools import zip_longest + +import itertools +import tensorflow as tf +from functools import reduce +from operator import mul +import numpy as np + +VERY_BIG_NUMBER = 1e30 +VERY_SMALL_NUMBER = 1e-30 +VERY_POSITIVE_NUMBER = VERY_BIG_NUMBER +VERY_NEGATIVE_NUMBER = -VERY_BIG_NUMBER + + +def get_initializer(matrix): + def _initializer(shape, dtype=None, partition_info=None, **kwargs): return matrix + return _initializer + + +def variable_on_cpu(name, shape, initializer): + """Helper to create a Variable stored on CPU memory. + + Args: + name: name of the variable + shape: list of ints + initializer: initializer for Variable + + Returns: + Variable Tensor + """ + with tf.device('/cpu:0'): + var = tf.get_variable(name, shape, initializer=initializer) + return var + + +def variable_with_weight_decay(name, shape, stddev, wd): + """Helper to create an initialized Variable with weight decay. + + Note that the Variable is initialized with a truncated normal distribution. + A weight decay is added only if one is specified. + + Args: + name: name of the variable + shape: list of ints + stddev: standard deviation of a truncated Gaussian + wd: add L2Loss weight decay multiplied by this float. If None, weight + decay is not added for this Variable. + + Returns: + Variable Tensor + """ + var = variable_on_cpu(name, shape, + tf.truncated_normal_initializer(stddev=stddev)) + if wd: + weight_decay = tf.multiply(tf.nn.l2_loss(var), wd, name='weight_loss') + tf.add_to_collection('losses', weight_decay) + return var + + +def average_gradients(tower_grads): + """Calculate the average gradient for each shared variable across all towers. + + Note that this function provides a synchronization point across all towers. + + Args: + tower_grads: List of lists of (gradient, variable) tuples. The outer list + is over individual gradients. The inner list is over the gradient + calculation for each tower. + Returns: + List of pairs of (gradient, variable) where the gradient has been averaged + across all towers. + """ + average_grads = [] + for grad_and_vars in zip(*tower_grads): + # Note that each grad_and_vars looks like the following: + # ((grad0_gpu0, var0_gpu0), ... , (grad0_gpuN, var0_gpuN)) + grads = [] + for g, var in grad_and_vars: + # Add 0 dimension to the gradients to represent the tower. + assert g is not None, var.name + expanded_g = tf.expand_dims(g, 0) + + # Append on a 'tower' dimension which we will average over below. + grads.append(expanded_g) + + # Average over the 'tower' dimension. + grad = tf.concat(axis=0, values=grads) + grad = tf.reduce_mean(grad, 0) + + # Keep in mind that the Variables are redundant because they are shared + # across towers. So .. we will just return the first tower's pointer to + # the Variable. + v = grad_and_vars[0][1] + grad_and_var = (grad, v) + average_grads.append(grad_and_var) + return average_grads + + +def mask(val, mask, name=None): + if name is None: + name = 'mask' + return tf.multiply(val, tf.cast(mask, 'float'), name=name) + + +def exp_mask(val, mask, name=None): + """Give very negative number to unmasked elements in val. + For example, [-3, -2, 10], [True, True, False] -> [-3, -2, -1e9]. + Typically, this effectively masks in exponential space (e.g. softmax) + Args: + val: values to be masked + mask: masking boolean tensor, same shape as tensor + name: name for output tensor + + Returns: + Same shape as val, where some elements are very small (exponentially zero) + """ + if name is None: + name = "exp_mask" + return tf.add(val, (1 - tf.cast(mask, 'float')) * VERY_NEGATIVE_NUMBER, name=name) + + +def flatten(tensor, keep): + fixed_shape = tensor.get_shape().as_list() + start = len(fixed_shape) - keep + left = reduce(mul, [fixed_shape[i] or tf.shape(tensor)[i] for i in range(start)]) + out_shape = [left] + [fixed_shape[i] or tf.shape(tensor)[i] for i in range(start, len(fixed_shape))] + flat = tf.reshape(tensor, out_shape) + return flat + + +def reconstruct(tensor, ref, keep): + ref_shape = ref.get_shape().as_list() + tensor_shape = tensor.get_shape().as_list() + ref_stop = len(ref_shape) - keep + tensor_start = len(tensor_shape) - keep + pre_shape = [ref_shape[i] or tf.shape(ref)[i] for i in range(ref_stop)] + keep_shape = [tensor_shape[i] or tf.shape(tensor)[i] for i in range(tensor_start, len(tensor_shape))] + # pre_shape = [tf.shape(ref)[i] for i in range(len(ref.get_shape().as_list()[:-keep]))] + # keep_shape = tensor.get_shape().as_list()[-keep:] + target_shape = pre_shape + keep_shape + out = tf.reshape(tensor, target_shape) + return out + + +def add_wd(wd, scope=None): + scope = scope or tf.get_variable_scope().name + variables = tf.get_collection(tf.GraphKeys.TRAINABLE_VARIABLES, scope=scope) + with tf.name_scope("weight_decay"): + for var in variables: + weight_decay = tf.multiply(tf.nn.l2_loss(var), wd, name="{}/wd".format(var.op.name)) + tf.add_to_collection('losses', weight_decay) + + +def grouper(iterable, n, fillvalue=None, shorten=False, num_groups=None): + args = [iter(iterable)] * n + out = zip_longest(*args, fillvalue=fillvalue) + out = list(out) + if num_groups is not None: + default = (fillvalue, ) * n + assert isinstance(num_groups, int) + out = list(each for each, _ in zip_longest(out, range(num_groups), fillvalue=default)) + if shorten: + assert fillvalue is None + out = (tuple(e for e in each if e is not None) for each in out) + return out + +def padded_reshape(tensor, shape, mode='CONSTANT', name=None): + paddings = [[0, shape[i] - tf.shape(tensor)[i]] for i in range(len(shape))] + return tf.pad(tensor, paddings, mode=mode, name=name) + + +def get_num_params(): + num_params = 0 + for variable in tf.trainable_variables(): + shape = variable.get_shape() + num_params += reduce(mul, [dim.value for dim in shape], 1) + return num_params diff --git a/tensorflow/SQuAD/my/tensorflow/nn.py b/tensorflow/SQuAD/my/tensorflow/nn.py new file mode 100644 index 0000000..d8c70b7 --- /dev/null +++ b/tensorflow/SQuAD/my/tensorflow/nn.py @@ -0,0 +1,180 @@ +from tensorflow.python.ops.rnn_cell_impl import _linear +from tensorflow.python.util import nest +import tensorflow as tf + +from my.tensorflow import flatten, reconstruct, add_wd, exp_mask + + +def linear(args, output_size, bias, bias_start=0.0, scope=None, squeeze=False, wd=0.0, input_keep_prob=1.0, + is_train=None): + if args is None or (nest.is_sequence(args) and not args): + raise ValueError("`args` must be specified") + if not nest.is_sequence(args): + args = [args] + + flat_args = [flatten(arg, 1) for arg in args] + if input_keep_prob < 1.0: + assert is_train is not None + flat_args = [tf.cond(is_train, lambda: tf.nn.dropout(arg, input_keep_prob), lambda: arg) + for arg in flat_args] + with tf.variable_scope(scope or 'Linear'): + flat_out = _linear(flat_args, output_size, bias, bias_initializer=tf.constant_initializer(bias_start)) + out = reconstruct(flat_out, args[0], 1) + if squeeze: + out = tf.squeeze(out, [len(args[0].get_shape().as_list())-1]) + if wd: + add_wd(wd) + + return out + + +def dropout(x, keep_prob, is_train, noise_shape=None, seed=None, name=None): + with tf.name_scope(name or "dropout"): + if keep_prob < 1.0: + d = tf.nn.dropout(x, keep_prob, noise_shape=noise_shape, seed=seed) + out = tf.cond(is_train, lambda: d, lambda: x) + return out + return x + + +def softmax(logits, mask=None, scope=None): + with tf.name_scope(scope or "Softmax"): + if mask is not None: + logits = exp_mask(logits, mask) + flat_logits = flatten(logits, 1) + flat_out = tf.nn.softmax(flat_logits) + out = reconstruct(flat_out, logits, 1) + + return out + + +def softsel(target, logits, mask=None, scope=None): + """ + + :param target: [ ..., J, d] dtype=float + :param logits: [ ..., J], dtype=float + :param mask: [ ..., J], dtype=bool + :param scope: + :return: [..., d], dtype=float + """ + with tf.name_scope(scope or "Softsel"): + a = softmax(logits, mask=mask) + target_rank = len(target.get_shape().as_list()) + out = tf.reduce_sum(tf.expand_dims(a, -1) * target, target_rank - 2) + return out + + +def double_linear_logits(args, size, bias, bias_start=0.0, scope=None, mask=None, wd=0.0, input_keep_prob=1.0, is_train=None): + with tf.variable_scope(scope or "Double_Linear_Logits"): + first = tf.tanh(linear(args, size, bias, bias_start=bias_start, scope='first', + wd=wd, input_keep_prob=input_keep_prob, is_train=is_train)) + second = linear(first, 1, bias, bias_start=bias_start, squeeze=True, scope='second', + wd=wd, input_keep_prob=input_keep_prob, is_train=is_train) + if mask is not None: + second = exp_mask(second, mask) + return second + + +def linear_logits(args, bias, bias_start=0.0, scope=None, mask=None, wd=0.0, input_keep_prob=1.0, is_train=None): + with tf.variable_scope(scope or "Linear_Logits"): + logits = linear(args, 1, bias, bias_start=bias_start, squeeze=True, scope='first', + wd=wd, input_keep_prob=input_keep_prob, is_train=is_train) + if mask is not None: + logits = exp_mask(logits, mask) + return logits + + +def sum_logits(args, mask=None, name=None): + with tf.name_scope(name or "sum_logits"): + if args is None or (nest.is_sequence(args) and not args): + raise ValueError("`args` must be specified") + if not nest.is_sequence(args): + args = [args] + rank = len(args[0].get_shape()) + logits = sum(tf.reduce_sum(arg, rank-1) for arg in args) + if mask is not None: + logits = exp_mask(logits, mask) + return logits + + +def get_logits(args, size, bias, bias_start=0.0, scope=None, mask=None, wd=0.0, input_keep_prob=1.0, is_train=None, func=None): + if func is None: + func = "sum" + if func == 'sum': + return sum_logits(args, mask=mask, name=scope) + elif func == 'linear': + return linear_logits(args, bias, bias_start=bias_start, scope=scope, mask=mask, wd=wd, input_keep_prob=input_keep_prob, + is_train=is_train) + elif func == 'double': + return double_linear_logits(args, size, bias, bias_start=bias_start, scope=scope, mask=mask, wd=wd, input_keep_prob=input_keep_prob, + is_train=is_train) + elif func == 'dot': + assert len(args) == 2 + arg = args[0] * args[1] + return sum_logits([arg], mask=mask, name=scope) + elif func == 'mul_linear': + assert len(args) == 2 + arg = args[0] * args[1] + return linear_logits([arg], bias, bias_start=bias_start, scope=scope, mask=mask, wd=wd, input_keep_prob=input_keep_prob, + is_train=is_train) + elif func == 'proj': + assert len(args) == 2 + d = args[1].get_shape()[-1] + proj = linear([args[0]], d, False, bias_start=bias_start, scope=scope, wd=wd, input_keep_prob=input_keep_prob, + is_train=is_train) + return sum_logits([proj * args[1]], mask=mask) + elif func == 'tri_linear': + assert len(args) == 2 + new_arg = args[0] * args[1] + return linear_logits([args[0], args[1], new_arg], bias, bias_start=bias_start, scope=scope, mask=mask, wd=wd, input_keep_prob=input_keep_prob, + is_train=is_train) + else: + raise Exception() + + +def highway_layer(arg, bias, bias_start=0.0, scope=None, wd=0.0, input_keep_prob=1.0, is_train=None): + with tf.variable_scope(scope or "highway_layer"): + d = arg.get_shape()[-1] + trans = linear([arg], d, bias, bias_start=bias_start, scope='trans', wd=wd, input_keep_prob=input_keep_prob, is_train=is_train) + trans = tf.nn.relu(trans) + gate = linear([arg], d, bias, bias_start=bias_start, scope='gate', wd=wd, input_keep_prob=input_keep_prob, is_train=is_train) + gate = tf.nn.sigmoid(gate) + out = gate * trans + (1 - gate) * arg + return out + + +def highway_network(arg, num_layers, bias, bias_start=0.0, scope=None, wd=0.0, input_keep_prob=1.0, is_train=None): + with tf.variable_scope(scope or "highway_network"): + prev = arg + cur = None + for layer_idx in range(num_layers): + cur = highway_layer(prev, bias, bias_start=bias_start, scope="layer_{}".format(layer_idx), wd=wd, + input_keep_prob=input_keep_prob, is_train=is_train) + prev = cur + return cur + + +def conv1d(in_, filter_size, height, padding, is_train=None, keep_prob=1.0, scope=None): + with tf.variable_scope(scope or "conv1d"): + num_channels = in_.get_shape()[-1] + filter_ = tf.get_variable("filter", shape=[1, height, num_channels, filter_size], dtype='float') + bias = tf.get_variable("bias", shape=[filter_size], dtype='float') + strides = [1, 1, 1, 1] + if is_train is not None and keep_prob < 1.0: + in_ = dropout(in_, keep_prob, is_train) + xxc = tf.nn.conv2d(in_, filter_, strides, padding) + bias # [N*M, JX, W/filter_stride, d] + out = tf.reduce_max(tf.nn.relu(xxc), 2) # [-1, JX, d] + return out + + +def multi_conv1d(in_, filter_sizes, heights, padding, is_train=None, keep_prob=1.0, scope=None): + with tf.variable_scope(scope or "multi_conv1d"): + assert len(filter_sizes) == len(heights) + outs = [] + for filter_size, height in zip(filter_sizes, heights): + if filter_size == 0: + continue + out = conv1d(in_, filter_size, height, padding, is_train=is_train, keep_prob=keep_prob, scope="conv1d_{}".format(height)) + outs.append(out) + concat_out = tf.concat(axis=2, values=outs) + return concat_out diff --git a/tensorflow/SQuAD/my/tensorflow/rnn.py b/tensorflow/SQuAD/my/tensorflow/rnn.py new file mode 100644 index 0000000..d873419 --- /dev/null +++ b/tensorflow/SQuAD/my/tensorflow/rnn.py @@ -0,0 +1,81 @@ +import tensorflow as tf +from tensorflow.python.ops.rnn import dynamic_rnn as _dynamic_rnn, \ + bidirectional_dynamic_rnn as _bidirectional_dynamic_rnn + +from my.tensorflow import flatten, reconstruct + + +def dynamic_rnn(cell, inputs, sequence_length=None, initial_state=None, + dtype=None, parallel_iterations=None, swap_memory=False, + time_major=False, scope=None): + assert not time_major # TODO : to be implemented later! + flat_inputs = flatten(inputs, 2) # [-1, J, d] + flat_len = None if sequence_length is None else tf.cast(flatten(sequence_length, 0), 'int64') + + flat_outputs, final_state = _dynamic_rnn(cell, flat_inputs, sequence_length=flat_len, + initial_state=initial_state, dtype=dtype, + parallel_iterations=parallel_iterations, swap_memory=swap_memory, + time_major=time_major, scope=scope) + + outputs = reconstruct(flat_outputs, inputs, 2) + return outputs, final_state + + +def bw_dynamic_rnn(cell, inputs, sequence_length=None, initial_state=None, + dtype=None, parallel_iterations=None, swap_memory=False, + time_major=False, scope=None): + assert not time_major # TODO : to be implemented later! + + flat_inputs = flatten(inputs, 2) # [-1, J, d] + flat_len = None if sequence_length is None else tf.cast(flatten(sequence_length, 0), 'int64') + + flat_inputs = tf.reverse(flat_inputs, 1) if sequence_length is None \ + else tf.reverse_sequence(flat_inputs, sequence_length, 1) + flat_outputs, final_state = _dynamic_rnn(cell, flat_inputs, sequence_length=flat_len, + initial_state=initial_state, dtype=dtype, + parallel_iterations=parallel_iterations, swap_memory=swap_memory, + time_major=time_major, scope=scope) + flat_outputs = tf.reverse(flat_outputs, 1) if sequence_length is None \ + else tf.reverse_sequence(flat_outputs, sequence_length, 1) + + outputs = reconstruct(flat_outputs, inputs, 2) + return outputs, final_state + + +def bidirectional_dynamic_rnn(cell_fw, cell_bw, inputs, sequence_length=None, + initial_state_fw=None, initial_state_bw=None, + dtype=None, parallel_iterations=None, + swap_memory=False, time_major=False, scope=None): + assert not time_major + + flat_inputs = flatten(inputs, 2) # [-1, J, d] + flat_len = None if sequence_length is None else tf.cast(flatten(sequence_length, 0), 'int64') + + (flat_fw_outputs, flat_bw_outputs), final_state = \ + _bidirectional_dynamic_rnn(cell_fw, cell_bw, flat_inputs, sequence_length=flat_len, + initial_state_fw=initial_state_fw, initial_state_bw=initial_state_bw, + dtype=dtype, parallel_iterations=parallel_iterations, swap_memory=swap_memory, + time_major=time_major, scope=scope) + + fw_outputs = reconstruct(flat_fw_outputs, inputs, 2) + bw_outputs = reconstruct(flat_bw_outputs, inputs, 2) + # FIXME : final state is not reshaped! + return (fw_outputs, bw_outputs), final_state + + +def bidirectional_rnn(cell_fw, cell_bw, inputs, + initial_state_fw=None, initial_state_bw=None, + dtype=None, sequence_length=None, scope=None): + + flat_inputs = flatten(inputs, 2) # [-1, J, d] + flat_len = None if sequence_length is None else tf.cast(flatten(sequence_length, 0), 'int64') + + (flat_fw_outputs, flat_bw_outputs), final_state = \ + tf.nn.bidirectional_dynamic_rnn(cell_fw, cell_bw, flat_inputs, sequence_length=flat_len, + initial_state_fw=initial_state_fw, initial_state_bw=initial_state_bw, + dtype=dtype, scope=scope) + + fw_outputs = reconstruct(flat_fw_outputs, inputs, 2) + bw_outputs = reconstruct(flat_bw_outputs, inputs, 2) + # FIXME : final state is not reshaped! + return (fw_outputs, bw_outputs), final_state diff --git a/tensorflow/SQuAD/my/tensorflow/rnn_cell.py b/tensorflow/SQuAD/my/tensorflow/rnn_cell.py new file mode 100644 index 0000000..bc64d04 --- /dev/null +++ b/tensorflow/SQuAD/my/tensorflow/rnn_cell.py @@ -0,0 +1,223 @@ +import tensorflow as tf +from tensorflow.contrib.rnn import DropoutWrapper, RNNCell, LSTMStateTuple + +from my.tensorflow import exp_mask, flatten +from my.tensorflow.nn import linear, softsel, double_linear_logits + + +class SwitchableDropoutWrapper(DropoutWrapper): + def __init__(self, cell, is_train, input_keep_prob=1.0, output_keep_prob=1.0, + seed=None): + super(SwitchableDropoutWrapper, self).__init__(cell, input_keep_prob=input_keep_prob, output_keep_prob=output_keep_prob, + seed=seed) + self.is_train = is_train + + def __call__(self, inputs, state, scope=None): + outputs_do, new_state_do = super(SwitchableDropoutWrapper, self).__call__(inputs, state, scope=scope) + tf.get_variable_scope().reuse_variables() + outputs, new_state = self._cell(inputs, state, scope) + outputs = tf.cond(self.is_train, lambda: outputs_do, lambda: outputs) + if isinstance(state, tuple): + new_state = state.__class__(*[tf.cond(self.is_train, lambda: new_state_do_i, lambda: new_state_i) + for new_state_do_i, new_state_i in zip(new_state_do, new_state)]) + else: + new_state = tf.cond(self.is_train, lambda: new_state_do, lambda: new_state) + return outputs, new_state + + +class TreeRNNCell(RNNCell): + def __init__(self, cell, input_size, reduce_func): + self._cell = cell + self._input_size = input_size + self._reduce_func = reduce_func + + def __call__(self, inputs, state, scope=None): + """ + :param inputs: [N*B, I + B] + :param state: [N*B, d] + :param scope: + :return: [N*B, d] + """ + with tf.variable_scope(scope or self.__class__.__name__): + d = self.state_size + x = tf.slice(inputs, [0, 0], [-1, self._input_size]) # [N*B, I] + mask = tf.slice(inputs, [0, self._input_size], [-1, -1]) # [N*B, B] + B = tf.shape(mask)[1] + prev_state = tf.expand_dims(tf.reshape(state, [-1, B, d]), 1) # [N, B, d] -> [N, 1, B, d] + mask = tf.tile(tf.expand_dims(tf.reshape(mask, [-1, B, B]), -1), [1, 1, 1, d]) # [N, B, B, d] + # prev_state = self._reduce_func(tf.tile(prev_state, [1, B, 1, 1]), 2) + prev_state = self._reduce_func(exp_mask(prev_state, mask), 2) # [N, B, d] + prev_state = tf.reshape(prev_state, [-1, d]) # [N*B, d] + return self._cell(x, prev_state) + + @property + def state_size(self): + return self._cell.state_size + + @property + def output_size(self): + return self._cell.output_size + + +class NoOpCell(RNNCell): + def __init__(self, num_units): + self._num_units = num_units + + def __call__(self, inputs, state, scope=None): + return state, state + + @property + def state_size(self): + return self._num_units + + @property + def output_size(self): + return self._num_units + + +class MatchCell(RNNCell): + def __init__(self, cell, input_size, q_len): + self._cell = cell + self._input_size = input_size + # FIXME : This won't be needed with good shape guessing + self._q_len = q_len + + @property + def state_size(self): + return self._cell.state_size + + @property + def output_size(self): + return self._cell.output_size + + def __call__(self, inputs, state, scope=None): + """ + + :param inputs: [N, d + JQ + JQ * d] + :param state: [N, d] + :param scope: + :return: + """ + with tf.variable_scope(scope or self.__class__.__name__): + c_prev, h_prev = state + x = tf.slice(inputs, [0, 0], [-1, self._input_size]) + q_mask = tf.slice(inputs, [0, self._input_size], [-1, self._q_len]) # [N, JQ] + qs = tf.slice(inputs, [0, self._input_size + self._q_len], [-1, -1]) + qs = tf.reshape(qs, [-1, self._q_len, self._input_size]) # [N, JQ, d] + x_tiled = tf.tile(tf.expand_dims(x, 1), [1, self._q_len, 1]) # [N, JQ, d] + h_prev_tiled = tf.tile(tf.expand_dims(h_prev, 1), [1, self._q_len, 1]) # [N, JQ, d] + f = tf.tanh(linear([qs, x_tiled, h_prev_tiled], self._input_size, True, scope='f')) # [N, JQ, d] + a = tf.nn.softmax(exp_mask(linear(f, 1, True, squeeze=True, scope='a'), q_mask)) # [N, JQ] + q = tf.reduce_sum(qs * tf.expand_dims(a, -1), 1) + z = tf.concat(axis=1, values=[x, q]) # [N, 2d] + return self._cell(z, state) + + +class AttentionCell(RNNCell): + def __init__(self, cell, memory, mask=None, controller=None, mapper=None, input_keep_prob=1.0, is_train=None): + """ + Early fusion attention cell: uses the (inputs, state) to control the current attention. + + :param cell: + :param memory: [N, M, m] + :param mask: + :param controller: (inputs, prev_state, memory) -> memory_logits + """ + self._cell = cell + self._memory = memory + self._mask = mask + self._flat_memory = flatten(memory, 2) + self._flat_mask = flatten(mask, 1) + if controller is None: + controller = AttentionCell.get_linear_controller(True, is_train=is_train) + self._controller = controller + if mapper is None: + mapper = AttentionCell.get_concat_mapper() + elif mapper == 'sim': + mapper = AttentionCell.get_sim_mapper() + self._mapper = mapper + + @property + def state_size(self): + return self._cell.state_size + + @property + def output_size(self): + return self._cell.output_size + + def __call__(self, inputs, state, scope=None): + with tf.variable_scope(scope or "AttentionCell"): + memory_logits = self._controller(inputs, state, self._flat_memory) + sel_mem = softsel(self._flat_memory, memory_logits, mask=self._flat_mask) # [N, m] + new_inputs, new_state = self._mapper(inputs, state, sel_mem) + return self._cell(new_inputs, state) + + @staticmethod + def get_double_linear_controller(size, bias, input_keep_prob=1.0, is_train=None): + def double_linear_controller(inputs, state, memory): + """ + + :param inputs: [N, i] + :param state: [N, d] + :param memory: [N, M, m] + :return: [N, M] + """ + rank = len(memory.get_shape()) + _memory_size = tf.shape(memory)[rank-2] + tiled_inputs = tf.tile(tf.expand_dims(inputs, 1), [1, _memory_size, 1]) + if isinstance(state, tuple): + tiled_states = [tf.tile(tf.expand_dims(each, 1), [1, _memory_size, 1]) + for each in state] + else: + tiled_states = [tf.tile(tf.expand_dims(state, 1), [1, _memory_size, 1])] + + # [N, M, d] + in_ = tf.concat([tiled_inputs] + tiled_states + [memory], axis=2) + out = double_linear_logits(in_, size, bias, input_keep_prob=input_keep_prob, + is_train=is_train) + return out + return double_linear_controller + + @staticmethod + def get_linear_controller(bias, input_keep_prob=1.0, is_train=None): + def linear_controller(inputs, state, memory): + rank = len(memory.get_shape()) + _memory_size = tf.shape(memory)[rank-2] + tiled_inputs = tf.tile(tf.expand_dims(inputs, 1), [1, _memory_size, 1]) + if isinstance(state, tuple): + tiled_states = [tf.tile(tf.expand_dims(each, 1), [1, _memory_size, 1]) + for each in state] + else: + tiled_states = [tf.tile(tf.expand_dims(state, 1), [1, _memory_size, 1])] + + # [N, M, d] + in_ = tf.concat([tiled_inputs] + tiled_states + [memory], axis=2) + out = linear(in_, 1, bias, squeeze=True, input_keep_prob=input_keep_prob, is_train=is_train) + return out + return linear_controller + + @staticmethod + def get_concat_mapper(): + def concat_mapper(inputs, state, sel_mem): + """ + + :param inputs: [N, i] + :param state: [N, d] + :param sel_mem: [N, m] + :return: (new_inputs, new_state) tuple + """ + return tf.concat(axis=1, values=[inputs, sel_mem]), state + return concat_mapper + + @staticmethod + def get_sim_mapper(): + def sim_mapper(inputs, state, sel_mem): + """ + Assume that inputs and sel_mem are the same size + :param inputs: [N, i] + :param state: [N, d] + :param sel_mem: [N, i] + :return: (new_inputs, new_state) tuple + """ + return tf.concat(axis=1, values=[inputs, sel_mem, inputs * sel_mem, tf.abs(inputs - sel_mem)]), state + return sim_mapper diff --git a/tensorflow/SQuAD/my/utils.py b/tensorflow/SQuAD/my/utils.py new file mode 100644 index 0000000..9f3161c --- /dev/null +++ b/tensorflow/SQuAD/my/utils.py @@ -0,0 +1,58 @@ +import json +from collections import deque + +import numpy as np +from tqdm import tqdm + + +def mytqdm(list_, desc="", show=True): + if show: + pbar = tqdm(list_) + pbar.set_description(desc) + return pbar + return list_ + + +def json_pretty_dump(obj, fh): + return json.dump(obj, fh, sort_keys=True, indent=2, separators=(',', ': ')) + + +def index(l, i): + return index(l[i[0]], i[1:]) if len(i) > 1 else l[i[0]] + + +def fill(l, shape, dtype=None): + out = np.zeros(shape, dtype=dtype) + stack = deque() + stack.appendleft(((), l)) + while len(stack) > 0: + indices, cur = stack.pop() + if len(indices) < shape: + for i, sub in enumerate(cur): + stack.appendleft([indices + (i,), sub]) + else: + out[indices] = cur + return out + + +def short_floats(o, precision): + class ShortFloat(float): + def __repr__(self): + return '%.{}g'.format(precision) % self + + def _short_floats(obj): + if isinstance(obj, float): + return ShortFloat(obj) + elif isinstance(obj, dict): + return dict((k, _short_floats(v)) for k, v in obj.items()) + elif isinstance(obj, (list, tuple)): + return tuple(map(_short_floats, obj)) + return obj + + return _short_floats(o) + + +def argmax(x): + return np.unravel_index(x.argmax(), x.shape) + + diff --git a/tensorflow/SQuAD/my/zip_save.py b/tensorflow/SQuAD/my/zip_save.py new file mode 100644 index 0000000..5696423 --- /dev/null +++ b/tensorflow/SQuAD/my/zip_save.py @@ -0,0 +1,50 @@ +import argparse +import os + +import shutil +from zipfile import ZipFile + +from tqdm import tqdm + + +def get_args(): + parser = argparse.ArgumentParser() + parser.add_argument('paths', nargs='+') + parser.add_argument('-o', '--out', default='save.zip') + args = parser.parse_args() + return args + + +def zip_save(args): + temp_dir = "." + save_dir = os.path.join(temp_dir, "save") + if not os.path.exists(save_dir): + os.makedirs(save_dir) + for save_source_path in tqdm(args.paths): + # path = "out/basic/30/save/basic-18000" + # target_path = "save_dir/30/save" + # also output full path name to "save_dir/30/readme.txt + # need to also extract "out/basic/30/shared.json" + temp, _ = os.path.split(save_source_path) # "out/basic/30/save", _ + model_dir, _ = os.path.split(temp) # "out/basic/30, _ + _, model_name = os.path.split(model_dir) + cur_dir = os.path.join(save_dir, model_name) + if not os.path.exists(cur_dir): + os.makedirs(cur_dir) + save_target_path = os.path.join(cur_dir, "save") + shared_target_path = os.path.join(cur_dir, "shared.json") + readme_path = os.path.join(cur_dir, "readme.txt") + shared_source_path = os.path.join(model_dir, "shared.json") + shutil.copy(save_source_path, save_target_path) + shutil.copy(shared_source_path, shared_target_path) + with open(readme_path, 'w') as fh: + fh.write(save_source_path) + + os.system("zip {} -r {}".format(args.out, save_dir)) + +def main(): + args = get_args() + zip_save(args) + +if __name__ == "__main__": + main() diff --git a/tensorflow/SQuAD/requirements.txt b/tensorflow/SQuAD/requirements.txt new file mode 100644 index 0000000..7246e34 --- /dev/null +++ b/tensorflow/SQuAD/requirements.txt @@ -0,0 +1,3 @@ +nltk +tqdm +jinja2 diff --git a/tensorflow/SQuAD/run_training.sh b/tensorflow/SQuAD/run_training.sh new file mode 100755 index 0000000..2259acf --- /dev/null +++ b/tensorflow/SQuAD/run_training.sh @@ -0,0 +1 @@ +python3 -m basic.cli --mode train --noload --len_opt --cluster diff --git a/tensorflow/SQuAD/squad/__init__.py b/tensorflow/SQuAD/squad/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/tensorflow/SQuAD/squad/aug_squad.py b/tensorflow/SQuAD/squad/aug_squad.py new file mode 100644 index 0000000..c173fe1 --- /dev/null +++ b/tensorflow/SQuAD/squad/aug_squad.py @@ -0,0 +1,157 @@ +import json +import sys + +from tqdm import tqdm + +from my.corenlp_interface import CoreNLPInterface + +in_path = sys.argv[1] +out_path = sys.argv[2] +url = sys.argv[3] +port = int(sys.argv[4]) +data = json.load(open(in_path, 'r')) + +h = CoreNLPInterface(url, port) + + +def find_all(a_str, sub): + start = 0 + while True: + start = a_str.find(sub, start) + if start == -1: return + yield start + start += len(sub) # use start += 1 to find overlapping matches + + +def to_hex(s): + return " ".join(map(hex, map(ord, s))) + + +def handle_nobreak(cand, text): + if cand == text: + return cand + if cand.replace(u'\u00A0', ' ') == text: + return cand + elif cand == text.replace(u'\u00A0', ' '): + return text + raise Exception("{} '{}' {} '{}'".format(cand, to_hex(cand), text, to_hex(text))) + + +# resolving unicode complication + +wrong_loc_count = 0 +loc_diffs = [] + +for article in data['data']: + for para in article['paragraphs']: + para['context'] = para['context'].replace(u'\u000A', '') + para['context'] = para['context'].replace(u'\u00A0', ' ') + context = para['context'] + for qa in para['qas']: + for answer in qa['answers']: + answer['text'] = answer['text'].replace(u'\u00A0', ' ') + text = answer['text'] + answer_start = answer['answer_start'] + if context[answer_start:answer_start + len(text)] == text: + if text.lstrip() == text: + pass + else: + answer_start += len(text) - len(text.lstrip()) + answer['answer_start'] = answer_start + text = text.lstrip() + answer['text'] = text + else: + wrong_loc_count += 1 + text = text.lstrip() + answer['text'] = text + starts = list(find_all(context, text)) + if len(starts) == 1: + answer_start = starts[0] + elif len(starts) > 1: + new_answer_start = min(starts, key=lambda s: abs(s - answer_start)) + loc_diffs.append(abs(new_answer_start - answer_start)) + answer_start = new_answer_start + else: + raise Exception() + answer['answer_start'] = answer_start + + answer_stop = answer_start + len(text) + answer['answer_stop'] = answer_stop + assert para['context'][answer_start:answer_stop] == answer['text'], "{} {}".format( + para['context'][answer_start:answer_stop], answer['text']) + +print(wrong_loc_count, loc_diffs) + +mismatch_count = 0 +dep_fail_count = 0 +no_answer_count = 0 + +size = sum(len(article['paragraphs']) for article in data['data']) +pbar = tqdm(range(size)) + +for ai, article in enumerate(data['data']): + for pi, para in enumerate(article['paragraphs']): + context = para['context'] + sents = h.split_doc(context) + words = h.split_sent(context) + sent_starts = [] + ref_idx = 0 + for sent in sents: + new_idx = context.find(sent, ref_idx) + sent_starts.append(new_idx) + ref_idx = new_idx + len(sent) + para['sents'] = sents + para['words'] = words + para['sent_starts'] = sent_starts + + consts = list(map(h.get_const, sents)) + para['consts'] = consts + deps = list(map(h.get_dep, sents)) + para['deps'] = deps + + for qa in para['qas']: + question = qa['question'] + question_const = h.get_const(question) + qa['const'] = question_const + question_dep = h.get_dep(question) + qa['dep'] = question_dep + qa['words'] = h.split_sent(question) + + for answer in qa['answers']: + answer_start = answer['answer_start'] + text = answer['text'] + answer_stop = answer_start + len(text) + # answer_words = h.split_sent(text) + word_idxs = [] + answer_words = [] + for sent_idx, (sent, sent_start, dep) in enumerate(zip(sents, sent_starts, deps)): + if dep is None: + print("dep parse failed at {} {} {}".format(ai, pi, sent_idx)) + dep_fail_count += 1 + continue + nodes, edges = dep + words = [node[0] for node in nodes] + + for word_idx, (word, _, _, start, _) in enumerate(nodes): + global_start = sent_start + start + global_stop = global_start + len(word) + if answer_start <= global_start < answer_stop or answer_start < global_stop <= answer_stop: + word_idxs.append((sent_idx, word_idx)) + answer_words.append(word) + if len(word_idxs) > 0: + answer['answer_word_start'] = word_idxs[0] + answer['answer_word_stop'] = word_idxs[-1][0], word_idxs[-1][1] + 1 + if not text.startswith(answer_words[0]): + print("'{}' '{}'".format(text, ' '.join(answer_words))) + mismatch_count += 1 + else: + answer['answer_word_start'] = None + answer['answer_word_stop'] = None + no_answer_count += 1 + pbar.update(1) +pbar.close() + +print(mismatch_count, dep_fail_count, no_answer_count) + +print("saving...") +json.dump(data, open(out_path, 'w')) \ No newline at end of file diff --git a/tensorflow/SQuAD/squad/eda_aug_dev.ipynb b/tensorflow/SQuAD/squad/eda_aug_dev.ipynb new file mode 100644 index 0000000..1cfefd9 --- /dev/null +++ b/tensorflow/SQuAD/squad/eda_aug_dev.ipynb @@ -0,0 +1,271 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [ + "import json\n", + "\n", + "aug_data_path = \"/Users/minjoons/data/squad/dev-v1.0-aug.json\"\n", + "aug_data = json.load(open(aug_data_path, 'r'))" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "(['Denver', 'Broncos'], 'Denver Broncos')\n", + "(['Denver', 'Broncos'], 'Denver Broncos')\n", + "(['Denver', 'Broncos'], 'Denver Broncos ')\n", + "(['Carolina', 'Panthers'], 'Carolina Panthers')\n" + ] + } + ], + "source": [ + "def compare_answers():\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " deps = para['deps']\n", + " nodess = []\n", + " for dep in deps:\n", + " nodes, edges = dep\n", + " if dep is not None:\n", + " nodess.append(nodes)\n", + " else:\n", + " nodess.append([])\n", + " wordss = [[node[0] for node in nodes] for nodes in nodess]\n", + " for qa in para['qas']:\n", + " for answer in qa['answers']:\n", + " text = answer['text']\n", + " word_start = answer['answer_word_start']\n", + " word_stop = answer['answer_word_stop']\n", + " answer_words = wordss[word_start[0]][word_start[1]:word_stop[1]]\n", + " yield answer_words, text\n", + "\n", + "ca = compare_answers()\n", + "print(next(ca))\n", + "print(next(ca))\n", + "print(next(ca))\n", + "print(next(ca))" + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "8\n" + ] + } + ], + "source": [ + "def counter():\n", + " count = 0\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " deps = para['deps']\n", + " nodess = []\n", + " for dep in deps:\n", + " if dep is None:\n", + " count += 1\n", + " print(count)\n", + "counter()\n" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "0\n" + ] + } + ], + "source": [ + "def bad_node_counter():\n", + " count = 0\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " sents = para['sents']\n", + " deps = para['deps']\n", + " nodess = []\n", + " for dep in deps:\n", + " if dep is not None:\n", + " nodes, edges = dep\n", + " for node in nodes:\n", + " if len(node) != 5:\n", + " count += 1\n", + " print(count)\n", + "bad_node_counter() " + ] + }, + { + "cell_type": "code", + "execution_count": 20, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "7\n" + ] + } + ], + "source": [ + "def noanswer_counter():\n", + " count = 0\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " deps = para['deps']\n", + " nodess = []\n", + " for dep in deps:\n", + " if dep is not None:\n", + " nodes, edges = dep\n", + " nodess.append(nodes)\n", + " else:\n", + " nodess.append([])\n", + " wordss = [[node[0] for node in nodes] for nodes in nodess]\n", + " for qa in para['qas']:\n", + " for answer in qa['answers']:\n", + " text = answer['text']\n", + " word_start = answer['answer_word_start']\n", + " word_stop = answer['answer_word_stop']\n", + " if word_start is None:\n", + " count += 1\n", + " print(count)\n", + "noanswer_counter()" + ] + }, + { + "cell_type": "code", + "execution_count": 22, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "10600\n" + ] + } + ], + "source": [ + "print(sum(len(para['qas']) for a in aug_data['data'] for para in a['paragraphs']))" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "10348\n" + ] + } + ], + "source": [ + "import nltk\n", + "\n", + "def _set_span(t, i):\n", + " if isinstance(t[0], str):\n", + " t.span = (i, i+len(t))\n", + " else:\n", + " first = True\n", + " for c in t:\n", + " cur_span = _set_span(c, i)\n", + " i = cur_span[1]\n", + " if first:\n", + " min_ = cur_span[0]\n", + " first = False\n", + " max_ = cur_span[1]\n", + " t.span = (min_, max_)\n", + " return t.span\n", + "\n", + "\n", + "def set_span(t):\n", + " assert isinstance(t, nltk.tree.Tree)\n", + " try:\n", + " return _set_span(t, 0)\n", + " except:\n", + " print(t)\n", + " exit()\n", + "\n", + "def same_span_counter():\n", + " count = 0\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " consts = para['consts']\n", + " for const in consts:\n", + " tree = nltk.tree.Tree.fromstring(const)\n", + " set_span(tree)\n", + " if len(list(tree.subtrees())) > len(set(t.span for t in tree.subtrees())):\n", + " count += 1\n", + " print(count)\n", + "same_span_counter()" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.5.1" + } + }, + "nbformat": 4, + "nbformat_minor": 0 +} diff --git a/tensorflow/SQuAD/squad/eda_aug_train.ipynb b/tensorflow/SQuAD/squad/eda_aug_train.ipynb new file mode 100644 index 0000000..7780024 --- /dev/null +++ b/tensorflow/SQuAD/squad/eda_aug_train.ipynb @@ -0,0 +1,314 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [ + "import json\n", + "\n", + "aug_data_path = \"/Users/minjoons/data/squad/train-v1.0-aug.json\"\n", + "aug_data = json.load(open(aug_data_path, 'r'))" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "(['Saint', 'Bernadette', 'Soubirous'], 'Saint Bernadette Soubirous')\n", + "(['a', 'copper', 'statue', 'of', 'Christ'], 'a copper statue of Christ')\n", + "(['the', 'Main', 'Building'], 'the Main Building')\n", + "(['a', 'Marian', 'place', 'of', 'prayer', 'and', 'reflection'], 'a Marian place of prayer and reflection')\n" + ] + } + ], + "source": [ + "def compare_answers():\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " deps = para['deps']\n", + " nodess = []\n", + " for dep in deps:\n", + " nodes, edges = dep\n", + " if dep is not None:\n", + " nodess.append(nodes)\n", + " else:\n", + " nodess.append([])\n", + " wordss = [[node[0] for node in nodes] for nodes in nodess]\n", + " for qa in para['qas']:\n", + " for answer in qa['answers']:\n", + " text = answer['text']\n", + " word_start = answer['answer_word_start']\n", + " word_stop = answer['answer_word_stop']\n", + " answer_words = wordss[word_start[0]][word_start[1]:word_stop[1]]\n", + " yield answer_words, text\n", + "\n", + "ca = compare_answers()\n", + "print(next(ca))\n", + "print(next(ca))\n", + "print(next(ca))\n", + "print(next(ca))" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "q: k\n", + "q: j\n", + "q: n\n", + "q: b\n", + "q: v\n", + "x: .\n", + "x: :208\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "q: dd\n", + "q: dd\n", + "q: dd\n", + "q: dd\n", + "q: d\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: :411\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: :40\n", + "x: .\n", + "x: *\n", + "x: :14\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: :131\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "x: .\n", + "53 10\n" + ] + } + ], + "source": [ + "def nodep_counter():\n", + " x_count = 0\n", + " q_count = 0\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " deps = para['deps']\n", + " nodess = []\n", + " for sent, dep in zip(para['sents'], deps):\n", + " if dep is None:\n", + " print(\"x:\", sent)\n", + " x_count += 1\n", + " for qa in para['qas']:\n", + " if qa['dep'] is None:\n", + " print(\"q:\", qa['question'])\n", + " q_count += 1\n", + " print(x_count, q_count)\n", + "nodep_counter()\n" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "0\n" + ] + } + ], + "source": [ + "def bad_node_counter():\n", + " count = 0\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " sents = para['sents']\n", + " deps = para['deps']\n", + " nodess = []\n", + " for dep in deps:\n", + " if dep is not None:\n", + " nodes, edges = dep\n", + " for node in nodes:\n", + " if len(node) != 5:\n", + " count += 1\n", + " print(count)\n", + "bad_node_counter() " + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "36\n" + ] + } + ], + "source": [ + "def noanswer_counter():\n", + " count = 0\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " deps = para['deps']\n", + " nodess = []\n", + " for dep in deps:\n", + " if dep is not None:\n", + " nodes, edges = dep\n", + " nodess.append(nodes)\n", + " else:\n", + " nodess.append([])\n", + " wordss = [[node[0] for node in nodes] for nodes in nodess]\n", + " for qa in para['qas']:\n", + " for answer in qa['answers']:\n", + " text = answer['text']\n", + " word_start = answer['answer_word_start']\n", + " word_stop = answer['answer_word_stop']\n", + " if word_start is None:\n", + " count += 1\n", + " print(count)\n", + "noanswer_counter()" + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "106\n" + ] + } + ], + "source": [ + "def mult_sent_answer_counter():\n", + " count = 0\n", + " for article in aug_data['data']:\n", + " for para in article['paragraphs']:\n", + " for qa in para['qas']:\n", + " for answer in qa['answers']:\n", + " text = answer['text']\n", + " word_start = answer['answer_word_start']\n", + " word_stop = answer['answer_word_stop']\n", + " if word_start is not None and word_start[0] != word_stop[0]:\n", + " count += 1\n", + " print(count)\n", + "mult_sent_answer_counter()" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.5.1" + } + }, + "nbformat": 4, + "nbformat_minor": 0 +} diff --git a/tensorflow/SQuAD/squad/evaluate-v1.1.py b/tensorflow/SQuAD/squad/evaluate-v1.1.py new file mode 100644 index 0000000..0137fbc --- /dev/null +++ b/tensorflow/SQuAD/squad/evaluate-v1.1.py @@ -0,0 +1,94 @@ +""" Official evaluation script for v1.1 of the SQuAD dataset. """ +from __future__ import print_function +from collections import Counter +import string +import re +import argparse +import json +import sys + + +def normalize_answer(s): + """Lower text and remove punctuation, articles and extra whitespace.""" + def remove_articles(text): + return re.sub(r'\b(a|an|the)\b', ' ', text) + + def white_space_fix(text): + return ' '.join(text.split()) + + def remove_punc(text): + exclude = set(string.punctuation) + return ''.join(ch for ch in text if ch not in exclude) + + def lower(text): + return text.lower() + + return white_space_fix(remove_articles(remove_punc(lower(s)))) + + +def f1_score(prediction, ground_truth): + prediction_tokens = normalize_answer(prediction).split() + ground_truth_tokens = normalize_answer(ground_truth).split() + common = Counter(prediction_tokens) & Counter(ground_truth_tokens) + num_same = sum(common.values()) + if num_same == 0: + return 0 + precision = 1.0 * num_same / len(prediction_tokens) + recall = 1.0 * num_same / len(ground_truth_tokens) + f1 = (2 * precision * recall) / (precision + recall) + return f1 + + +def exact_match_score(prediction, ground_truth): + return (normalize_answer(prediction) == normalize_answer(ground_truth)) + + +def metric_max_over_ground_truths(metric_fn, prediction, ground_truths): + scores_for_ground_truths = [] + for ground_truth in ground_truths: + score = metric_fn(prediction, ground_truth) + scores_for_ground_truths.append(score) + return max(scores_for_ground_truths) + + +def evaluate(dataset, predictions): + f1 = exact_match = total = 0 + for article in dataset: + for paragraph in article['paragraphs']: + for qa in paragraph['qas']: + total += 1 + if qa['id'] not in predictions: + message = 'Unanswered question ' + qa['id'] + \ + ' will receive score 0.' + print(message, file=sys.stderr) + continue + ground_truths = list(map(lambda x: x['text'], qa['answers'])) + prediction = predictions[qa['id']] + exact_match += metric_max_over_ground_truths( + exact_match_score, prediction, ground_truths) + f1 += metric_max_over_ground_truths( + f1_score, prediction, ground_truths) + + exact_match = 100.0 * exact_match / total + f1 = 100.0 * f1 / total + + return {'exact_match': exact_match, 'f1': f1} + + +if __name__ == '__main__': + expected_version = '1.1' + parser = argparse.ArgumentParser( + description='Evaluation for SQuAD ' + expected_version) + parser.add_argument('dataset_file', help='Dataset file') + parser.add_argument('prediction_file', help='Prediction File') + args = parser.parse_args() + with open(args.dataset_file) as dataset_file: + dataset_json = json.load(dataset_file) + if (dataset_json['version'] != expected_version): + print('Evaluation expects v-' + expected_version + + ', but got dataset with v-' + dataset_json['version'], + file=sys.stderr) + dataset = dataset_json['data'] + with open(args.prediction_file) as prediction_file: + predictions = json.load(prediction_file) + print(json.dumps(evaluate(dataset, predictions))) diff --git a/tensorflow/SQuAD/squad/evaluate.py b/tensorflow/SQuAD/squad/evaluate.py new file mode 100644 index 0000000..2bf15f0 --- /dev/null +++ b/tensorflow/SQuAD/squad/evaluate.py @@ -0,0 +1,94 @@ +""" Official evaluation script for v1.1 of the SQuAD dataset. [Changed name for external importing]""" +from __future__ import print_function +from collections import Counter +import string +import re +import argparse +import json +import sys + + +def normalize_answer(s): + """Lower text and remove punctuation, articles and extra whitespace.""" + def remove_articles(text): + return re.sub(r'\b(a|an|the)\b', ' ', text) + + def white_space_fix(text): + return ' '.join(text.split()) + + def remove_punc(text): + exclude = set(string.punctuation) + return ''.join(ch for ch in text if ch not in exclude) + + def lower(text): + return text.lower() + + return white_space_fix(remove_articles(remove_punc(lower(s)))) + + +def f1_score(prediction, ground_truth): + prediction_tokens = normalize_answer(prediction).split() + ground_truth_tokens = normalize_answer(ground_truth).split() + common = Counter(prediction_tokens) & Counter(ground_truth_tokens) + num_same = sum(common.values()) + if num_same == 0: + return 0 + precision = 1.0 * num_same / len(prediction_tokens) + recall = 1.0 * num_same / len(ground_truth_tokens) + f1 = (2 * precision * recall) / (precision + recall) + return f1 + + +def exact_match_score(prediction, ground_truth): + return (normalize_answer(prediction) == normalize_answer(ground_truth)) + + +def metric_max_over_ground_truths(metric_fn, prediction, ground_truths): + scores_for_ground_truths = [] + for ground_truth in ground_truths: + score = metric_fn(prediction, ground_truth) + scores_for_ground_truths.append(score) + return max(scores_for_ground_truths) + + +def evaluate(dataset, predictions): + f1 = exact_match = total = 0 + for article in dataset: + for paragraph in article['paragraphs']: + for qa in paragraph['qas']: + total += 1 + if qa['id'] not in predictions: + message = 'Unanswered question ' + qa['id'] + \ + ' will receive score 0.' + print(message, file=sys.stderr) + continue + ground_truths = list(map(lambda x: x['text'], qa['answers'])) + prediction = predictions[qa['id']] + exact_match += metric_max_over_ground_truths( + exact_match_score, prediction, ground_truths) + f1 += metric_max_over_ground_truths( + f1_score, prediction, ground_truths) + + exact_match = 100.0 * exact_match / total + f1 = 100.0 * f1 / total + + return {'exact_match': exact_match, 'f1': f1} + + +if __name__ == '__main__': + expected_version = '1.1' + parser = argparse.ArgumentParser( + description='Evaluation for SQuAD ' + expected_version) + parser.add_argument('dataset_file', help='Dataset file') + parser.add_argument('prediction_file', help='Prediction File') + args = parser.parse_args() + with open(args.dataset_file) as dataset_file: + dataset_json = json.load(dataset_file) + if (dataset_json['version'] != expected_version): + print('Evaluation expects v-' + expected_version + + ', but got dataset with v-' + dataset_json['version'], + file=sys.stderr) + dataset = dataset_json['data'] + with open(args.prediction_file) as prediction_file: + predictions = json.load(prediction_file) + print(json.dumps(evaluate(dataset, predictions))) diff --git a/tensorflow/SQuAD/squad/neg_squad.py b/tensorflow/SQuAD/squad/neg_squad.py new file mode 100644 index 0000000..2bba23d --- /dev/null +++ b/tensorflow/SQuAD/squad/neg_squad.py @@ -0,0 +1,50 @@ +import argparse +import json +import os +# data: q, cq, (dq), (pq), y, *x, *cx +# shared: x, cx, (dx), (px), word_counter, char_counter, word2vec +# no metadata +import random +from collections import Counter + +from tqdm import tqdm + +from squad.utils import get_word_span, get_word_idx, process_tokens + + +def main(): + args = get_args() + neg_squad(args) + + +def get_args(): + parser = argparse.ArgumentParser() + home = os.path.expanduser("~") + parser.add_argument("source_path") + parser.add_argument("target_path") + parser.add_argument('-d', "--debug", action='store_true') + parser.add_argument('-r', "--aug_ratio", default=1, type=int) + # TODO : put more args here + return parser.parse_args() + + +def neg_squad(args): + with open(args.source_path, 'r') as fp: + squad = json.load(fp) + with open(args.source_path, 'r') as fp: + ref_squad = json.load(fp) + + for ai, article in enumerate(ref_squad['data']): + for pi, para in enumerate(article['paragraphs']): + cands = list(range(pi)) + list(range(pi+1, len(article['paragraphs']))) + samples = random.sample(cands, args.aug_ratio) + for sample in samples: + for qi, ques in enumerate(article['paragraphs'][sample]['qas']): + new_ques = {'question': ques['question'], 'answers': [], 'answer_start': 0, 'id': "neg_" + ques['id']} + squad['data'][ai]['paragraphs'][pi]['qas'].append(new_ques) + + with open(args.target_path, 'w') as fp: + json.dump(squad, fp) + +if __name__ == "__main__": + main() \ No newline at end of file diff --git a/tensorflow/SQuAD/squad/prepro.py b/tensorflow/SQuAD/squad/prepro.py new file mode 100644 index 0000000..a3c8f45 --- /dev/null +++ b/tensorflow/SQuAD/squad/prepro.py @@ -0,0 +1,241 @@ +import argparse +import json +import os +# data: q, cq, (dq), (pq), y, *x, *cx +# shared: x, cx, (dx), (px), word_counter, char_counter, word2vec +# no metadata +from collections import Counter + +from tqdm import tqdm + +from squad.utils import get_word_span, get_word_idx, process_tokens + + +def main(): + args = get_args() + prepro(args) + + +def get_args(): + parser = argparse.ArgumentParser() + home = os.path.expanduser("~") + source_dir = os.path.join(home, "data", "squad") + target_dir = "data/squad" + glove_dir = os.path.join(home, "data", "glove") + parser.add_argument('-s', "--source_dir", default=source_dir) + parser.add_argument('-t', "--target_dir", default=target_dir) + parser.add_argument("--train_name", default='train-v1.1.json') + parser.add_argument('-d', "--debug", action='store_true') + parser.add_argument("--train_ratio", default=0.9, type=int) + parser.add_argument("--glove_corpus", default="6B") + parser.add_argument("--glove_dir", default=glove_dir) + parser.add_argument("--glove_vec_size", default=100, type=int) + parser.add_argument("--mode", default="full", type=str) + parser.add_argument("--single_path", default="", type=str) + parser.add_argument("--tokenizer", default="PTB", type=str) + parser.add_argument("--url", default="vision-server2.corp.ai2", type=str) + parser.add_argument("--port", default=8000, type=int) + parser.add_argument("--split", action='store_true') + parser.add_argument("--suffix", default="") + # TODO : put more args here + return parser.parse_args() + + +def create_all(args): + out_path = os.path.join(args.source_dir, "all-v1.1.json") + if os.path.exists(out_path): + return + train_path = os.path.join(args.source_dir, args.train_name) + train_data = json.load(open(train_path, 'r')) + dev_path = os.path.join(args.source_dir, args.dev_name) + dev_data = json.load(open(dev_path, 'r')) + train_data['data'].extend(dev_data['data']) + print("dumping all data ...") + json.dump(train_data, open(out_path, 'w')) + + +def prepro(args): + if not os.path.exists(args.target_dir): + os.makedirs(args.target_dir) + + if args.mode == 'full': + prepro_each(args, 'train', out_name='train') + prepro_each(args, 'dev', out_name='dev') + prepro_each(args, 'dev', out_name='test') + elif args.mode == 'all': + create_all(args) + prepro_each(args, 'dev', 0.0, 0.0, out_name='dev') + prepro_each(args, 'dev', 0.0, 0.0, out_name='test') + prepro_each(args, 'all', out_name='train') + elif args.mode == 'single': + assert len(args.single_path) > 0 + prepro_each(args, "NULL", out_name="single", in_path=args.single_path) + else: + prepro_each(args, 'train', 0.0, args.train_ratio, out_name='train') + prepro_each(args, 'train', args.train_ratio, 1.0, out_name='dev') + prepro_each(args, 'dev', out_name='test') + + +def save(args, data, shared, data_type): + data_path = os.path.join(args.target_dir, "data_{}.json".format(data_type)) + shared_path = os.path.join(args.target_dir, "shared_{}.json".format(data_type)) + json.dump(data, open(data_path, 'w')) + json.dump(shared, open(shared_path, 'w')) + + +def get_word2vec(args, word_counter): + glove_path = os.path.join(args.glove_dir, "glove.{}.{}d.txt".format(args.glove_corpus, args.glove_vec_size)) + sizes = {'6B': int(4e5), '42B': int(1.9e6), '840B': int(2.2e6), '2B': int(1.2e6)} + total = sizes[args.glove_corpus] + word2vec_dict = {} + with open(glove_path, 'r', encoding='utf-8') as fh: + for line in tqdm(fh, total=total): + array = line.lstrip().rstrip().split(" ") + word = array[0] + vector = list(map(float, array[1:])) + if word in word_counter: + word2vec_dict[word] = vector + elif word.capitalize() in word_counter: + word2vec_dict[word.capitalize()] = vector + elif word.lower() in word_counter: + word2vec_dict[word.lower()] = vector + elif word.upper() in word_counter: + word2vec_dict[word.upper()] = vector + + print("{}/{} of word vocab have corresponding vectors in {}".format(len(word2vec_dict), len(word_counter), glove_path)) + return word2vec_dict + + +def prepro_each(args, data_type, start_ratio=0.0, stop_ratio=1.0, out_name="default", in_path=None): + if args.tokenizer == "PTB": + import nltk + sent_tokenize = nltk.sent_tokenize + def word_tokenize(tokens): + return [token.replace("''", '"').replace("``", '"') for token in nltk.word_tokenize(tokens)] + elif args.tokenizer == 'Stanford': + from my.corenlp_interface import CoreNLPInterface + interface = CoreNLPInterface(args.url, args.port) + sent_tokenize = interface.split_doc + word_tokenize = interface.split_sent + else: + raise Exception() + + if not args.split: + sent_tokenize = lambda para: [para] + + source_path = in_path or os.path.join(args.source_dir, "{}-{}v1.1.json".format(data_type, args.suffix)) + source_data = json.load(open(source_path, 'r')) + + q, cq, y, rx, rcx, ids, idxs = [], [], [], [], [], [], [] + na = [] + cy = [] + x, cx = [], [] + answerss = [] + p = [] + word_counter, char_counter, lower_word_counter = Counter(), Counter(), Counter() + start_ai = int(round(len(source_data['data']) * start_ratio)) + stop_ai = int(round(len(source_data['data']) * stop_ratio)) + for ai, article in enumerate(tqdm(source_data['data'][start_ai:stop_ai])): + xp, cxp = [], [] + pp = [] + x.append(xp) + cx.append(cxp) + p.append(pp) + for pi, para in enumerate(article['paragraphs']): + # wordss + context = para['context'] + context = context.replace("''", '" ') + context = context.replace("``", '" ') + xi = list(map(word_tokenize, sent_tokenize(context))) + xi = [process_tokens(tokens) for tokens in xi] # process tokens + # given xi, add chars + cxi = [[list(xijk) for xijk in xij] for xij in xi] + xp.append(xi) + cxp.append(cxi) + pp.append(context) + + for xij in xi: + for xijk in xij: + word_counter[xijk] += len(para['qas']) + lower_word_counter[xijk.lower()] += len(para['qas']) + for xijkl in xijk: + char_counter[xijkl] += len(para['qas']) + + rxi = [ai, pi] + assert len(x) - 1 == ai + assert len(x[ai]) - 1 == pi + for qa in para['qas']: + # get words + qi = word_tokenize(qa['question']) + qi = process_tokens(qi) + cqi = [list(qij) for qij in qi] + yi = [] + cyi = [] + answers = [] + for answer in qa['answers']: + answer_text = answer['text'] + answers.append(answer_text) + answer_start = answer['answer_start'] + answer_stop = answer_start + len(answer_text) + # TODO : put some function that gives word_start, word_stop here + yi0, yi1 = get_word_span(context, xi, answer_start, answer_stop) + # yi0 = answer['answer_word_start'] or [0, 0] + # yi1 = answer['answer_word_stop'] or [0, 1] + assert len(xi[yi0[0]]) > yi0[1] + assert len(xi[yi1[0]]) >= yi1[1] + w0 = xi[yi0[0]][yi0[1]] + w1 = xi[yi1[0]][yi1[1]-1] + i0 = get_word_idx(context, xi, yi0) + i1 = get_word_idx(context, xi, (yi1[0], yi1[1]-1)) + cyi0 = answer_start - i0 + cyi1 = answer_stop - i1 - 1 + # print(answer_text, w0[cyi0:], w1[:cyi1+1]) + assert answer_text[0] == w0[cyi0], (answer_text, w0, cyi0) + assert answer_text[-1] == w1[cyi1] + assert cyi0 < 32, (answer_text, w0) + assert cyi1 < 32, (answer_text, w1) + + yi.append([yi0, yi1]) + cyi.append([cyi0, cyi1]) + + if len(qa['answers']) == 0: + yi.append([(0, 0), (0, 1)]) + cyi.append([0, 1]) + na.append(True) + else: + na.append(False) + + for qij in qi: + word_counter[qij] += 1 + lower_word_counter[qij.lower()] += 1 + for qijk in qij: + char_counter[qijk] += 1 + + q.append(qi) + cq.append(cqi) + y.append(yi) + cy.append(cyi) + rx.append(rxi) + rcx.append(rxi) + ids.append(qa['id']) + idxs.append(len(idxs)) + answerss.append(answers) + + if args.debug: + break + + word2vec_dict = get_word2vec(args, word_counter) + lower_word2vec_dict = get_word2vec(args, lower_word_counter) + + # add context here + data = {'q': q, 'cq': cq, 'y': y, '*x': rx, '*cx': rcx, 'cy': cy, + 'idxs': idxs, 'ids': ids, 'answerss': answerss, '*p': rx, 'na': na} + shared = {'x': x, 'cx': cx, 'p': p, + 'word_counter': word_counter, 'char_counter': char_counter, 'lower_word_counter': lower_word_counter, + 'word2vec': word2vec_dict, 'lower_word2vec': lower_word2vec_dict} + + print("saving ...") + save(args, data, shared, out_name) + +if __name__ == "__main__": + main() \ No newline at end of file diff --git a/tensorflow/SQuAD/squad/prepro_aug.py b/tensorflow/SQuAD/squad/prepro_aug.py new file mode 100644 index 0000000..a3459d4 --- /dev/null +++ b/tensorflow/SQuAD/squad/prepro_aug.py @@ -0,0 +1,183 @@ +import argparse +import json +import os +# data: q, cq, (dq), (pq), y, *x, *cx +# shared: x, cx, (dx), (px), word_counter, char_counter, word2vec +# no metadata +from collections import Counter + +import nltk +from tqdm import tqdm + +from my.nltk_utils import load_compressed_tree + + +def bool_(arg): + if arg == 'True': + return True + elif arg == 'False': + return False + raise Exception() + + +def main(): + args = get_args() + prepro(args) + + +def get_args(): + parser = argparse.ArgumentParser() + home = os.path.expanduser("~") + source_dir = os.path.join(home, "data", "squad") + target_dir = "data/squad" + glove_dir = os.path.join(home, "data", "glove") + parser.add_argument("--source_dir", default=source_dir) + parser.add_argument("--target_dir", default=target_dir) + parser.add_argument("--debug", default=False, type=bool_) + parser.add_argument("--train_ratio", default=0.9, type=int) + parser.add_argument("--glove_corpus", default="6B") + parser.add_argument("--glove_dir", default=glove_dir) + parser.add_argument("--glove_vec_size", default=100, type=int) + parser.add_argument("--full_train", default=False, type=bool_) + # TODO : put more args here + return parser.parse_args() + + +def prepro(args): + if not os.path.exists(args.target_dir): + os.makedirs(args.target_dir) + + if args.full_train: + data_train, shared_train = prepro_each(args, 'train') + data_dev, shared_dev = prepro_each(args, 'dev') + else: + data_train, shared_train = prepro_each(args, 'train', 0.0, args.train_ratio) + data_dev, shared_dev = prepro_each(args, 'train', args.train_ratio, 1.0) + data_test, shared_test = prepro_each(args, 'dev') + + print("saving ...") + save(args, data_train, shared_train, 'train') + save(args, data_dev, shared_dev, 'dev') + save(args, data_test, shared_test, 'test') + + +def save(args, data, shared, data_type): + data_path = os.path.join(args.target_dir, "data_{}.json".format(data_type)) + shared_path = os.path.join(args.target_dir, "shared_{}.json".format(data_type)) + json.dump(data, open(data_path, 'w')) + json.dump(shared, open(shared_path, 'w')) + + +def get_word2vec(args, word_counter): + glove_path = os.path.join(args.glove_dir, "glove.{}.{}d.txt".format(args.glove_corpus, args.glove_vec_size)) + sizes = {'6B': int(4e5), '42B': int(1.9e6), '840B': int(2.2e6), '2B': int(1.2e6)} + total = sizes[args.glove_corpus] + word2vec_dict = {} + with open(glove_path, 'r') as fh: + for line in tqdm(fh, total=total): + array = line.lstrip().rstrip().split(" ") + word = array[0] + vector = list(map(float, array[1:])) + if word in word_counter: + word2vec_dict[word] = vector + elif word.capitalize() in word_counter: + word2vec_dict[word.capitalize()] = vector + elif word.lower() in word_counter: + word2vec_dict[word.lower()] = vector + elif word.upper() in word_counter: + word2vec_dict[word.upper()] = vector + + print("{}/{} of word vocab have corresponding vectors in {}".format(len(word2vec_dict), len(word_counter), glove_path)) + return word2vec_dict + + +def prepro_each(args, data_type, start_ratio=0.0, stop_ratio=1.0): + source_path = os.path.join(args.source_dir, "{}-v1.0-aug.json".format(data_type)) + source_data = json.load(open(source_path, 'r')) + + q, cq, y, rx, rcx, ids, idxs = [], [], [], [], [], [], [] + x, cx, tx, stx = [], [], [], [] + answerss = [] + word_counter, char_counter, lower_word_counter = Counter(), Counter(), Counter() + pos_counter = Counter() + start_ai = int(round(len(source_data['data']) * start_ratio)) + stop_ai = int(round(len(source_data['data']) * stop_ratio)) + for ai, article in enumerate(tqdm(source_data['data'][start_ai:stop_ai])): + xp, cxp, txp, stxp = [], [], [], [] + x.append(xp) + cx.append(cxp) + tx.append(txp) + stx.append(stxp) + for pi, para in enumerate(article['paragraphs']): + xi = [] + for dep in para['deps']: + if dep is None: + xi.append([]) + else: + xi.append([node[0] for node in dep[0]]) + cxi = [[list(xijk) for xijk in xij] for xij in xi] + xp.append(xi) + cxp.append(cxi) + txp.append(para['consts']) + stxp.append([str(load_compressed_tree(s)) for s in para['consts']]) + trees = map(nltk.tree.Tree.fromstring, para['consts']) + for tree in trees: + for subtree in tree.subtrees(): + pos_counter[subtree.label()] += 1 + + for xij in xi: + for xijk in xij: + word_counter[xijk] += len(para['qas']) + lower_word_counter[xijk.lower()] += len(para['qas']) + for xijkl in xijk: + char_counter[xijkl] += len(para['qas']) + + rxi = [ai, pi] + assert len(x) - 1 == ai + assert len(x[ai]) - 1 == pi + for qa in para['qas']: + dep = qa['dep'] + qi = [] if dep is None else [node[0] for node in dep[0]] + cqi = [list(qij) for qij in qi] + yi = [] + answers = [] + for answer in qa['answers']: + answers.append(answer['text']) + yi0 = answer['answer_word_start'] or [0, 0] + yi1 = answer['answer_word_stop'] or [0, 1] + assert len(xi[yi0[0]]) > yi0[1] + assert len(xi[yi1[0]]) >= yi1[1] + yi.append([yi0, yi1]) + + for qij in qi: + word_counter[qij] += 1 + lower_word_counter[qij.lower()] += 1 + for qijk in qij: + char_counter[qijk] += 1 + + q.append(qi) + cq.append(cqi) + y.append(yi) + rx.append(rxi) + rcx.append(rxi) + ids.append(qa['id']) + idxs.append(len(idxs)) + answerss.append(answers) + + if args.debug: + break + + word2vec_dict = get_word2vec(args, word_counter) + lower_word2vec_dict = get_word2vec(args, lower_word_counter) + + data = {'q': q, 'cq': cq, 'y': y, '*x': rx, '*cx': rcx, '*tx': rx, '*stx': rx, + 'idxs': idxs, 'ids': ids, 'answerss': answerss} + shared = {'x': x, 'cx': cx, 'tx': tx, 'stx': stx, + 'word_counter': word_counter, 'char_counter': char_counter, 'lower_word_counter': lower_word_counter, + 'word2vec': word2vec_dict, 'lower_word2vec': lower_word2vec_dict, 'pos_counter': pos_counter} + + return data, shared + + +if __name__ == "__main__": + main() \ No newline at end of file diff --git a/tensorflow/SQuAD/squad/utils.py b/tensorflow/SQuAD/squad/utils.py new file mode 100644 index 0000000..cc4d50b --- /dev/null +++ b/tensorflow/SQuAD/squad/utils.py @@ -0,0 +1,146 @@ +import re +import numpy as np + + +def get_2d_spans(text, tokenss): + spanss = [] + cur_idx = 0 + for tokens in tokenss: + spans = [] + for token in tokens: + if text.find(token, cur_idx) < 0: + print(tokens) + print("{} {} {}".format(token, cur_idx, text)) + raise Exception() + cur_idx = text.find(token, cur_idx) + spans.append((cur_idx, cur_idx + len(token))) + cur_idx += len(token) + spanss.append(spans) + return spanss + + +def get_word_span(context, wordss, start, stop): + spanss = get_2d_spans(context, wordss) + idxs = [] + for sent_idx, spans in enumerate(spanss): + for word_idx, span in enumerate(spans): + if not (stop <= span[0] or start >= span[1]): + idxs.append((sent_idx, word_idx)) + + assert len(idxs) > 0, "{} {} {} {}".format(context, spanss, start, stop) + return idxs[0], (idxs[-1][0], idxs[-1][1] + 1) + + +def get_phrase(context, wordss, span): + """ + Obtain phrase as substring of context given start and stop indices in word level + :param context: + :param wordss: + :param start: [sent_idx, word_idx] + :param stop: [sent_idx, word_idx] + :return: + """ + start, stop = span + flat_start = get_flat_idx(wordss, start) + flat_stop = get_flat_idx(wordss, stop) + words = sum(wordss, []) + char_idx = 0 + char_start, char_stop = None, None + for word_idx, word in enumerate(words): + char_idx = context.find(word, char_idx) + assert char_idx >= 0 + if word_idx == flat_start: + char_start = char_idx + char_idx += len(word) + if word_idx == flat_stop - 1: + char_stop = char_idx + assert char_start is not None + assert char_stop is not None + return context[char_start:char_stop] + + +def get_flat_idx(wordss, idx): + return sum(len(words) for words in wordss[:idx[0]]) + idx[1] + + +def get_word_idx(context, wordss, idx): + spanss = get_2d_spans(context, wordss) + return spanss[idx[0]][idx[1]][0] + + +def process_tokens(temp_tokens): + tokens = [] + for token in temp_tokens: + flag = False + l = ("-", "\u2212", "\u2014", "\u2013", "/", "~", '"', "'", "\u201C", "\u2019", "\u201D", "\u2018", "\u00B0") + # \u2013 is en-dash. Used for number to nubmer + # l = ("-", "\u2212", "\u2014", "\u2013") + # l = ("\u2013",) + tokens.extend(re.split("([{}])".format("".join(l)), token)) + return tokens + + +def get_best_span(ypi, yp2i): + max_val = 0 + best_word_span = (0, 1) + best_sent_idx = 0 + for f, (ypif, yp2if) in enumerate(zip(ypi, yp2i)): + argmax_j1 = 0 + for j in range(len(ypif)): + val1 = ypif[argmax_j1] + if val1 < ypif[j]: + val1 = ypif[j] + argmax_j1 = j + + val2 = yp2if[j] + if val1 * val2 > max_val: + best_word_span = (argmax_j1, j) + best_sent_idx = f + max_val = val1 * val2 + return ((best_sent_idx, best_word_span[0]), (best_sent_idx, best_word_span[1] + 1)), float(max_val) + + +def get_best_span_wy(wypi, th): + chunk_spans = [] + scores = [] + chunk_start = None + score = 0 + l = 0 + th = min(th, np.max(wypi)) + for f, wypif in enumerate(wypi): + for j, wypifj in enumerate(wypif): + if wypifj >= th: + if chunk_start is None: + chunk_start = f, j + score += wypifj + l += 1 + else: + if chunk_start is not None: + chunk_stop = f, j + chunk_spans.append((chunk_start, chunk_stop)) + scores.append(score/l) + score = 0 + l = 0 + chunk_start = None + if chunk_start is not None: + chunk_stop = f, j+1 + chunk_spans.append((chunk_start, chunk_stop)) + scores.append(score/l) + score = 0 + l = 0 + chunk_start = None + + return max(zip(chunk_spans, scores), key=lambda pair: pair[1]) + + +def get_span_score_pairs(ypi, yp2i): + span_score_pairs = [] + for f, (ypif, yp2if) in enumerate(zip(ypi, yp2i)): + for j in range(len(ypif)): + for k in range(j, len(yp2if)): + span = ((f, j), (f, k+1)) + score = ypif[j] * yp2if[k] + span_score_pairs.append((span, score)) + return span_score_pairs + + diff --git a/tensorflow/SQuAD/tree/__init__.py b/tensorflow/SQuAD/tree/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/tensorflow/SQuAD/tree/cli.py b/tensorflow/SQuAD/tree/cli.py new file mode 100644 index 0000000..01b6240 --- /dev/null +++ b/tensorflow/SQuAD/tree/cli.py @@ -0,0 +1,57 @@ +import os +from pprint import pprint + +import tensorflow as tf + +from tree.main import main as m + +flags = tf.app.flags + +flags.DEFINE_string("model_name", "tree", "Model name [tree]") +flags.DEFINE_string("data_dir", "data/squad", "Data dir [data/squad]") +flags.DEFINE_integer("run_id", 0, "Run ID [0]") + +flags.DEFINE_integer("batch_size", 128, "Batch size [128]") +flags.DEFINE_float("init_lr", 0.5, "Initial learning rate [0.5]") +flags.DEFINE_integer("num_epochs", 50, "Total number of epochs for training [50]") +flags.DEFINE_integer("num_steps", 0, "Number of steps [0]") +flags.DEFINE_integer("eval_num_batches", 100, "eval num batches [100]") +flags.DEFINE_integer("load_step", 0, "load step [0]") +flags.DEFINE_integer("early_stop", 4, "early stop [4]") + +flags.DEFINE_string("mode", "test", "train | test | forward [test]") +flags.DEFINE_boolean("load", True, "load saved data? [True]") +flags.DEFINE_boolean("progress", True, "Show progress? [True]") +flags.DEFINE_integer("log_period", 100, "Log period [100]") +flags.DEFINE_integer("eval_period", 1000, "Eval period [1000]") +flags.DEFINE_integer("save_period", 1000, "Save Period [1000]") +flags.DEFINE_float("decay", 0.9, "Exponential moving average decay [0.9]") + +flags.DEFINE_boolean("draft", False, "Draft for quick testing? [False]") + +flags.DEFINE_integer("hidden_size", 32, "Hidden size [32]") +flags.DEFINE_float("input_keep_prob", 0.5, "Input keep prob [0.5]") +flags.DEFINE_integer("char_emb_size", 8, "Char emb size [8]") +flags.DEFINE_integer("char_filter_height", 5, "Char filter height [5]") +flags.DEFINE_float("wd", 0.0001, "Weight decay [0.001]") +flags.DEFINE_bool("lower_word", True, "lower word [True]") +flags.DEFINE_bool("dump_eval", True, "dump eval? [True]") + +flags.DEFINE_integer("word_count_th", 100, "word count th [100]") +flags.DEFINE_integer("char_count_th", 500, "char count th [500]") +flags.DEFINE_integer("sent_size_th", 64, "sent size th [64]") +flags.DEFINE_integer("num_sents_th", 8, "num sents th [8]") +flags.DEFINE_integer("ques_size_th", 64, "ques size th [64]") +flags.DEFINE_integer("word_size_th", 16, "word size th [16]") +flags.DEFINE_integer("tree_height_th", 16, "tree height th [16]") + + +def main(_): + config = flags.FLAGS + + config.out_dir = os.path.join("out", config.model_name, str(config.run_id).zfill(2)) + + m(config) + +if __name__ == "__main__": + tf.app.run() diff --git a/tensorflow/SQuAD/tree/evaluator.py b/tensorflow/SQuAD/tree/evaluator.py new file mode 100644 index 0000000..e763905 --- /dev/null +++ b/tensorflow/SQuAD/tree/evaluator.py @@ -0,0 +1,197 @@ +import numpy as np +import tensorflow as tf + +from tree.read_data import DataSet +from my.nltk_utils import span_f1 + + +class Evaluation(object): + def __init__(self, data_type, global_step, idxs, yp): + self.data_type = data_type + self.global_step = global_step + self.idxs = idxs + self.yp = yp + self.num_examples = len(yp) + self.dict = {'data_type': data_type, + 'global_step': global_step, + 'yp': yp, + 'idxs': idxs, + 'num_examples': self.num_examples} + self.summaries = None + + def __repr__(self): + return "{} step {}".format(self.data_type, self.global_step) + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_yp = self.yp + other.yp + new_idxs = self.idxs + other.idxs + return Evaluation(self.data_type, self.global_step, new_idxs, new_yp) + + def __radd__(self, other): + return self.__add__(other) + + +class LabeledEvaluation(Evaluation): + def __init__(self, data_type, global_step, idxs, yp, y): + super(LabeledEvaluation, self).__init__(data_type, global_step, idxs, yp) + self.y = y + self.dict['y'] = y + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_yp = self.yp + other.yp + new_y = self.y + other.y + new_idxs = self.idxs + other.idxs + return LabeledEvaluation(self.data_type, self.global_step, new_idxs, new_yp, new_y) + + +class AccuracyEvaluation(LabeledEvaluation): + def __init__(self, data_type, global_step, idxs, yp, y, correct, loss): + super(AccuracyEvaluation, self).__init__(data_type, global_step, idxs, yp, y) + self.loss = loss + self.correct = correct + self.acc = sum(correct) / len(correct) + self.dict['loss'] = loss + self.dict['correct'] = correct + self.dict['acc'] = self.acc + loss_summary = tf.Summary(value=[tf.Summary.Value(tag='dev/loss', simple_value=self.loss)]) + acc_summary = tf.Summary(value=[tf.Summary.Value(tag='dev/acc', simple_value=self.acc)]) + self.summaries = [loss_summary, acc_summary] + + def __repr__(self): + return "{} step {}: accuracy={}, loss={}".format(self.data_type, self.global_step, self.acc, self.loss) + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_idxs = self.idxs + other.idxs + new_yp = self.yp + other.yp + new_y = self.y + other.y + new_correct = self.correct + other.correct + new_loss = (self.loss * self.num_examples + other.loss * other.num_examples) / len(new_correct) + return AccuracyEvaluation(self.data_type, self.global_step, new_idxs, new_yp, new_y, new_correct, new_loss) + + +class Evaluator(object): + def __init__(self, config, model): + self.config = config + self.model = model + + def get_evaluation(self, sess, batch): + idxs, data_set = batch + feed_dict = self.model.get_feed_dict(data_set, False, supervised=False) + global_step, yp = sess.run([self.model.global_step, self.model.yp], feed_dict=feed_dict) + yp = yp[:data_set.num_examples] + e = Evaluation(data_set.data_type, int(global_step), idxs, yp.tolist()) + return e + + def get_evaluation_from_batches(self, sess, batches): + e = sum(self.get_evaluation(sess, batch) for batch in batches) + return e + + +class LabeledEvaluator(Evaluator): + def get_evaluation(self, sess, batch): + idxs, data_set = batch + feed_dict = self.model.get_feed_dict(data_set, False, supervised=False) + global_step, yp = sess.run([self.model.global_step, self.model.yp], feed_dict=feed_dict) + yp = yp[:data_set.num_examples] + y = feed_dict[self.model.y] + e = LabeledEvaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), y.tolist()) + return e + + +class AccuracyEvaluator(LabeledEvaluator): + def get_evaluation(self, sess, batch): + idxs, data_set = batch + assert isinstance(data_set, DataSet) + feed_dict = self.model.get_feed_dict(data_set, False) + global_step, yp, loss = sess.run([self.model.global_step, self.model.yp, self.model.loss], feed_dict=feed_dict) + y = feed_dict[self.model.y] + yp = yp[:data_set.num_examples] + correct = [self.__class__.compare(yi, ypi) for yi, ypi in zip(y, yp)] + e = AccuracyEvaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), y.tolist(), correct, float(loss)) + return e + + @staticmethod + def compare(yi, ypi): + return int(np.argmax(yi)) == int(np.argmax(ypi)) + + +class AccuracyEvaluator2(AccuracyEvaluator): + @staticmethod + def compare(yi, ypi): + i = int(np.argmax(yi.flatten())) + j = int(np.argmax(ypi.flatten())) + # print(i, j, i == j) + return i == j + + +class TempEvaluation(AccuracyEvaluation): + def __init__(self, data_type, global_step, idxs, yp, yp2, y, y2, correct, loss, f1s): + super(TempEvaluation, self).__init__(data_type, global_step, idxs, yp, y, correct, loss) + self.y2 = y2 + self.yp2 = yp2 + self.f1s = f1s + self.f1 = float(np.mean(f1s)) + self.dict['y2'] = y2 + self.dict['yp2'] = yp2 + self.dict['f1s'] = f1s + self.dict['f1'] = self.f1 + f1_summary = tf.Summary(value=[tf.Summary.Value(tag='dev/f1', simple_value=self.f1)]) + self.summaries.append(f1_summary) + + def __add__(self, other): + if other == 0: + return self + assert self.data_type == other.data_type + assert self.global_step == other.global_step + new_idxs = self.idxs + other.idxs + new_yp = self.yp + other.yp + new_yp2 = self.yp2 + other.yp2 + new_y = self.y + other.y + new_y2 = self.y2 + other.y2 + new_correct = self.correct + other.correct + new_f1s = self.f1s + other.f1s + new_loss = (self.loss * self.num_examples + other.loss * other.num_examples) / len(new_correct) + return TempEvaluation(self.data_type, self.global_step, new_idxs, new_yp, new_yp2, new_y, new_y2, new_correct, new_loss, new_f1s) + + +class TempEvaluator(LabeledEvaluator): + def get_evaluation(self, sess, batch): + idxs, data_set = batch + assert isinstance(data_set, DataSet) + feed_dict = self.model.get_feed_dict(data_set, False) + global_step, yp, yp2, loss = sess.run([self.model.global_step, self.model.yp, self.model.yp2, self.model.loss], feed_dict=feed_dict) + y, y2 = feed_dict[self.model.y], feed_dict[self.model.y2] + yp, yp2 = yp[:data_set.num_examples], yp2[:data_set.num_examples] + correct = [self.__class__.compare(yi, y2i, ypi, yp2i) for yi, y2i, ypi, yp2i in zip(y, y2, yp, yp2)] + f1s = [self.__class__.span_f1(yi, y2i, ypi, yp2i) for yi, y2i, ypi, yp2i in zip(y, y2, yp, yp2)] + e = TempEvaluation(data_set.data_type, int(global_step), idxs, yp.tolist(), yp2.tolist(), y.tolist(), y2.tolist(), correct, float(loss), f1s) + return e + + @staticmethod + def compare(yi, y2i, ypi, yp2i): + i = int(np.argmax(yi.flatten())) + j = int(np.argmax(ypi.flatten())) + k = int(np.argmax(y2i.flatten())) + l = int(np.argmax(yp2i.flatten())) + # print(i, j, i == j) + return i == j and k == l + + @staticmethod + def span_f1(yi, y2i, ypi, yp2i): + true_span = (np.argmax(yi.flatten()), np.argmax(y2i.flatten())+1) + pred_span = (np.argmax(ypi.flatten()), np.argmax(yp2i.flatten())+1) + f1 = span_f1(true_span, pred_span) + return f1 + diff --git a/tensorflow/SQuAD/tree/graph_handler.py b/tensorflow/SQuAD/tree/graph_handler.py new file mode 100644 index 0000000..8115249 --- /dev/null +++ b/tensorflow/SQuAD/tree/graph_handler.py @@ -0,0 +1,54 @@ +import json +from json import encoder +import os + +import tensorflow as tf + +from tree.evaluator import Evaluation +from my.utils import short_floats + + +class GraphHandler(object): + def __init__(self, config): + self.config = config + self.saver = tf.train.Saver() + self.writer = None + self.save_path = os.path.join(config.save_dir, config.model_name) + + def initialize(self, sess): + if self.config.load: + self._load(sess) + else: + sess.run(tf.global_variables_initializer()) + + if self.config.mode == 'train': + self.writer = tf.summary.FileWriter(self.config.log_dir, graph=tf.get_default_graph()) + + def save(self, sess, global_step=None): + self.saver.save(sess, self.save_path, global_step=global_step) + + def _load(self, sess): + config = self.config + if config.load_step > 0: + save_path = os.path.join(config.save_dir, "{}-{}".format(config.model_name, config.load_step)) + else: + save_dir = config.save_dir + checkpoint = tf.train.get_checkpoint_state(save_dir) + assert checkpoint is not None, "cannot load checkpoint at {}".format(save_dir) + save_path = checkpoint.model_checkpoint_path + print("Loading saved model from {}".format(save_path)) + self.saver.restore(sess, save_path) + + def add_summary(self, summary, global_step): + self.writer.add_summary(summary, global_step) + + def add_summaries(self, summaries, global_step): + for summary in summaries: + self.add_summary(summary, global_step) + + def dump_eval(self, e, precision=2): + assert isinstance(e, Evaluation) + path = os.path.join(self.config.eval_dir, "{}-{}.json".format(e.data_type, str(e.global_step).zfill(6))) + with open(path, 'w') as fh: + json.dump(short_floats(e.dict, precision), fh) + diff --git a/tensorflow/SQuAD/tree/main.py b/tensorflow/SQuAD/tree/main.py new file mode 100644 index 0000000..393a563 --- /dev/null +++ b/tensorflow/SQuAD/tree/main.py @@ -0,0 +1,187 @@ +import argparse +import json +import math +import os +import shutil +from pprint import pprint + +import tensorflow as tf +from tqdm import tqdm +import numpy as np + +from tree.evaluator import AccuracyEvaluator2, Evaluator +from tree.graph_handler import GraphHandler +from tree.model import Model +from tree.trainer import Trainer + +from tree.read_data import load_metadata, read_data, get_squad_data_filter, update_config + + +def main(config): + set_dirs(config) + if config.mode == 'train': + _train(config) + elif config.mode == 'test': + _test(config) + elif config.mode == 'forward': + _forward(config) + else: + raise ValueError("invalid value for 'mode': {}".format(config.mode)) + + +def _config_draft(config): + if config.draft: + config.num_steps = 10 + config.eval_period = 10 + config.log_period = 1 + config.save_period = 10 + config.eval_num_batches = 1 + + +def _train(config): + # load_metadata(config, 'train') # this updates the config file according to metadata file + + data_filter = get_squad_data_filter(config) + train_data = read_data(config, 'train', config.load, data_filter=data_filter) + dev_data = read_data(config, 'dev', True, data_filter=data_filter) + update_config(config, [train_data, dev_data]) + + _config_draft(config) + + word2vec_dict = train_data.shared['lower_word2vec'] if config.lower_word else train_data.shared['word2vec'] + word2idx_dict = train_data.shared['word2idx'] + idx2vec_dict = {word2idx_dict[word]: vec for word, vec in word2vec_dict.items() if word in word2idx_dict} + print("{}/{} unique words have corresponding glove vectors.".format(len(idx2vec_dict), len(word2idx_dict))) + emb_mat = np.array([idx2vec_dict[idx] if idx in idx2vec_dict + else np.random.multivariate_normal(np.zeros(config.word_emb_size), np.eye(config.word_emb_size)) + for idx in range(config.word_vocab_size)]) + config.emb_mat = emb_mat + + # construct model graph and variables (using default graph) + pprint(config.__flags, indent=2) + model = Model(config) + trainer = Trainer(config, model) + evaluator = AccuracyEvaluator2(config, model) + graph_handler = GraphHandler(config) # controls all tensors and variables in the graph, including loading /saving + + # Variables + sess = tf.Session() + graph_handler.initialize(sess) + + # begin training + num_steps = config.num_steps or int(config.num_epochs * train_data.num_examples / config.batch_size) + max_acc = 0 + noupdate_count = 0 + global_step = 0 + for _, batch in tqdm(train_data.get_batches(config.batch_size, num_batches=num_steps, shuffle=True), total=num_steps): + global_step = sess.run(model.global_step) + 1 # +1 because all calculations are done after step + get_summary = global_step % config.log_period == 0 + loss, summary, train_op = trainer.step(sess, batch, get_summary=get_summary) + if get_summary: + graph_handler.add_summary(summary, global_step) + + # Occasional evaluation and saving + if global_step % config.save_period == 0: + graph_handler.save(sess, global_step=global_step) + if global_step % config.eval_period == 0: + num_batches = math.ceil(dev_data.num_examples / config.batch_size) + if 0 < config.eval_num_batches < num_batches: + num_batches = config.eval_num_batches + e = evaluator.get_evaluation_from_batches( + sess, tqdm(dev_data.get_batches(config.batch_size, num_batches=num_batches), total=num_batches)) + graph_handler.add_summaries(e.summaries, global_step) + if e.acc > max_acc: + max_acc = e.acc + noupdate_count = 0 + else: + noupdate_count += 1 + if noupdate_count == config.early_stop: + break + if config.dump_eval: + graph_handler.dump_eval(e) + if global_step % config.save_period != 0: + graph_handler.save(sess, global_step=global_step) + + +def _test(config): + test_data = read_data(config, 'test', True) + update_config(config, [test_data]) + + _config_draft(config) + + pprint(config.__flags, indent=2) + model = Model(config) + evaluator = AccuracyEvaluator2(config, model) + graph_handler = GraphHandler(config) # controls all tensors and variables in the graph, including loading /saving + + sess = tf.Session() + graph_handler.initialize(sess) + + num_batches = math.ceil(test_data.num_examples / config.batch_size) + if 0 < config.eval_num_batches < num_batches: + num_batches = config.eval_num_batches + e = evaluator.get_evaluation_from_batches(sess, tqdm(test_data.get_batches(config.batch_size, num_batches=num_batches), total=num_batches)) + print(e) + if config.dump_eval: + graph_handler.dump_eval(e) + + +def _forward(config): + + forward_data = read_data(config, 'forward', True) + + _config_draft(config) + + pprint(config.__flag, indent=2) + model = Model(config) + evaluator = Evaluator(config, model) + graph_handler = GraphHandler(config) # controls all tensors and variables in the graph, including loading /saving + + sess = tf.Session() + graph_handler.initialize(sess) + + num_batches = math.ceil(forward_data.num_examples / config.batch_size) + if 0 < config.eval_num_batches < num_batches: + num_batches = config.eval_num_batches + e = evaluator.get_evaluation_from_batches(sess, tqdm(forward_data.get_batches(config.batch_size, num_batches=num_batches), total=num_batches)) + print(e) + if config.dump_eval: + graph_handler.dump_eval(e) + + +def set_dirs(config): + # create directories + if not config.load and os.path.exists(config.out_dir): + shutil.rmtree(config.out_dir) + + config.save_dir = os.path.join(config.out_dir, "save") + config.log_dir = os.path.join(config.out_dir, "log") + config.eval_dir = os.path.join(config.out_dir, "eval") + if not os.path.exists(config.out_dir): + os.makedirs(config.out_dir) + if not os.path.exists(config.save_dir): + os.mkdir(config.save_dir) + if not os.path.exists(config.log_dir): + os.mkdir(config.eval_dir) + + +def _get_args(): + parser = argparse.ArgumentParser() + parser.add_argument("config_path") + return parser.parse_args() + + +class Config(object): + def __init__(self, **entries): + self.__dict__.update(entries) + + +def _run(): + args = _get_args() + with open(args.config_path, 'r') as fh: + config = Config(**json.load(fh)) + main(config) + + +if __name__ == "__main__": + _run() diff --git a/tensorflow/SQuAD/tree/model.py b/tensorflow/SQuAD/tree/model.py new file mode 100644 index 0000000..2d63ef4 --- /dev/null +++ b/tensorflow/SQuAD/tree/model.py @@ -0,0 +1,248 @@ +import nltk +import numpy as np +import tensorflow as tf +from tensorflow.python.ops.rnn_cell import BasicLSTMCell + +from my.nltk_utils import tree2matrix, find_max_f1_subtree, load_compressed_tree, set_span +from tree.read_data import DataSet +from my.tensorflow import exp_mask, get_initializer +from my.tensorflow.nn import linear +from my.tensorflow.rnn import bidirectional_dynamic_rnn, dynamic_rnn +from my.tensorflow.rnn_cell import SwitchableDropoutWrapper, NoOpCell, TreeRNNCell + + +class Model(object): + def __init__(self, config): + self.config = config + self.global_step = tf.get_variable('global_step', shape=[], dtype='int32', + initializer=tf.constant_initializer(0), trainable=False) + + # Define forward inputs here + N, M, JX, JQ, VW, VC, W, H = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size, config.max_word_size, config.max_tree_height + self.x = tf.placeholder('int32', [None, M, JX], name='x') + self.cx = tf.placeholder('int32', [None, M, JX, W], name='cx') + self.q = tf.placeholder('int32', [None, JQ], name='q') + self.cq = tf.placeholder('int32', [None, JQ, W], name='cq') + self.tx = tf.placeholder('int32', [None, M, H, JX], name='tx') + self.tx_edge_mask = tf.placeholder('bool', [None, M, H, JX, JX], name='tx_edge_mask') + self.y = tf.placeholder('bool', [None, M, H, JX], name='y') + self.is_train = tf.placeholder('bool', [], name='is_train') + + # Define misc + + # Forward outputs / loss inputs + self.logits = None + self.yp = None + self.var_list = None + + # Loss outputs + self.loss = None + + self._build_forward() + self._build_loss() + + self.ema_op = self._get_ema_op() + self.summary = tf.summary.merge_all() + + def _build_forward(self): + config = self.config + N, M, JX, JQ, VW, VC, d, dc, W = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size, config.hidden_size, \ + config.char_emb_size, config.max_word_size + H = config.max_tree_height + + x_mask = self.x > 0 + q_mask = self.q > 0 + tx_mask = self.tx > 0 # [N, M, H, JX] + + with tf.variable_scope("char_emb"): + char_emb_mat = tf.get_variable("char_emb_mat", shape=[VC, dc], dtype='float') + Acx = tf.nn.embedding_lookup(char_emb_mat, self.cx) # [N, M, JX, W, dc] + Acq = tf.nn.embedding_lookup(char_emb_mat, self.cq) # [N, JQ, W, dc] + + filter = tf.get_variable("filter", shape=[1, config.char_filter_height, dc, d], dtype='float') + bias = tf.get_variable("bias", shape=[d], dtype='float') + strides = [1, 1, 1, 1] + Acx = tf.reshape(Acx, [-1, JX, W, dc]) + Acq = tf.reshape(Acq, [-1, JQ, W, dc]) + xxc = tf.nn.conv2d(Acx, filter, strides, "VALID") + bias # [N*M, JX, W/filter_stride, d] + qqc = tf.nn.conv2d(Acq, filter, strides, "VALID") + bias # [N, JQ, W/filter_stride, d] + xxc = tf.reshape(tf.reduce_max(tf.nn.relu(xxc), 2), [-1, M, JX, d]) + qqc = tf.reshape(tf.reduce_max(tf.nn.relu(qqc), 2), [-1, JQ, d]) + + with tf.variable_scope("word_emb"): + if config.mode == 'train': + word_emb_mat = tf.get_variable("word_emb_mat", dtype='float', shape=[VW, config.word_emb_size], initializer=get_initializer(config.emb_mat)) + else: + word_emb_mat = tf.get_variable("word_emb_mat", shape=[VW, config.word_emb_size], dtype='float') + Ax = tf.nn.embedding_lookup(word_emb_mat, self.x) # [N, M, JX, d] + Aq = tf.nn.embedding_lookup(word_emb_mat, self.q) # [N, JQ, d] + # Ax = linear([Ax], d, False, scope='Ax_reshape') + # Aq = linear([Aq], d, False, scope='Aq_reshape') + + xx = tf.concat(axis=3, values=[xxc, Ax]) # [N, M, JX, 2d] + qq = tf.concat(axis=2, values=[qqc, Aq]) # [N, JQ, 2d] + D = d + config.word_emb_size + + with tf.variable_scope("pos_emb"): + pos_emb_mat = tf.get_variable("pos_emb_mat", shape=[config.pos_vocab_size, d], dtype='float') + Atx = tf.nn.embedding_lookup(pos_emb_mat, self.tx) # [N, M, H, JX, d] + + cell = BasicLSTMCell(D, state_is_tuple=True) + cell = SwitchableDropoutWrapper(cell, self.is_train, input_keep_prob=config.input_keep_prob) + x_len = tf.reduce_sum(tf.cast(x_mask, 'int32'), 2) # [N, M] + q_len = tf.reduce_sum(tf.cast(q_mask, 'int32'), 1) # [N] + + with tf.variable_scope("rnn"): + (fw_h, bw_h), _ = bidirectional_dynamic_rnn(cell, cell, xx, x_len, dtype='float', scope='start') # [N, M, JX, 2d] + tf.get_variable_scope().reuse_variables() + (fw_us, bw_us), (_, (fw_u, bw_u)) = bidirectional_dynamic_rnn(cell, cell, qq, q_len, dtype='float', scope='start') # [N, J, d], [N, d] + u = (fw_u + bw_u) / 2.0 + h = (fw_h + bw_h) / 2.0 + + with tf.variable_scope("h"): + no_op_cell = NoOpCell(D) + tree_rnn_cell = TreeRNNCell(no_op_cell, d, tf.reduce_max) + initial_state = tf.reshape(h, [N*M*JX, D]) # [N*M*JX, D] + inputs = tf.concat(axis=4, values=[Atx, tf.cast(self.tx_edge_mask, 'float')]) # [N, M, H, JX, d+JX] + inputs = tf.reshape(tf.transpose(inputs, [0, 1, 3, 2, 4]), [N*M*JX, H, d + JX]) # [N*M*JX, H, d+JX] + length = tf.reshape(tf.reduce_sum(tf.cast(tx_mask, 'int32'), 2), [N*M*JX]) + # length = tf.reshape(tf.reduce_sum(tf.cast(tf.transpose(tx_mask, [0, 1, 3, 2]), 'float'), 3), [-1]) + h, _ = dynamic_rnn(tree_rnn_cell, inputs, length, initial_state=initial_state) # [N*M*JX, H, D] + h = tf.transpose(tf.reshape(h, [N, M, JX, H, D]), [0, 1, 3, 2, 4]) # [N, M, H, JX, D] + + u = tf.expand_dims(tf.expand_dims(tf.expand_dims(u, 1), 1), 1) # [N, 1, 1, 1, 4d] + dot = linear(h * u, 1, True, squeeze=True, scope='dot') # [N, M, H, JX] + # self.logits = tf.reshape(dot, [N, M * H * JX]) + self.logits = tf.reshape(exp_mask(dot, tx_mask), [N, M * H * JX]) # [N, M, H, JX] + self.yp = tf.reshape(tf.nn.softmax(self.logits), [N, M, H, JX]) + + def _build_loss(self): + config = self.config + N, M, JX, JQ, VW, VC = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size + H = config.max_tree_height + ce_loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits( + logits=self.logits, labels=tf.cast(tf.reshape(self.y, [N, M * H * JX]), 'float'))) + tf.add_to_collection('losses', ce_loss) + self.loss = tf.add_n(tf.get_collection('losses'), name='loss') + tf.summary.scalar(self.loss.op.name, self.loss) + tf.add_to_collection('ema/scalar', self.loss) + + def _get_ema_op(self): + ema = tf.train.ExponentialMovingAverage(self.config.decay) + ema_op = ema.apply(tf.get_collection("ema/scalar") + tf.get_collection("ema/histogram")) + for var in tf.get_collection("ema/scalar"): + ema_var = ema.average(var) + tf.summary.scalar(ema_var.op.name, ema_var) + for var in tf.get_collection("ema/histogram"): + ema_var = ema.average(var) + tf.summary.histogram(ema_var.op.name, ema_var) + return ema_op + + def get_loss(self): + return self.loss + + def get_global_step(self): + return self.global_step + + def get_var_list(self): + return self.var_list + + def get_feed_dict(self, batch, is_train, supervised=True): + assert isinstance(batch, DataSet) + config = self.config + N, M, JX, JQ, VW, VC, d, W, H = \ + config.batch_size, config.max_num_sents, config.max_sent_size, \ + config.max_ques_size, config.word_vocab_size, config.char_vocab_size, config.hidden_size, config.max_word_size, \ + config.max_tree_height + feed_dict = {} + + x = np.zeros([N, M, JX], dtype='int32') + cx = np.zeros([N, M, JX, W], dtype='int32') + q = np.zeros([N, JQ], dtype='int32') + cq = np.zeros([N, JQ, W], dtype='int32') + tx = np.zeros([N, M, H, JX], dtype='int32') + tx_edge_mask = np.zeros([N, M, H, JX, JX], dtype='bool') + + feed_dict[self.x] = x + feed_dict[self.cx] = cx + feed_dict[self.q] = q + feed_dict[self.cq] = cq + feed_dict[self.tx] = tx + feed_dict[self.tx_edge_mask] = tx_edge_mask + feed_dict[self.is_train] = is_train + + def _get_word(word): + d = batch.shared['word2idx'] + for each in (word, word.lower(), word.capitalize(), word.upper()): + if each in d: + return d[each] + return 1 + + def _get_char(char): + d = batch.shared['char2idx'] + if char in d: + return d[char] + return 1 + + def _get_pos(tree): + d = batch.shared['pos2idx'] + if tree.label() in d: + return d[tree.label()] + return 1 + + for i, xi in enumerate(batch.data['x']): + for j, xij in enumerate(xi): + for k, xijk in enumerate(xij): + x[i, j, k] = _get_word(xijk) + + for i, cxi in enumerate(batch.data['cx']): + for j, cxij in enumerate(cxi): + for k, cxijk in enumerate(cxij): + for l, cxijkl in enumerate(cxijk): + cx[i, j, k, l] = _get_char(cxijkl) + if l + 1 == config.max_word_size: + break + + for i, qi in enumerate(batch.data['q']): + for j, qij in enumerate(qi): + q[i, j] = _get_word(qij) + + for i, cqi in enumerate(batch.data['cq']): + for j, cqij in enumerate(cqi): + for k, cqijk in enumerate(cqij): + cq[i, j, k] = _get_char(cqijk) + if k + 1 == config.max_word_size: + break + + for i, txi in enumerate(batch.data['stx']): + for j, txij in enumerate(txi): + txij_mat, txij_mask = tree2matrix(nltk.tree.Tree.fromstring(txij), _get_pos, row_size=H, col_size=JX) + tx[i, j, :, :], tx_edge_mask[i, j, :, :, :] = txij_mat, txij_mask + + if supervised: + y = np.zeros([N, M, H, JX], dtype='bool') + feed_dict[self.y] = y + for i, yi in enumerate(batch.data['y']): + start_idx, stop_idx = yi + sent_idx = start_idx[0] + if start_idx[0] == stop_idx[0]: + span = [start_idx[1], stop_idx[1]] + else: + span = [start_idx[1], len(batch.data['x'][sent_idx])] + tree = nltk.tree.Tree.fromstring(batch.data['stx'][i][sent_idx]) + set_span(tree) + best_subtree = find_max_f1_subtree(tree, span) + + def _get_y(t): + return t == best_subtree + + yij, _ = tree2matrix(tree, _get_y, H, JX, dtype='bool') + y[i, sent_idx, :, :] = yij + + return feed_dict diff --git a/tensorflow/SQuAD/tree/read_data.py b/tensorflow/SQuAD/tree/read_data.py new file mode 100644 index 0000000..55295fc --- /dev/null +++ b/tensorflow/SQuAD/tree/read_data.py @@ -0,0 +1,159 @@ +import json +import os +import random +import itertools +import math + +import nltk + +from my.nltk_utils import load_compressed_tree +from my.utils import index + + +class DataSet(object): + def __init__(self, data, data_type, shared=None, valid_idxs=None): + total_num_examples = len(next(iter(data.values()))) + self.data = data # e.g. {'X': [0, 1, 2], 'Y': [2, 3, 4]} + self.data_type = data_type + self.shared = shared + self.valid_idxs = range(total_num_examples) if valid_idxs is None else valid_idxs + self.num_examples = len(self.valid_idxs) + + def get_batches(self, batch_size, num_batches=None, shuffle=False): + num_batches_per_epoch = int(math.ceil(self.num_examples / batch_size)) + if num_batches is None: + num_batches = num_batches_per_epoch + num_epochs = int(math.ceil(num_batches / num_batches_per_epoch)) + + idxs = itertools.chain.from_iterable(random.sample(self.valid_idxs, len(self.valid_idxs)) + if shuffle else self.valid_idxs + for _ in range(num_epochs)) + for _ in range(num_batches): + batch_idxs = tuple(itertools.islice(idxs, batch_size)) + batch_data = {} + for key, val in self.data.items(): + if key.startswith('*'): + assert self.shared is not None + shared_key = key[1:] + batch_data[shared_key] = [index(self.shared[shared_key], val[idx]) for idx in batch_idxs] + else: + batch_data[key] = list(map(val.__getitem__, batch_idxs)) + + batch_ds = DataSet(batch_data, self.data_type, shared=self.shared) + yield batch_idxs, batch_ds + + +class SquadDataSet(DataSet): + def __init__(self, data, data_type, shared=None, valid_idxs=None): + super(SquadDataSet, self).__init__(data, data_type, shared=shared, valid_idxs=valid_idxs) + + +def load_metadata(config, data_type): + metadata_path = os.path.join(config.data_dir, "metadata_{}.json".format(data_type)) + with open(metadata_path, 'r') as fh: + metadata = json.load(fh) + for key, val in metadata.items(): + config.__setattr__(key, val) + return metadata + + +def read_data(config, data_type, ref, data_filter=None): + data_path = os.path.join(config.data_dir, "data_{}.json".format(data_type)) + shared_path = os.path.join(config.data_dir, "shared_{}.json".format(data_type)) + with open(data_path, 'r') as fh: + data = json.load(fh) + with open(shared_path, 'r') as fh: + shared = json.load(fh) + + num_examples = len(next(iter(data.values()))) + if data_filter is None: + valid_idxs = range(num_examples) + else: + mask = [] + keys = data.keys() + values = data.values() + for vals in zip(*values): + each = {key: val for key, val in zip(keys, vals)} + mask.append(data_filter(each, shared)) + valid_idxs = [idx for idx in range(len(mask)) if mask[idx]] + + print("Loaded {}/{} examples from {}".format(len(valid_idxs), num_examples, data_type)) + + shared_path = os.path.join(config.out_dir, "shared.json") + if not ref: + word_counter = shared['lower_word_counter'] if config.lower_word else shared['word_counter'] + char_counter = shared['char_counter'] + pos_counter = shared['pos_counter'] + shared['word2idx'] = {word: idx + 2 for idx, word in + enumerate(word for word, count in word_counter.items() + if count > config.word_count_th)} + shared['char2idx'] = {char: idx + 2 for idx, char in + enumerate(char for char, count in char_counter.items() + if count > config.char_count_th)} + shared['pos2idx'] = {pos: idx + 2 for idx, pos in enumerate(pos_counter.keys())} + NULL = "-NULL-" + UNK = "-UNK-" + shared['word2idx'][NULL] = 0 + shared['word2idx'][UNK] = 1 + shared['char2idx'][NULL] = 0 + shared['char2idx'][UNK] = 1 + shared['pos2idx'][NULL] = 0 + shared['pos2idx'][UNK] = 1 + json.dump({'word2idx': shared['word2idx'], 'char2idx': shared['char2idx'], + 'pos2idx': shared['pos2idx']}, open(shared_path, 'w')) + else: + new_shared = json.load(open(shared_path, 'r')) + for key, val in new_shared.items(): + shared[key] = val + + data_set = DataSet(data, data_type, shared=shared, valid_idxs=valid_idxs) + return data_set + + +def get_squad_data_filter(config): + def data_filter(data_point, shared): + assert shared is not None + rx, rcx, q, cq, y = (data_point[key] for key in ('*x', '*cx', 'q', 'cq', 'y')) + x, cx, stx = shared['x'], shared['cx'], shared['stx'] + if len(q) > config.ques_size_th: + return False + xi = x[rx[0]][rx[1]] + if len(xi) > config.num_sents_th: + return False + if any(len(xij) > config.sent_size_th for xij in xi): + return False + stxi = stx[rx[0]][rx[1]] + if any(nltk.tree.Tree.fromstring(s).height() > config.tree_height_th for s in stxi): + return False + return True + return data_filter + + +def update_config(config, data_sets): + config.max_num_sents = 0 + config.max_sent_size = 0 + config.max_ques_size = 0 + config.max_word_size = 0 + config.max_tree_height = 0 + for data_set in data_sets: + data = data_set.data + shared = data_set.shared + for idx in data_set.valid_idxs: + rx = data['*x'][idx] + q = data['q'][idx] + sents = shared['x'][rx[0]][rx[1]] + trees = map(nltk.tree.Tree.fromstring, shared['stx'][rx[0]][rx[1]]) + config.max_tree_height = max(config.max_tree_height, max(tree.height() for tree in trees)) + config.max_num_sents = max(config.max_num_sents, len(sents)) + config.max_sent_size = max(config.max_sent_size, max(map(len, sents))) + config.max_word_size = max(config.max_word_size, max(len(word) for sent in sents for word in sent)) + if len(q) > 0: + config.max_ques_size = max(config.max_ques_size, len(q)) + config.max_word_size = max(config.max_word_size, max(len(word) for word in q)) + + config.max_word_size = min(config.max_word_size, config.word_size_th) + + config.char_vocab_size = len(data_sets[0].shared['char2idx']) + config.word_emb_size = len(next(iter(data_sets[0].shared['word2vec'].values()))) + config.word_vocab_size = len(data_sets[0].shared['word2idx']) + config.pos_vocab_size = len(data_sets[0].shared['pos2idx']) diff --git a/tensorflow/SQuAD/tree/templates/visualizer.html b/tensorflow/SQuAD/tree/templates/visualizer.html new file mode 100644 index 0000000..5f8bb74 --- /dev/null +++ b/tensorflow/SQuAD/tree/templates/visualizer.html @@ -0,0 +1,67 @@ + + + + + {{ title }} + + + + + + +

    {{ title }}

    + + + + + + + + {% for row in rows %} + + + + + + + {% endfor %} +
    IDQuestionAnswerParagraph
    {{ row.id }} + {% for qj in row.ques %} + {{ qj }} + {% endfor %} + {{ row.a }} + + {% for xj, yj, y2j, ypj, yp2j in zip(row.para, row.y, row.y2, row.yp, row.yp2) %} + + {% for xjk, yjk, y2jk, ypjk in zip(xj, yj, y2j, ypj) %} + + {% endfor %} + + + {% for xjk, yp2jk in zip(xj, yp2j) %} + + {% endfor %} + + {% endfor %} +
    + {% if yjk or y2jk %} + {{ xjk }} + {% else %} + {{ xjk }} + {% endif %} +
    -
    +
    + + \ No newline at end of file diff --git a/tensorflow/SQuAD/tree/test.ipynb b/tensorflow/SQuAD/tree/test.ipynb new file mode 100644 index 0000000..0115baf --- /dev/null +++ b/tensorflow/SQuAD/tree/test.ipynb @@ -0,0 +1,294 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [ + "import nltk\n", + "import matplotlib.pyplot as plt\n", + "%matplotlib inline" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "(S (PRP I) (VP (VBP am) (NNP Sam)) (. .))\n", + "(PRP I)\n", + "(VP (VBP am) (NNP Sam))\n", + "(VBP am)\n", + "(NNP Sam)\n", + "(. .)\n", + "(S (PRP I) (VP (VBP am) (NNP Sam)) (. .))\n" + ] + } + ], + "source": [ + "string = \"(ROOT(S(NP (PRP I))(VP (VBP am)(NP (NNP Sam)))(. .)))\"\n", + "tree = nltk.tree.Tree.fromstring(string)\n", + "\n", + "def load_compressed_tree(s):\n", + "\n", + " def compress_tree(tree):\n", + " if len(tree) == 1:\n", + " if isinstance(tree[0], nltk.tree.Tree):\n", + " return compress_tree(tree[0])\n", + " else:\n", + " return tree\n", + " else:\n", + " for i, t in enumerate(tree):\n", + " tree[i] = compress_tree(t)\n", + " return tree\n", + "\n", + " return compress_tree(nltk.tree.Tree.fromstring(s))\n", + "tree = load_compressed_tree(string)\n", + "for t in tree.subtrees():\n", + " print(t)\n", + " \n", + "print(str(tree))" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "(ROOT I am Sam .)\n" + ] + } + ], + "source": [ + "print(tree.flatten())" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "['ROOT', 'S', 'NP', 'PRP', 'VP', 'VBP', 'NP', 'NNP', '.']\n" + ] + } + ], + "source": [ + "print(list(t.label() for t in tree.subtrees()))" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [ + "import json\n", + "d = json.load(open(\"data/squad/shared_dev.json\", 'r'))" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "text/plain": [ + "73" + ] + }, + "execution_count": 12, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "len(d['pos_counter'])" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "data": { + "text/plain": [ + "{'#': 6,\n", + " '$': 80,\n", + " \"''\": 1291,\n", + " ',': 14136,\n", + " '-LRB-': 1926,\n", + " '-RRB-': 1925,\n", + " '.': 9505,\n", + " ':': 1455,\n", + " 'ADJP': 3426,\n", + " 'ADVP': 4936,\n", + " 'CC': 9300,\n", + " 'CD': 6216,\n", + " 'CONJP': 191,\n", + " 'DT': 26286,\n", + " 'EX': 288,\n", + " 'FRAG': 107,\n", + " 'FW': 96,\n", + " 'IN': 32564,\n", + " 'INTJ': 12,\n", + " 'JJ': 21452,\n", + " 'JJR': 563,\n", + " 'JJS': 569,\n", + " 'LS': 7,\n", + " 'LST': 1,\n", + " 'MD': 1051,\n", + " 'NAC': 19,\n", + " 'NN': 34750,\n", + " 'NNP': 28392,\n", + " 'NNPS': 1400,\n", + " 'NNS': 16716,\n", + " 'NP': 91636,\n", + " 'NP-TMP': 236,\n", + " 'NX': 108,\n", + " 'PDT': 89,\n", + " 'POS': 1451,\n", + " 'PP': 33278,\n", + " 'PRN': 2085,\n", + " 'PRP': 2320,\n", + " 'PRP$': 1959,\n", + " 'PRT': 450,\n", + " 'QP': 838,\n", + " 'RB': 7611,\n", + " 'RBR': 301,\n", + " 'RBS': 252,\n", + " 'ROOT': 9587,\n", + " 'RP': 454,\n", + " 'RRC': 19,\n", + " 'S': 21557,\n", + " 'SBAR': 5009,\n", + " 'SBARQ': 6,\n", + " 'SINV': 135,\n", + " 'SQ': 5,\n", + " 'SYM': 17,\n", + " 'TO': 5167,\n", + " 'UCP': 143,\n", + " 'UH': 15,\n", + " 'VB': 4197,\n", + " 'VBD': 8377,\n", + " 'VBG': 3570,\n", + " 'VBN': 7218,\n", + " 'VBP': 2897,\n", + " 'VBZ': 4146,\n", + " 'VP': 33696,\n", + " 'WDT': 1368,\n", + " 'WHADJP': 5,\n", + " 'WHADVP': 439,\n", + " 'WHNP': 1927,\n", + " 'WHPP': 153,\n", + " 'WP': 482,\n", + " 'WP$': 50,\n", + " 'WRB': 442,\n", + " 'X': 23,\n", + " '``': 1269}" + ] + }, + "execution_count": 13, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "d['pos_counter']" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": { + "collapsed": false + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "[[False False False False]\n", + " [False True False False]\n", + " [False False False False]]\n", + "[[0 2 2 0]\n", + " [2 2 0 2]\n", + " [2 0 0 0]]\n" + ] + } + ], + "source": [ + "from my.nltk_utils import tree2matrix, load_compressed_tree, find_max_f1_subtree, set_span\n", + "string = \"(ROOT(S(NP (PRP I))(VP (VBP am)(NP (NNP Sam)))(. .)))\"\n", + "tree = load_compressed_tree(string)\n", + "span = (1, 3)\n", + "set_span(tree)\n", + "subtree = find_max_f1_subtree(tree, span)\n", + "f = lambda t: t == subtree\n", + "g = lambda t: 1 if isinstance(t, str) else 2\n", + "a, b = tree2matrix(tree, f, dtype='bool')\n", + "c, d = tree2matrix(tree, g, dtype='int32')\n", + "print(a)\n", + "print(c)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.5.1" + } + }, + "nbformat": 4, + "nbformat_minor": 0 +} diff --git a/tensorflow/SQuAD/tree/trainer.py b/tensorflow/SQuAD/tree/trainer.py new file mode 100644 index 0000000..0dd85e3 --- /dev/null +++ b/tensorflow/SQuAD/tree/trainer.py @@ -0,0 +1,36 @@ +import tensorflow as tf + +from tree.model import Model + + +class Trainer(object): + def __init__(self, config, model): + assert isinstance(model, Model) + self.config = config + self.model = model + self.opt = tf.train.AdagradOptimizer(config.init_lr) + self.loss = model.get_loss() + self.var_list = model.get_var_list() + self.global_step = model.get_global_step() + self.ema_op = model.ema_op + self.summary = model.summary + self.grads = self.opt.compute_gradients(self.loss, var_list=self.var_list) + opt_op = self.opt.apply_gradients(self.grads, global_step=self.global_step) + + # Define train op + with tf.control_dependencies([opt_op]): + self.train_op = tf.group(self.ema_op) + + def get_train_op(self): + return self.train_op + + def step(self, sess, batch, get_summary=False): + assert isinstance(sess, tf.Session) + feed_dict = self.model.get_feed_dict(batch, True) + if get_summary: + loss, summary, train_op = \ + sess.run([self.loss, self.summary, self.train_op], feed_dict=feed_dict) + else: + loss, train_op = sess.run([self.loss, self.train_op], feed_dict=feed_dict) + summary = None + return loss, summary, train_op diff --git a/tensorflow/SQuAD/tree/visualizer.py b/tensorflow/SQuAD/tree/visualizer.py new file mode 100644 index 0000000..d2e503c --- /dev/null +++ b/tensorflow/SQuAD/tree/visualizer.py @@ -0,0 +1,122 @@ +import shutil +from collections import OrderedDict +import http.server +import socketserver +import argparse +import json +import os +import numpy as np +from tqdm import tqdm + +from jinja2 import Environment, FileSystemLoader + + +def bool_(string): + if string == 'True': + return True + elif string == 'False': + return False + else: + raise Exception() + +def get_args(): + parser = argparse.ArgumentParser() + parser.add_argument("--model_name", type=str, default='basic') + parser.add_argument("--data_type", type=str, default='dev') + parser.add_argument("--step", type=int, default=5000) + parser.add_argument("--template_name", type=str, default="visualizer.html") + parser.add_argument("--num_per_page", type=int, default=100) + parser.add_argument("--data_dir", type=str, default="data/squad") + parser.add_argument("--port", type=int, default=8000) + parser.add_argument("--host", type=str, default="0.0.0.0") + parser.add_argument("--open", type=str, default='False') + parser.add_argument("--run_id", type=str, default="0") + + args = parser.parse_args() + return args + + +def _decode(decoder, sent): + return " ".join(decoder[idx] for idx in sent) + + +def accuracy2_visualizer(args): + model_name = args.model_name + data_type = args.data_type + num_per_page = args.num_per_page + data_dir = args.data_dir + run_id = args.run_id.zfill(2) + step = args.step + + eval_path =os.path.join("out", model_name, run_id, "eval", "{}-{}.json".format(data_type, str(step).zfill(6))) + eval_ = json.load(open(eval_path, 'r')) + + _id = 0 + html_dir = "/tmp/list_results%d" % _id + while os.path.exists(html_dir): + _id += 1 + html_dir = "/tmp/list_results%d" % _id + + if os.path.exists(html_dir): + shutil.rmtree(html_dir) + os.mkdir(html_dir) + + cur_dir = os.path.dirname(os.path.realpath(__file__)) + templates_dir = os.path.join(cur_dir, 'templates') + env = Environment(loader=FileSystemLoader(templates_dir)) + env.globals.update(zip=zip, reversed=reversed) + template = env.get_template(args.template_name) + + data_path = os.path.join(data_dir, "data_{}.json".format(data_type)) + shared_path = os.path.join(data_dir, "shared_{}.json".format(data_type)) + data = json.load(open(data_path, 'r')) + shared = json.load(open(shared_path, 'r')) + + rows = [] + for i, (idx, yi, ypi) in enumerate(zip(*[eval_[key] for key in ('idxs', 'y', 'yp')])): + id_, q, rx = (data[key][idx] for key in ('ids', 'q', '*x')) + x = shared['x'][rx[0]][rx[1]] + ques = [" ".join(q)] + para = [[word for word in sent] for sent in x] + row = { + 'id': id_, + 'title': "Hello world!", + 'ques': ques, + 'para': para, + 'y': yi, + 'y2': yi, + 'yp': ypi, + 'yp2': ypi, + 'a': "" + } + rows.append(row) + + if i % num_per_page == 0: + html_path = os.path.join(html_dir, "%s.html" % str(i).zfill(8)) + + if (i + 1) % num_per_page == 0 or (i + 1) == len(eval_['y']): + var_dict = {'title': "Accuracy Visualization", + 'rows': rows + } + with open(html_path, "wb") as f: + f.write(template.render(**var_dict).encode('UTF-8')) + rows = [] + + os.chdir(html_dir) + port = args.port + host = args.host + # Overriding to suppress log message + class MyHandler(http.server.SimpleHTTPRequestHandler): + def log_message(self, format, *args): + pass + handler = MyHandler + httpd = socketserver.TCPServer((host, port), handler) + if args.open == 'True': + os.system("open http://%s:%d" % (args.host, args.port)) + print("serving at %s:%d" % (host, port)) + httpd.serve_forever() + + +if __name__ == "__main__": + ARGS = get_args() + accuracy2_visualizer(ARGS) \ No newline at end of file diff --git a/tensorflow/SQuAD/visualization/compare_models.py b/tensorflow/SQuAD/visualization/compare_models.py new file mode 100644 index 0000000..8dc9c7e --- /dev/null +++ b/tensorflow/SQuAD/visualization/compare_models.py @@ -0,0 +1,244 @@ +import numpy as np +from collections import Counter +import string +import re +import argparse +import os +import json +import nltk +from matplotlib_venn import venn2 +from matplotlib import pyplot as plt + + +class Question: + def __init__(self, id, question_text, ground_truth, model_names): + self.id = id + self.question_text = self.normalize_answer(question_text) + self.question_head_ngram = [] + self.question_tokens = nltk.word_tokenize(self.question_text) + for nc in range(3): + self.question_head_ngram.append(' '.join(self.question_tokens[0:nc])) + self.ground_truth = ground_truth + self.model_names = model_names + self.em = np.zeros(2) + self.f1 = np.zeros(2) + self.answer_text = [] + + def add_answers(self, answer_model_1, answer_model_2): + self.answer_text.append(answer_model_1) + self.answer_text.append(answer_model_2) + self.eval() + + def eval(self): + for model_count in range(2): + self.em[model_count] = self.metric_max_over_ground_truths(self.exact_match_score, self.answer_text[model_count], self.ground_truth) + self.f1[model_count] = self.metric_max_over_ground_truths(self.f1_score, self.answer_text[model_count], self.ground_truth) + + def normalize_answer(self, s): + """Lower text and remove punctuation, articles and extra whitespace.""" + def remove_articles(text): + return re.sub(r'\b(a|an|the)\b', ' ', text) + + def white_space_fix(text): + return ' '.join(text.split()) + + def remove_punc(text): + exclude = set(string.punctuation) + return ''.join(ch for ch in text if ch not in exclude) + + def lower(text): + return text.lower() + + return white_space_fix(remove_articles(remove_punc(lower(s)))) + + def f1_score(self, prediction, ground_truth): + prediction_tokens = self.normalize_answer(prediction).split() + ground_truth_tokens = self.normalize_answer(ground_truth).split() + common = Counter(prediction_tokens) & Counter(ground_truth_tokens) + num_same = sum(common.values()) + if num_same == 0: + return 0 + precision = 1.0 * num_same / len(prediction_tokens) + recall = 1.0 * num_same / len(ground_truth_tokens) + f1 = (2 * precision * recall) / (precision + recall) + return f1 + + def exact_match_score(self, prediction, ground_truth): + return (self.normalize_answer(prediction) == self.normalize_answer(ground_truth)) + + def metric_max_over_ground_truths(self, metric_fn, prediction, ground_truths): + scores_for_ground_truths = [] + for ground_truth in ground_truths: + score = metric_fn(prediction, ground_truth) + scores_for_ground_truths.append(score) + return max(scores_for_ground_truths) + + +def safe_dict_access(in_dict, in_key, default_string='some junk string'): + if in_key in in_dict: + return in_dict[in_key] + else: + return default_string + + +def aggregate_metrics(questions): + total = len(questions) + exact_match = np.zeros(2) + f1_scores = np.zeros(2) + + for mc in range(2): + exact_match[mc] = 100 * np.sum(np.array([questions[x].em[mc] for x in questions])) / total + f1_scores[mc] = 100 * np.sum(np.array([questions[x].f1[mc] for x in questions])) / total + + model_names = questions[list(questions.keys())[0]].model_names + print('\nAggregate Scores:') + for model_count in range(2): + print('Model {0} EM = {1:.2f}'.format(model_names[model_count], exact_match[model_count])) + print('Model {0} F1 = {1:.2f}'.format(model_names[model_count], f1_scores[model_count])) + + +def venn_diagram(questions, output_dir): + em_model1_ids = [x for x in questions if questions[x].em[0] == 1] + em_model2_ids = [x for x in questions if questions[x].em[1] == 1] + model_names = questions[list(questions.keys())[0]].model_names + print('\nVenn diagram') + + correct_model1 = em_model1_ids + correct_model2 = em_model2_ids + correct_model1_and_model2 = list(set(em_model1_ids).intersection(set(em_model2_ids))) + correct_model1_and_not_model2 = list(set(em_model1_ids) - set(em_model2_ids)) + correct_model2_and_not_model1 = list(set(em_model2_ids) - set(em_model1_ids)) + + print('{0} answers correctly = {1}'.format(model_names[0], len(correct_model1))) + print('{0} answers correctly = {1}'.format(model_names[1], len(correct_model2))) + print('Both answer correctly = {1}'.format(model_names[0], len(correct_model1_and_model2))) + print('{0} correct & {1} incorrect = {2}'.format(model_names[0], model_names[1], len(correct_model1_and_not_model2))) + print('{0} correct & {1} incorrect = {2}'.format(model_names[1], model_names[0], len(correct_model2_and_not_model1))) + + plt.clf() + venn_diagram_plot = venn2( + subsets=(len(correct_model1_and_not_model2), len(correct_model2_and_not_model1), len(correct_model1_and_model2)), + set_labels=('{0} correct'.format(model_names[0]), '{0} correct'.format(model_names[1]), 'Both correct'), + set_colors=('r', 'b'), + alpha=0.3, + normalize_to=1 + ) + plt.savefig(os.path.join(output_dir, 'venn_diagram.png')) + plt.close() + return correct_model1, correct_model2, correct_model1_and_model2, correct_model1_and_not_model2, correct_model2_and_not_model1 + + +def get_head_ngrams(questions, num_grams): + head_ngrams = [] + for question in questions.values(): + head_ngrams.append(question.question_head_ngram[num_grams]) + return head_ngrams + + +def get_head_ngram_frequencies(questions, head_ngrams, num_grams): + head_ngram_frequencies = {} + for current_ngram in head_ngrams: + head_ngram_frequencies[current_ngram] = 0 + for question in questions.values(): + head_ngram_frequencies[question.question_head_ngram[num_grams]] += 1 + return head_ngram_frequencies + + +def get_head_ngram_statistics(questions, correct_model1, correct_model2, correct_model1_and_model2, correct_model1_and_not_model2, correct_model2_and_not_model1, output_dir, num_grams=2, top_count=25): + # Head ngram statistics + head_ngrams = get_head_ngrams(questions, num_grams) + + # Get head_ngram_frequencies (hnf) + hnf_all = get_head_ngram_frequencies(questions, head_ngrams, num_grams) + hnf_correct_model1 = get_head_ngram_frequencies({qid: questions[qid] for qid in correct_model1}, head_ngrams, num_grams) + hnf_correct_model2 = get_head_ngram_frequencies({qid: questions[qid] for qid in correct_model2}, head_ngrams, num_grams) + hnf_correct_model1_and_model2 = get_head_ngram_frequencies({qid: questions[qid] for qid in correct_model1_and_model2}, head_ngrams, num_grams) + hnf_correct_model1_and_not_model2 = get_head_ngram_frequencies({qid: questions[qid] for qid in correct_model1_and_not_model2}, head_ngrams, num_grams) + hnf_correct_model2_and_not_model1 = get_head_ngram_frequencies({qid: questions[qid] for qid in correct_model2_and_not_model1}, head_ngrams, num_grams) + + sorted_bigrams_all = sorted(hnf_all.items(), key=lambda x: x[1], reverse=True) + top_bigrams = [x[0] for x in sorted_bigrams_all[0:top_count]] + + counts_total = [hnf_all[x] for x in top_bigrams] + counts_model1 = [hnf_correct_model1[x] for x in top_bigrams] + counts_model2 = [hnf_correct_model2[x] for x in top_bigrams] + counts_model1_and_model2 = [hnf_correct_model1_and_model2[x] for x in top_bigrams] + counts_model1_and_not_model2 = [hnf_correct_model1_and_not_model2[x] for x in top_bigrams] + counts_model2_and_not_model1 = [hnf_correct_model2_and_not_model1[x] for x in top_bigrams] + + top_bigrams_with_counts = [] + for cc in range(len(top_bigrams)): + top_bigrams_with_counts.append('{0} ({1})'.format(top_bigrams[cc], counts_total[cc])) + + plt.clf() + fig, ax = plt.subplots(figsize=(6, 10)) + + ylocs = list(range(top_count)) + counts_model1_percent = 100 * np.array(counts_model1) / np.array(counts_total) + plt.barh([top_count - x for x in ylocs], counts_model1_percent, height=0.4, alpha=0.5, color='#EE3224', label=top_bigrams) + counts_model2_percent = 100 * np.array(counts_model2) / np.array(counts_total) + plt.barh([top_count - x+0.4 for x in ylocs], counts_model2_percent, height=0.4, alpha=0.5, color='#2432EE', label=top_bigrams ) + ax.set_yticks([top_count - x + 0.4 for x in ylocs]) + ax.set_yticklabels(top_bigrams_with_counts) + ax.set_ylim([0.5, top_count+1]) + ax.set_xlim([0, 100]) + plt.subplots_adjust(left=0.28, right=0.9, top=0.9, bottom=0.1) + plt.xlabel('Percentage of questions with correct answers') + plt.ylabel('Top N-grams') + plt.savefig(os.path.join(output_dir, 'ngram_stats_{0}.png'.format(num_grams))) + plt.close() + + +def read_json(filename): + with open(filename) as filepoint: + data = json.load(filepoint) + return data + + +def compare_models(dataset_file, predictions_m1_file, predictions_m2_file, output_dir, name_m1='Model 1', name_m2='Model 2'): + dataset = read_json(dataset_file)['data'] + predictions_m1 = read_json(predictions_m1_file) + predictions_m2 = read_json(predictions_m2_file) + + # Read in data + total = 0 + questions = {} + for article in dataset: + for paragraph in article['paragraphs']: + for qa in paragraph['qas']: + current_question = Question(id=qa['id'], question_text=qa['question'], ground_truth=list(map(lambda x: x['text'], qa['answers'])), model_names=[name_m1, name_m2]) + current_question.add_answers(answer_model_1=safe_dict_access(predictions_m1, qa['id']), answer_model_2=safe_dict_access(predictions_m2, qa['id'])) + questions[current_question.id] = current_question + total += 1 + model_names = questions[list(questions.keys())[0]].model_names + print('Read in {0} questions'.format(total)) + + # Aggregate scores + aggregate_metrics(questions) + + # Venn diagram + correct_model1, correct_model2, correct_model1_and_model2, correct_model1_and_not_model2, correct_model2_and_not_model1 = venn_diagram(questions, output_dir=output_dir) + + # Head Unigram statistics + get_head_ngram_statistics(questions, correct_model1, correct_model2, correct_model1_and_model2, correct_model1_and_not_model2, + correct_model2_and_not_model1, output_dir, num_grams=1, top_count=10) + + # Head Bigram statistics + get_head_ngram_statistics(questions, correct_model1, correct_model2, correct_model1_and_model2, correct_model1_and_not_model2, + correct_model2_and_not_model1, output_dir, num_grams=2, top_count=10) + + +if __name__ == '__main__': + parser = argparse.ArgumentParser(description='Compare two QA models') + parser.add_argument('-dataset', action='store', dest='dataset', required=True, help='Dataset file') + parser.add_argument('-model1', action='store', dest='predictions_m1', required=True, help='Prediction file for model 1') + parser.add_argument('-model2', action='store', dest='predictions_m2', required=True, help='Prediction file for model 2') + parser.add_argument('-name1', action='store', dest='name_m1', help='Name for model 1') + parser.add_argument('-name2', action='store', dest='name_m2', help='Name for model 2') + parser.add_argument('-output', action='store', dest='output_dir', help='Output directory for visualizations') + results = parser.parse_args() + + if results.name_m1 is not None and results.name_m2 is not None: + compare_models(dataset_file=results.dataset, predictions_m1_file=results.predictions_m1, predictions_m2_file=results.predictions_m2, output_dir=results.output_dir, name_m1=results.name_m1, name_m2=results.name_m2) + else: + compare_models(dataset_file=results.dataset, predictions_m1_file=results.predictions_m1, predictions_m2_file=results.predictions_m2, output_dir=results.output_dir)