109 lines
3.5 KiB
Python
109 lines
3.5 KiB
Python
|
'''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)
|