Remove pfor dependency in BUILD file, and strengthen unit tests for clip_and_aggregate_gradients.py.

PiperOrigin-RevId: 482050282

tensorflow_privacy/privacy/optimizers/BUILD

@@ -15,7 +15,6 @@ py_library(
         "clip_and_aggregate_gradients.py",
     ],
     srcs_version = "PY3",
-    deps = ["//third_party/tensorflow/python/ops/parallel_for:control_flow_ops"],
 )
 
 py_library(

tensorflow_privacy/privacy/optimizers/clip_and_aggregate_gradients.py

@@ -191,9 +191,8 @@ def clip_and_aggregate_gradients(
     output = control_flow_ops.pfor(loop_fn, target_size)
   except ValueError as err:
     raise ValueError(
-        'Encountered an exception while vectorizing the '
+        'Encountered an exception while vectorizing the jacobian computation. '
-        'batch_jacobian computation. Vectorization can be disabled by '
+        'Consider using a non-vectorized optimizer instead.') from err
-        'setting experimental_use_pfor to False.') from err
   grads = []
   for i, out in enumerate(output):
     if out is not None:

tensorflow_privacy/privacy/optimizers/clip_and_aggregate_gradients_test.py

@@ -22,8 +22,8 @@ from tensorflow_privacy.privacy.optimizers import clip_and_aggregate_gradients a
 class ClipAndAggregateGradientsTest(tf.test.TestCase, parameterized.TestCase):
   """Tests clip_and_aggreate_gradients."""
 
-  def _get_loss_and_vars_fn(self, n, keepdims=False):
+  def _get_loss_and_vars(self, n, keepdims=False, is_callable=False):
-    """Returns the function for creating the loss and variables."""
+    """Returns the loss and variable tensors or callables."""
     # The "model" here consists of both sparse and dense parameters to make sure
     # `clip_and_aggregate_gradients` computes the gradients in the correct way
     # and in the right format. The sparse layer is the embedding layer `emb0`,
@@ -51,9 +51,20 @@ class ClipAndAggregateGradientsTest(tf.test.TestCase, parameterized.TestCase):
       return 0.5 * tf.reduce_sum(
           input_tensor=tf.math.squared_difference(val0, val1), axis=1)
 
-    def _loss_and_vars_fn():
+    if is_callable:
-      # We concatenate the embeddings with some constant values to make sure
+
-      # backprop does only go through those gathered indices.
+      def loss_fn():
+        val0 = tf.concat([emb0(ind0), tf.constant([[0.0, 0.0]])], axis=0)
+        return tf.reduce_sum(
+            tf.reshape(_loss(data0, val0) + _loss(data1, var1), [n, -1]),
+            keepdims=keepdims,
+            axis=1)
+
+      def vars_fn():
+        return (emb0.embeddings, var1, dummy_var)
+
+      return loss_fn, vars_fn
+    else:
       val0 = tf.concat([emb0(ind0), tf.constant([[0.0, 0.0]])], axis=0)
       loss = tf.reduce_sum(
           tf.reshape(_loss(data0, val0) + _loss(data1, var1), [n, -1]),
@@ -61,8 +72,6 @@ class ClipAndAggregateGradientsTest(tf.test.TestCase, parameterized.TestCase):
           axis=1)
       return loss, (emb0.embeddings, var1, dummy_var)
 
-    return _loss_and_vars_fn
-
   def _get_true_grads(self,
                       n,
                       normalize=False,
@@ -130,15 +139,50 @@ class ClipAndAggregateGradientsTest(tf.test.TestCase, parameterized.TestCase):
       (3, False, 1.0, 'sum', 2, 'zero'),
       (1, True, 0.5, 'mean', 3, 'none'),
   )
-  def testCorrect(self, n, normalize, l2_norm_clip, agg_method,
+  def testCorrectTensor(self, n, normalize, l2_norm_clip, agg_method,
                         keep_sparse_threshold, unconnected):
-    """Tests the correctness of the computation."""
+    """Tests the correctness of the computation for tensors."""
-    loss_and_vars_fn = self._get_loss_and_vars_fn(n)
     true_grads = self._get_true_grads(n, normalize, l2_norm_clip, agg_method,
                                       unconnected)
 
-    with tf.GradientTape() as tape:
+    tape = tf.GradientTape()
-      loss, test_vars = loss_and_vars_fn()
+    with tape:
+      loss, test_vars = self._get_loss_and_vars(n)
+    results = cag.clip_and_aggregate_gradients(
+        tape,
+        loss,
+        test_vars,
+        normalize=normalize,
+        l2_norm_clip=l2_norm_clip,
+        aggregate_method=agg_method,
+        unconnected_gradients=unconnected,
+        keep_sparse_threshold=keep_sparse_threshold)
+    for r, t in zip(results, true_grads):
+      if t is None:
+        self.assertIsNone(r)
+      else:
+        r = self._to_dense_array(r)
+        self.assertAllCloseAccordingToType(r, t)
+
+  @parameterized.parameters(
+      (6, False, None, 'mean', -1, 'none'),
+      (6, True, None, 'sum', 1, 'none'),
+      (2, False, None, 'sum', 3, 'none'),
+      (2, True, 100.0, 'mean', 1, 'zero'),
+      (3, False, 1.0, 'sum', 2, 'zero'),
+      (1, True, 0.5, 'mean', 3, 'none'),
+  )
+  def testCorrectCallable(self, n, normalize, l2_norm_clip, agg_method,
+                          keep_sparse_threshold, unconnected):
+    """Tests the correctness of the computation for callables."""
+    true_grads = self._get_true_grads(n, normalize, l2_norm_clip, agg_method,
+                                      unconnected)
+
+    loss_fn, vars_fn = self._get_loss_and_vars(n, is_callable=True)
+    tape = tf.GradientTape()
+    with tape:
+      loss = loss_fn()
+      test_vars = vars_fn()
     results = cag.clip_and_aggregate_gradients(
         tape,
         loss,
@@ -163,11 +207,10 @@ class ClipAndAggregateGradientsTest(tf.test.TestCase, parameterized.TestCase):
   )
   def testTargetShape(self, n, keepdims):
     """Tests target gets vectorized regardless of their original shape."""
-    loss_and_vars_fn = self._get_loss_and_vars_fn(n, keepdims)
     true_grads = self._get_true_grads(n)
 
     with tf.GradientTape() as tape:
-      loss, test_vars = loss_and_vars_fn()
+      loss, test_vars = self._get_loss_and_vars(n, keepdims)
       results = cag.clip_and_aggregate_gradients(tape, loss, test_vars)
     for r, t in zip(results, true_grads):
       if t is None:
@@ -184,9 +227,8 @@ class ClipAndAggregateGradientsTest(tf.test.TestCase, parameterized.TestCase):
   )
   def testSparse(self, keep_sparse_threshold):
     """Tests the outcome is in the desired (dense or sparse) tensor form."""
-    loss_and_vars_fn = self._get_loss_and_vars_fn(3)
     with tf.GradientTape() as tape:
-      loss, test_vars = loss_and_vars_fn()
+      loss, test_vars = self._get_loss_and_vars(3)
       results = cag.clip_and_aggregate_gradients(
           tape,
           loss,