# Copyright 2017 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. # ============================================================================== import copy import os import pathlib import sys import weakref from absl.testing import parameterized from tensorflow.python.checkpoint import async_checkpoint_helper from tensorflow.python.checkpoint import checkpoint as trackable_utils from tensorflow.python.checkpoint import checkpoint_management from tensorflow.python.checkpoint import checkpoint_options from tensorflow.python.checkpoint import graph_view from tensorflow.python.checkpoint import save_util from tensorflow.python.eager import context from tensorflow.python.eager import def_function from tensorflow.python.framework import constant_op from tensorflow.python.framework import dtypes from tensorflow.python.framework import errors_impl from tensorflow.python.framework import ops from tensorflow.python.framework import stack from tensorflow.python.framework import test_util from tensorflow.python.ops import control_flow_ops from tensorflow.python.ops import init_ops from tensorflow.python.ops import resource_variable_ops from tensorflow.python.ops import state_ops from tensorflow.python.ops import template from tensorflow.python.ops import variable_scope from tensorflow.python.ops import variable_v1 from tensorflow.python.ops import variables as variables_lib from tensorflow.python.platform import test from tensorflow.python.platform import tf_logging as logging from tensorflow.python.saved_model import save as saved_model_save from tensorflow.python.trackable import autotrackable from tensorflow.python.trackable import base from tensorflow.python.training import checkpoint_utils from tensorflow.python.training import saver as saver_lib try: import psutil # pylint: disable=g-import-not-at-top psutil_import_succeeded = True except ImportError: psutil_import_succeeded = False class NonLayerTrackable(autotrackable.AutoTrackable): def __init__(self): super().__init__() self.a_variable = trackable_utils.add_variable( self, name="a_variable", shape=[]) class InterfaceTests(test.TestCase): @test_util.run_in_graph_and_eager_modes(assert_no_eager_garbage=True) def testAddVariable(self): obj = NonLayerTrackable() with self.assertRaisesRegex(ValueError, "do not specify shape"): trackable_utils.add_variable( obj, name="shape_specified_twice", shape=[], initializer=1) constant_initializer = trackable_utils.add_variable( obj, name="constant_initializer", initializer=1) with variable_scope.variable_scope("some_variable_scope"): ones_initializer = trackable_utils.add_variable( obj, name="ones_initializer", shape=[2], initializer=init_ops.ones_initializer(dtype=dtypes.float32)) bare_initializer = trackable_utils.add_variable( obj, name="bare_initializer", shape=[2, 2], dtype=dtypes.float64, initializer=init_ops.zeros_initializer) # Even in graph mode, there are no naming conflicts between objects, only # naming conflicts within an object. other_duplicate = resource_variable_ops.ResourceVariable( name="duplicate", initial_value=1.) duplicate = trackable_utils.add_variable( obj, name="duplicate", shape=[]) with self.assertRaisesRegex(ValueError, "'duplicate'.*already declared"): trackable_utils.add_variable(obj, name="duplicate", shape=[]) self.evaluate(trackable_utils.gather_initializers(obj)) self.assertEqual("constant_initializer:0", constant_initializer.name) self.assertEqual(1, self.evaluate(constant_initializer)) self.assertEqual("some_variable_scope/ones_initializer:0", ones_initializer.name) self.assertAllEqual([1, 1], self.evaluate(ones_initializer)) self.assertAllEqual([[0., 0.], [0., 0.]], self.evaluate(bare_initializer)) self.assertEqual("a_variable:0", obj.a_variable.name) self.assertEqual("duplicate:0", other_duplicate.name) if context.executing_eagerly(): # When executing eagerly, there's no uniquification of variable names. The # checkpoint name will be the same. self.assertEqual("duplicate:0", duplicate.name) else: # The .name attribute may be globally influenced, but the checkpoint name # won't be (tested below). self.assertEqual("duplicate_1:0", duplicate.name) expected_checkpoint_names = { "a_variable/.ATTRIBUTES/VARIABLE_VALUE", "bare_initializer/.ATTRIBUTES/VARIABLE_VALUE", "constant_initializer/.ATTRIBUTES/VARIABLE_VALUE", "duplicate/.ATTRIBUTES/VARIABLE_VALUE", "ones_initializer/.ATTRIBUTES/VARIABLE_VALUE", } actual_checkpoint_names = _get_all_checkpoint_names(obj) self.assertEqual(expected_checkpoint_names, set(actual_checkpoint_names)) def testInitNotCalled(self): class NoInit(autotrackable.AutoTrackable): def __init__(self): pass # __init__ for Trackable will be called implicitly. trackable_utils.add_variable(NoInit(), "var", shape=[]) def testShapeDtype(self): root = autotrackable.AutoTrackable() v1 = trackable_utils.add_variable( root, name="v1", initializer=3., dtype=dtypes.float64) self.assertEqual(dtypes.float64, v1.dtype) v2 = trackable_utils.add_variable( root, name="v2", shape=[3], initializer=init_ops.ones_initializer, dtype=dtypes.float64) self.assertEqual(dtypes.float64, v2.dtype) self.assertAllEqual([1., 1., 1.], self.evaluate(v2)) class _MirroringSaveable(saver_lib.BaseSaverBuilder.SaveableObject): def __init__(self, primary_variable, mirrored_variable, name): self._primary_variable = primary_variable self._mirrored_variable = mirrored_variable tensor = self._primary_variable.read_value() spec = saver_lib.BaseSaverBuilder.SaveSpec( tensor=tensor, slice_spec="", name=name) super().__init__(tensor, [spec], name) def restore(self, restored_tensors, restored_shapes): """Restore the same value into both variables.""" tensor, = restored_tensors return control_flow_ops.group( self._primary_variable.assign(tensor), self._mirrored_variable.assign(tensor)) class _OwnsMirroredVariables(base.Trackable): """A Trackable object which returns a more complex SaveableObject.""" def __init__(self): self.non_dep_variable = variable_scope.get_variable( name="non_dep_variable", initializer=6., use_resource=True) self.mirrored = variable_scope.get_variable( name="mirrored", initializer=15., use_resource=True) def _gather_saveables_for_checkpoint(self): def _saveable_factory(name=self.non_dep_variable.name): return _MirroringSaveable( primary_variable=self.non_dep_variable, mirrored_variable=self.mirrored, name=name) return {base.VARIABLE_VALUE_KEY: _saveable_factory} # The Saver sorts by name before parsing, so we need a name property. @property def name(self): return self.non_dep_variable.name def _get_all_checkpoint_names(root): serialized_tensors, _, _, _ = save_util.serialize_graph_view( graph_view.ObjectGraphView(root)) checkpoint_names = [] for tensor_dict in serialized_tensors.values(): checkpoint_names.extend(tensor_dict.keys()) return checkpoint_names class CheckpointingTests(parameterized.TestCase, test.TestCase): @parameterized.named_parameters( ("_enable_async_ckpt", True), ("_disable_async_ckpt", False) ) @test_util.run_in_graph_and_eager_modes def testMoreComplexSaveableReturned(self, enable_async_ckpt): v = _OwnsMirroredVariables() checkpoint = trackable_utils.Checkpoint(v=v) test_dir = self.get_temp_dir() prefix = os.path.join(test_dir, "ckpt") self.evaluate(v.non_dep_variable.assign(42.)) ckpt_options = checkpoint_options.CheckpointOptions( experimental_enable_async_checkpoint=enable_async_ckpt) save_path = checkpoint.save(file_prefix=prefix, options=ckpt_options) # TODO(chienchunh): Identify why sync needs to be called here. if enable_async_ckpt: checkpoint.sync() self.evaluate(v.non_dep_variable.assign(43.)) self.evaluate(v.mirrored.assign(44.)) checkpoint.restore(save_path).assert_consumed().initialize_or_restore() self.assertEqual(42., self.evaluate(v.non_dep_variable)) self.assertEqual(42., self.evaluate(v.mirrored)) self.evaluate(v.non_dep_variable.assign(44.)) save_path = checkpoint.save(file_prefix=prefix, options=ckpt_options) # TODO(chienchunh): Identify why sync needs to be called here. if enable_async_ckpt: checkpoint.sync() self.evaluate(v.non_dep_variable.assign(45.)) checkpoint.restore(save_path).assert_consumed().initialize_or_restore() self.assertEqual(44., self.evaluate(v.non_dep_variable)) self.assertEqual(44., self.evaluate(v.mirrored)) @test_util.run_in_graph_and_eager_modes def testMoreComplexSaveableReturnedWithGlobalName(self): # The same object can also be saved using the name-based saver. v = _OwnsMirroredVariables() saver = saver_lib.Saver(var_list=[v]) test_dir = self.get_temp_dir() prefix = os.path.join(test_dir, "ckpt") with self.cached_session() as sess: self.evaluate(v.non_dep_variable.assign(42.)) save_path = saver.save(sess, prefix) self.evaluate(v.non_dep_variable.assign(43.)) self.evaluate(v.mirrored.assign(44.)) saver.restore(sess, save_path) self.assertEqual(42., self.evaluate(v.non_dep_variable)) self.assertEqual(42., self.evaluate(v.mirrored)) @parameterized.named_parameters( ("_enable_async_ckpt", True), ("_disable_async_ckpt", False) ) @test_util.run_in_graph_and_eager_modes def testAssertConsumedNoCheckpoint(self, enable_async_ckpt): prefix = os.path.join(self.get_temp_dir(), "ckpt") v = variable_scope.get_variable(name="v", initializer=0.) self.evaluate(v.initializer) ckpt = trackable_utils.Checkpoint(v=v) self.evaluate(trackable_utils.gather_initializers(ckpt)) ckpt_options = checkpoint_options.CheckpointOptions( experimental_enable_async_checkpoint=enable_async_ckpt) save_path = ckpt.save(file_prefix=prefix, options=ckpt_options) status = ckpt.restore(save_path=save_path) del ckpt status.assert_consumed() def testDeepCopyCheckpoint(self): prefix = os.path.join(self.get_temp_dir(), "ckpt") v = variables_lib.Variable(1.) original_ckpt = trackable_utils.Checkpoint(v=v) copied_ckpt = copy.deepcopy(original_ckpt) copied_ckpt.v.assign(2.) self.assertAllClose(1., v) save_path = copied_ckpt.save(file_prefix=prefix) status = original_ckpt.restore(save_path=save_path) status.expect_partial() self.assertAllClose(2., v) def testDeepCopyCheckpointCopiesGraphViewRoot(self): original_ckpt = trackable_utils.Checkpoint(v=variables_lib.Variable(1.)) copied_ckpt = copy.deepcopy(original_ckpt) self.assertIs(copied_ckpt._saver._graph_view.root, copied_ckpt) self.assertIsNot(copied_ckpt._saver._graph_view.root, original_ckpt) self.assertIs(copied_ckpt._saver._graph_view.root.v, copied_ckpt.v) self.assertIsNot(copied_ckpt._saver._graph_view.root.v, original_ckpt.v) def testDeepCopyCheckpointCopiesGraphViewAttachedDependencies(self): root = autotrackable.AutoTrackable() child = variables_lib.Variable(1.) original_ckpt = trackable_utils.Checkpoint(root=root, child=child) copied_ckpt = copy.deepcopy(original_ckpt) object_graph_view = copied_ckpt._saver._graph_view attached_dependencies = { dependency.name: dependency.ref for dependency in object_graph_view.attached_dependencies } self.assertIs(object_graph_view.root, copied_ckpt.root) self.assertIsNot(object_graph_view.root, original_ckpt.root) self.assertIs(attached_dependencies["root"], copied_ckpt.root) self.assertIs(attached_dependencies["child"], copied_ckpt.child) self.assertIsNot(attached_dependencies["child"], original_ckpt.child) @test_util.run_in_graph_and_eager_modes def testPassingCheckpointOptions(self): localhost = "/job:localhost/device:CPU:0" options = checkpoint_options.CheckpointOptions( experimental_io_device=localhost) prefix = os.path.join(self.get_temp_dir(), "ckpt") v = variable_scope.get_variable(name="v", initializer=0.) self.evaluate(v.initializer) ckpt = trackable_utils.Checkpoint(v=v) self.evaluate(trackable_utils.gather_initializers(ckpt)) save_path = ckpt.save(file_prefix=prefix, options=options) status = ckpt.restore(save_path=save_path, options=options) del ckpt status.assert_consumed() # In graph mode, verify that the save and restore ops were set to run on # localhost. if not context.executing_eagerly(): for op in ops.get_default_graph().get_operations(): if op.type in ("SaveV2", "RestoreV2"): self.assertEqual(localhost, op.device) @test_util.run_in_graph_and_eager_modes def testFreezing(self): with test_util.use_gpu(): # Save an object-based checkpoint using a frozen saver directory = self.get_temp_dir() prefix = os.path.join(directory, "ckpt") v = resource_variable_ops.ResourceVariable(0, dtype=dtypes.int64) checkpoint = trackable_utils.Checkpoint(v=v) self.evaluate(v.assign(3)) # Create the save counter so assert_consumed doesn't complain about it not # existing in the checkpoint on restore. self.evaluate(checkpoint.save_counter.assign(12)) saver = trackable_utils.frozen_saver(checkpoint) with ops.device("cpu:0"): prefix_tensor = constant_op.constant(prefix) self.evaluate(saver.save(prefix_tensor)) self.evaluate(v.assign(10)) # Use the frozen saver to restore the same object graph self.evaluate(saver.restore(prefix_tensor)) self.assertEqual(3, self.evaluate(v)) # Restore using another frozen saver on an identical object graph del v, checkpoint, saver v = resource_variable_ops.ResourceVariable(0, dtype=dtypes.int64) checkpoint = trackable_utils.Checkpoint(v=v) saver = trackable_utils.frozen_saver(checkpoint) self.evaluate(saver.restore(prefix_tensor)) self.assertEqual(3, self.evaluate(v)) # Restore as an object-based checkpoint del v, checkpoint, saver checkpoint = trackable_utils.Checkpoint() status = checkpoint.restore(prefix) v = resource_variable_ops.ResourceVariable(0, dtype=dtypes.int64) if context.executing_eagerly(): self.assertEqual(12, self.evaluate(checkpoint.save_counter)) self.assertEqual(0, self.evaluate(v)) checkpoint.v = v status.assert_consumed().run_restore_ops() self.assertEqual(3, self.evaluate(v)) self.assertEqual(12, self.evaluate(checkpoint.save_counter)) @parameterized.named_parameters( ("_enable_async_ckpt", True), ("_disable_async_ckpt", False) ) @test_util.run_in_graph_and_eager_modes def testCustomNumbering(self, enable_async_ckpt): directory = self.get_temp_dir() prefix = os.path.join(directory, "ckpt") step = resource_variable_ops.ResourceVariable(0, dtype=dtypes.int64) checkpoint = trackable_utils.Checkpoint(step=step) ckpt_options = checkpoint_options.CheckpointOptions( experimental_enable_async_checkpoint=enable_async_ckpt) self.evaluate(step.initializer) for i in range(5): path = checkpoint.write("%s-%d" % (prefix, self.evaluate(step)), options=ckpt_options) expected_suffix = "-%d" % (2 * i,) if not path.endswith(expected_suffix): self.fail("%s should have suffix %s" % (path, expected_suffix)) self.evaluate(step.assign_add(2)) def testPartialRestoreWarningAttribute(self): with context.eager_mode(): original_root = trackable_utils.Checkpoint(v1=variables_lib.Variable(2.), v2=variables_lib.Variable(3.)) prefix = os.path.join(self.get_temp_dir(), "ckpt") save_path = original_root.save(prefix) partial_root = trackable_utils.Checkpoint(v1=base.Trackable(), v2=variables_lib.Variable(0.)) weak_partial_root = weakref.ref(partial_root) with test.mock.patch.object(logging, "warning") as mock_log: # Note: Unlike in testPartialRestoreWarningObject, the warning actually # prints immediately here, since all of the objects have been created # and there's no deferred restoration sitting around. partial_root.restore(save_path) self.assertEqual(3., partial_root.v2.numpy()) del partial_root self.assertIsNone(weak_partial_root()) messages = str(mock_log.call_args_list) self.assertIn("(root).v1", messages) self.assertNotIn("(root).v2", messages) self.assertIn("expect_partial()", messages) def testAttributeException(self): with context.eager_mode(): original_root = trackable_utils.Checkpoint(v1=variables_lib.Variable(2.), v2=variables_lib.Variable(3.)) prefix = os.path.join(self.get_temp_dir(), "ckpt") save_path = original_root.save(prefix) partial_root = trackable_utils.Checkpoint(v1=base.Trackable(), v2=variables_lib.Variable(0.)) status = partial_root.restore(save_path) with self.assertRaisesRegex(AssertionError, r"Unused attributes(.|\n)*\(root\).v1"): status.assert_consumed() def testSilencePartialWarning(self): with context.eager_mode(): original_root = trackable_utils.Checkpoint(v1=variables_lib.Variable(2.), v2=variables_lib.Variable(3.)) prefix = os.path.join(self.get_temp_dir(), "ckpt") save_path = original_root.save(prefix) partial_root = trackable_utils.Checkpoint(v1=variables_lib.Variable(0.)) weak_partial_root = weakref.ref(partial_root) weak_v1 = weakref.ref(partial_root.v1) partial_root.restore(save_path).expect_partial() self.assertEqual(2., partial_root.v1.numpy()) with test.mock.patch.object(logging, "warning") as mock_log: del partial_root self.assertIsNone(weak_partial_root()) self.assertIsNone(weak_v1()) self.assertEmpty(mock_log.call_args_list) def _get_checkpoint_name(self, name): root = autotrackable.AutoTrackable() trackable_utils.add_variable( root, name=name, shape=[1, 2], dtype=dtypes.float64) checkpoint_key = _get_all_checkpoint_names(root)[0] with ops.name_scope("root/" + checkpoint_key): pass # Make sure we can use this as an op name if we prefix it. return checkpoint_key @test_util.run_in_graph_and_eager_modes(assert_no_eager_garbage=True) def testVariableNameEscaping(self): suffix = "/.ATTRIBUTES/VARIABLE_VALUE" self.assertEqual(r"a.Sb.Sc" + suffix, self._get_checkpoint_name(r"a/b/c")) self.assertEqual(r"b" + suffix, self._get_checkpoint_name(r"b")) self.assertEqual(r"c.S" + suffix, self._get_checkpoint_name(r"c/")) self.assertEqual(r"d.S..S" + suffix, self._get_checkpoint_name(r"d/.S")) self.assertEqual(r"d.S..ATTRIBUTES.Sf" + suffix, self._get_checkpoint_name(r"d/.ATTRIBUTES/f")) @test_util.run_in_graph_and_eager_modes(assert_no_eager_garbage=True) def testNumberedPath(self): root = autotrackable.AutoTrackable() leaf = autotrackable.AutoTrackable() root.leaf = leaf trackable_utils.add_variable(leaf, name="v", shape=[]) checkpoint_key = _get_all_checkpoint_names(root)[0] self.assertEqual(r"leaf/v/.ATTRIBUTES/VARIABLE_VALUE", checkpoint_key) @test_util.run_in_graph_and_eager_modes def testLocalNameValidation(self): root = autotrackable.AutoTrackable() leaf = autotrackable.AutoTrackable() # Dots are escaped, which avoids conflicts with reserved names. root._track_trackable(leaf, name=".ATTRIBUTES") trackable_utils.add_variable(trackable=leaf, name="a", shape=[]) checkpoint_key = _get_all_checkpoint_names(root)[0] self.assertEqual("..ATTRIBUTES/a/.ATTRIBUTES/VARIABLE_VALUE", checkpoint_key) @test_util.run_in_graph_and_eager_modes def testLateDependencyTracking(self): class Dependency(autotrackable.AutoTrackable): def build(self): self.var = trackable_utils.add_variable( self, "var", initializer=0.) class LateDependencies(trackable_utils.Checkpoint): def add_dep(self): self.dep = Dependency() self.dep.build() original = LateDependencies() original.add_dep() self.evaluate(state_ops.assign(original.dep.var, 123.)) checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") save_path = original.save(checkpoint_prefix) load_into = LateDependencies() status = load_into.restore(save_path) status.assert_existing_objects_matched() with self.assertRaises(AssertionError): status.assert_consumed() load_into.add_dep() status.assert_consumed() status.assert_existing_objects_matched().run_restore_ops() self.assertEqual(123., self.evaluate(load_into.dep.var)) @test_util.run_in_graph_and_eager_modes def testDepAfterVar(self): class Dependency(autotrackable.AutoTrackable): def build(self): self.var = trackable_utils.add_variable( self, "var", initializer=0.) class DepAfterVar(trackable_utils.Checkpoint): def add_dep(self): dep = Dependency() dep.build() self.dep = dep dep_after_var = DepAfterVar() dep_after_var.add_dep() self.evaluate(state_ops.assign(dep_after_var.dep.var, -14.)) checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") save_path = dep_after_var.save(checkpoint_prefix) loaded_dep_after_var = DepAfterVar() status = loaded_dep_after_var.restore(save_path) loaded_dep_after_var.add_dep() status.assert_consumed() status.run_restore_ops() self.assertEqual(-14., self.evaluate(loaded_dep_after_var.dep.var)) @test_util.run_in_graph_and_eager_modes def testOverlappingRestores(self): checkpoint_directory = self.get_temp_dir() save_root = trackable_utils.Checkpoint() save_root.dep = autotrackable.AutoTrackable() save_root.dep.var = trackable_utils.add_variable( save_root.dep, name="var", initializer=0.) self.evaluate(state_ops.assign(save_root.dep.var, 12.)) first_path = save_root.save(os.path.join(checkpoint_directory, "first")) self.evaluate(state_ops.assign(save_root.dep.var, 13.)) second_path = save_root.save(os.path.join(checkpoint_directory, "second")) first_root = trackable_utils.Checkpoint() second_root = trackable_utils.Checkpoint() first_status = first_root.restore(first_path) second_status = second_root.restore(second_path) load_dep = autotrackable.AutoTrackable() load_dep.var = trackable_utils.add_variable( load_dep, name="var", shape=[]) first_root.dep = load_dep first_status.assert_consumed() first_status.run_restore_ops() self.assertEqual(12., self.evaluate(load_dep.var)) second_root.dep = load_dep second_status.assert_consumed() second_status.run_restore_ops() self.assertEqual(13., self.evaluate(load_dep.var)) # Try again with the order of the restore() reversed. The last restore # determines the final value. first_root = trackable_utils.Checkpoint() second_root = trackable_utils.Checkpoint() second_status = second_root.restore(second_path) first_status = first_root.restore(first_path) load_dep = autotrackable.AutoTrackable() load_dep.var = trackable_utils.add_variable( load_dep, name="var", shape=[]) first_root.dep = load_dep first_status.assert_consumed() first_status.run_restore_ops() self.assertEqual(12., self.evaluate(load_dep.var)) second_root.dep = load_dep second_status.assert_consumed() second_status.run_restore_ops() self.assertEqual(12., self.evaluate(load_dep.var)) @test_util.run_in_graph_and_eager_modes def testAmbiguousLoad(self): # Not OK to split one checkpoint object into two checkpoint_directory = self.get_temp_dir() save_root = trackable_utils.Checkpoint() save_root.dep_one = autotrackable.AutoTrackable() save_root.dep_two = autotrackable.AutoTrackable() dep_three = autotrackable.AutoTrackable() save_root.dep_one.dep_three = dep_three save_root.dep_two.dep_three = dep_three trackable_utils.add_variable(dep_three, name="var", initializer=0.) self.evaluate(trackable_utils.gather_initializers(save_root)) save_path = save_root.save(os.path.join(checkpoint_directory, "ckpt")) load_root = trackable_utils.Checkpoint() status = load_root.restore(save_path) load_root.dep_one = autotrackable.AutoTrackable() load_root.dep_two = autotrackable.AutoTrackable() load_root.dep_one.dep_three = autotrackable.AutoTrackable() load_root.dep_two.dep_three = autotrackable.AutoTrackable() trackable_utils.add_variable( load_root.dep_one.dep_three, name="var", initializer=0.) trackable_utils.add_variable( load_root.dep_two.dep_three, name="var", initializer=0.) with self.assertRaises(AssertionError): status.assert_consumed() with self.assertRaises(AssertionError): status.assert_existing_objects_matched() @test_util.run_in_graph_and_eager_modes def testObjectsCombined(self): # Currently fine to load two checkpoint objects into one Python object checkpoint_directory = self.get_temp_dir() save_root = trackable_utils.Checkpoint() save_root.dep_one = autotrackable.AutoTrackable() save_root.dep_two = autotrackable.AutoTrackable() trackable_utils.add_variable( save_root.dep_one, name="var1", initializer=32., dtype=dtypes.float64) trackable_utils.add_variable( save_root.dep_two, name="var2", initializer=64., dtype=dtypes.float64) self.evaluate(trackable_utils.gather_initializers(save_root)) save_path = save_root.save(os.path.join(checkpoint_directory, "ckpt")) load_root = trackable_utils.Checkpoint() load_root.dep_one = autotrackable.AutoTrackable() load_root.dep_two = load_root.dep_one v1 = trackable_utils.add_variable( load_root.dep_one, name="var1", shape=[], dtype=dtypes.float64) v2 = trackable_utils.add_variable( load_root.dep_one, name="var2", shape=[], dtype=dtypes.float64) status = load_root.restore( save_path).assert_consumed().assert_existing_objects_matched() status.run_restore_ops() self.assertEqual(32., self.evaluate(v1)) self.assertEqual(64., self.evaluate(v2)) @test_util.run_in_graph_and_eager_modes def testEmptyContainersIgnored(self): checkpoint_directory = self.get_temp_dir() save_root = trackable_utils.Checkpoint(a=[]) path = save_root.save(checkpoint_directory) load_root = trackable_utils.Checkpoint(b=[]) load_root.dep = [] load_root.dep.append([]) status = load_root.restore(path) status.assert_consumed() status.assert_existing_objects_matched() status.assert_nontrivial_match() @test_util.run_in_graph_and_eager_modes def testDependencyLoop(self): # Note: this test creates garbage during eager execution because it # purposefully creates a reference cycle. first = trackable_utils.Checkpoint() second = trackable_utils.Checkpoint() first.second = second second.first = first first.v = trackable_utils.add_variable( first, "v1", initializer=[3., 1., 4.]) second.v = trackable_utils.add_variable( second, "v2", initializer=[1., 1., 2., 3.]) self.evaluate(trackable_utils.gather_initializers(first)) checkpoint_directory = self.get_temp_dir() save_path = first.save(os.path.join(checkpoint_directory, "ckpt")) # Test deferred loading first_load = trackable_utils.Checkpoint() status = first_load.restore(save_path) second_load = autotrackable.AutoTrackable() first_load.second = second_load second_load.first = first_load with self.assertRaises(AssertionError): status.assert_consumed() first_load.v = trackable_utils.add_variable( first_load, "v1", shape=[3]) second_load.v = trackable_utils.add_variable( second_load, "v2", shape=[4]) status.assert_consumed() status.run_restore_ops() self.assertAllEqual([3., 1., 4.], self.evaluate(first_load.v)) self.assertAllEqual([1., 1., 2., 3.], self.evaluate(second_load.v)) # Test loading when variables have already been created self.evaluate(first_load.v.assign([2., 7., 1.])) self.assertAllEqual([2., 7., 1.], self.evaluate(first_load.v)) self.evaluate(second_load.v.assign([2., 7., 1., 8.])) self.assertAllEqual([2., 7., 1., 8.], self.evaluate(second_load.v)) status = first_load.restore(save_path).assert_consumed() status.run_restore_ops() self.assertAllEqual([3., 1., 4.], self.evaluate(first_load.v)) self.assertAllEqual([1., 1., 2., 3.], self.evaluate(second_load.v)) @test_util.run_in_graph_and_eager_modes def testRestoreOnAssign(self): checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") first = trackable_utils.Checkpoint() first.var1 = variables_lib.Variable(0., name="outside_var") first.var2 = variables_lib.Variable(0., name="blah") self.evaluate(first.var1.assign(4.)) self.evaluate(first.var2.assign(8.)) save_path = first.save(checkpoint_prefix) second = trackable_utils.Checkpoint() second.var2 = variables_lib.Variable(0., name="blah") status = second.restore(save_path) recreated_var1 = variables_lib.Variable(0., name="outside_var") status.run_restore_ops() self.assertEqual(8., self.evaluate(second.var2)) self.evaluate(recreated_var1.assign(-2.)) self.assertEqual(-2., self.evaluate(recreated_var1)) second.var1 = recreated_var1 status.run_restore_ops() self.assertEqual(4., self.evaluate(recreated_var1)) @test_util.run_in_graph_and_eager_modes def testCheckpointState(self): # No checkpoints are deleted by default checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") obj = autotrackable.AutoTrackable() obj.var = variable_scope.get_variable(name="v", initializer=0.) self.evaluate(trackable_utils.gather_initializers(obj)) saver = trackable_utils.Checkpoint(obj=obj) for _ in range(10): saver.save(checkpoint_prefix) expected_filenames = ["checkpoint"] for checkpoint_number in range(1, 11): expected_filenames.append("ckpt-%d.index" % (checkpoint_number,)) self.assertEmpty( set(expected_filenames) - set(os.listdir(checkpoint_directory))) @test_util.run_in_graph_and_eager_modes def testCheckpointStateChangingVarList(self): checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") obj = autotrackable.AutoTrackable() obj.var = variable_scope.get_variable(name="v", initializer=0.) self.evaluate(trackable_utils.gather_initializers(obj)) checkpoint = trackable_utils.Checkpoint(obj=obj) looped_variables = [] for iteration in range(10): new_variable = resource_variable_ops.ResourceVariable(iteration) self.evaluate(new_variable.initializer) setattr(checkpoint, "var_%d" % iteration, new_variable) checkpoint.save(checkpoint_prefix) looped_variables.append(new_variable) expected_filenames = ["checkpoint"] # We've copied the saver each time, but checkpoint management should still # be consistent. Nothing gets deleted. for checkpoint_number in range(1, 11): expected_filenames.append("ckpt-%d.index" % (checkpoint_number,)) self.assertEmpty( set(expected_filenames) - set(os.listdir(checkpoint_directory))) self.assertEqual( checkpoint_prefix + "-10", checkpoint_management.latest_checkpoint(checkpoint_directory)) # The checkpoint list only contains the most recent checkpoint, but they're # all on disk. This means we won't eventually run into proto size limits. self.assertEqual( [checkpoint_prefix + "-10"], (checkpoint_management.get_checkpoint_state(checkpoint_directory) .all_model_checkpoint_paths)) for v in looped_variables: self.evaluate(v.assign(314)) checkpoint.restore(checkpoint_prefix + "-6").run_restore_ops() self.assertEqual(314, self.evaluate(checkpoint.var_9)) self.assertEqual(314, self.evaluate(checkpoint.var_8)) self.assertEqual(314, self.evaluate(checkpoint.var_6)) self.assertEqual(5, self.evaluate(checkpoint.var_5)) self.assertEqual(1, self.evaluate(checkpoint.var_1)) self.assertEqual(0, self.evaluate(checkpoint.var_0)) checkpoint.restore(checkpoint_prefix + "-10").run_restore_ops() self.assertEqual(9, self.evaluate(checkpoint.var_9)) self.assertEqual(8, self.evaluate(checkpoint.var_8)) self.assertEqual(1, self.evaluate(checkpoint.var_1)) self.assertEqual(0, self.evaluate(checkpoint.var_0)) @test_util.run_in_graph_and_eager_modes def test_restore_after_adding_empty_trackable_data_structure(self): model = NonLayerTrackable() checkpoint = trackable_utils.Checkpoint(model=model) checkpoint.restore(None).initialize_or_restore() checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") save_path = checkpoint.save(checkpoint_prefix) del model, checkpoint model = NonLayerTrackable() model.dict = {"a": 1} model.list = {"b": 1} checkpoint = trackable_utils.Checkpoint(model=model) load_status = checkpoint.restore(save_path) load_status.assert_existing_objects_matched().run_restore_ops() @test_util.run_in_graph_and_eager_modes def test_write_checkpoint_path_str_from_function(self): checkpoint_prefix = os.path.join(self.get_temp_dir(), "ckpt") save_checkpoint = trackable_utils.Checkpoint(v=variables_lib.Variable(1.)) @def_function.function def _write_checkpoint(): save_path = save_checkpoint.write(checkpoint_prefix) return save_path self.evaluate([save_checkpoint.v.initializer]) self.evaluate(_write_checkpoint()) load_checkpoint = trackable_utils.Checkpoint(v=variables_lib.Variable(0.)) # Use read() instead of restore() which allows us to check that all # existing objects were loaded. status = load_checkpoint.read(checkpoint_prefix) status.assert_existing_objects_matched() status.assert_consumed() status.run_restore_ops() self.assertEqual(1., self.evaluate(load_checkpoint.v)) self.evaluate(save_checkpoint.v.assign(3.)) self.evaluate(_write_checkpoint()) self.evaluate(save_checkpoint.v.assign(0.)) status = load_checkpoint.read(checkpoint_prefix) status.assert_existing_objects_matched() status.assert_consumed() status.run_restore_ops() self.assertEqual(3., self.evaluate(load_checkpoint.v)) @test_util.run_in_graph_and_eager_modes def test_write_checkpoint_path_tensor_from_function(self): # Same as the previous test, but the path is a tensor not a python string. checkpoint_prefix = os.path.join(self.get_temp_dir(), "ckpt") checkpoint_prefix_tensor = constant_op.constant(checkpoint_prefix) save_checkpoint = trackable_utils.Checkpoint(v=variables_lib.Variable(1.)) @def_function.function def _write_checkpoint(prefix): save_path = save_checkpoint.write(prefix) return save_path self.evaluate([save_checkpoint.v.initializer]) self.evaluate(_write_checkpoint(checkpoint_prefix_tensor)) load_checkpoint = trackable_utils.Checkpoint(v=variables_lib.Variable(0.)) # Use read() instead of restore() which allows us to check that all # existing objects were loaded. status = load_checkpoint.read(checkpoint_prefix) status.assert_existing_objects_matched() status.assert_consumed() status.run_restore_ops() self.assertEqual(1., self.evaluate(load_checkpoint.v)) self.evaluate(save_checkpoint.v.assign(3.)) self.evaluate(_write_checkpoint(checkpoint_prefix_tensor)) self.evaluate(save_checkpoint.v.assign(0.)) status = load_checkpoint.read(checkpoint_prefix) status.assert_existing_objects_matched() status.assert_consumed() status.run_restore_ops() self.assertEqual(3., self.evaluate(load_checkpoint.v)) @test_util.run_in_graph_and_eager_modes def test_write_checkpoint_path_tensor_does_not_exist_from_function(self): # Same as the previous test, but the path is a tensor not a python string. checkpoint_prefix = os.path.join( self.get_temp_dir(), "DOES_NOT_EXIST", "ckpt") checkpoint_prefix_tensor = constant_op.constant(checkpoint_prefix) save_checkpoint = trackable_utils.Checkpoint(v=variables_lib.Variable(1.)) @def_function.function def _write_checkpoint(prefix): save_path = save_checkpoint.write(prefix) return save_path self.evaluate([save_checkpoint.v.initializer]) with self.assertRaises(errors_impl.NotFoundError): self.evaluate(_write_checkpoint(checkpoint_prefix_tensor)) @parameterized.named_parameters( ("_enable_async_ckpt", True), ("_disable_async_ckpt", False)) def test_inititialize_with_data_structures(self, enable_async_ckpt): checkpoint = trackable_utils.Checkpoint( a=[variables_lib.Variable(0.), variables_lib.Variable(1.)], b={"a": variables_lib.Variable(2.), "b": variables_lib.Variable(3.)}) checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") ckpt_options = checkpoint_options.CheckpointOptions( experimental_enable_async_checkpoint=enable_async_ckpt) save_path = checkpoint.save(file_prefix=checkpoint_prefix, options=ckpt_options) load_checkpoint = trackable_utils.Checkpoint( a=[variables_lib.Variable(4.), variables_lib.Variable(5.)], b={"a": variables_lib.Variable(6.), "b": variables_lib.Variable(7.)}) # When async checkpoint is enabled, we need to first make sure that the # checkpoint saving is fully complete before the checkpoint file can be # loaded by another checkpoint instance. Calling checkpoint.restore() is a # trick to make sure its async thread is joined. if enable_async_ckpt: checkpoint.restore(save_path) load_checkpoint.restore(save_path) self.assertAllClose(self.evaluate(load_checkpoint.a), [0, 1]) self.assertAllClose(self.evaluate(load_checkpoint.b), {"a": 2, "b": 3}) def _create_trackable(self): class Model(autotrackable.AutoTrackable): def __init__(self): self.v = variables_lib.Variable(2.) def __call__(self, x): return self.v * x return Model() def test_initialize_with_root_object(self): model = self._create_trackable() input_value = constant_op.constant([[3.]]) expected_output = self.evaluate(model(input_value)) model.deferred_variable = variables_lib.Variable(5.) checkpoint = trackable_utils.Checkpoint(model) checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") save_path = checkpoint.save(checkpoint_prefix) new_model = self._create_trackable() load_checkpoint = trackable_utils.Checkpoint(new_model) load_checkpoint.restore(save_path) self.assertAllClose(expected_output, new_model(input_value)) new_model.deferred_variable = variables_lib.Variable(1.) self.assertEqual(self.evaluate(new_model.deferred_variable), 5) def test_initialize_with_root_object_and_kwargs(self): model = self._create_trackable() model.v.assign(3.) separate_variable = variables_lib.Variable(5.) with self.assertRaisesRegex(ValueError, "root.v already exists"): trackable_utils.Checkpoint(model, v=separate_variable) checkpoint = trackable_utils.Checkpoint( model, separate_variable=separate_variable) checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") save_path = checkpoint.save(checkpoint_prefix) # Case 1: Loading checkpoint with same configuration. new_model = self._create_trackable() separate_variable = variables_lib.Variable(1.) load_checkpoint = trackable_utils.Checkpoint( new_model, separate_variable=separate_variable) load_checkpoint.restore(save_path).assert_consumed() self.assertEqual(self.evaluate(new_model.v), 3) self.assertEqual(self.evaluate(separate_variable), 5) self.assertEqual(self.evaluate(load_checkpoint.save_counter), 1) # Case 2: Loading checkpoint where v and separate_variable are swapped: # v is not attached to the root, while separate variable is attached to root new_model = autotrackable.AutoTrackable() new_model.separate_variable = variables_lib.Variable(200.) v = variables_lib.Variable(100.) load_checkpoint = trackable_utils.Checkpoint(new_model, v=v) load_checkpoint.restore(save_path).assert_consumed() self.assertEqual(self.evaluate(v), 3) self.assertEqual(self.evaluate(new_model.separate_variable), 5) self.assertEqual(self.evaluate(load_checkpoint.save_counter), 1) # Case 3: Loading checkpoint where no root object is specified separate_variable = variables_lib.Variable(200.) v = variables_lib.Variable(100.) load_checkpoint = trackable_utils.Checkpoint( v=v, separate_variable=separate_variable) load_checkpoint.restore(save_path).assert_consumed() self.assertEqual(self.evaluate(v), 3) self.assertEqual(self.evaluate(new_model.separate_variable), 5) self.assertEqual(self.evaluate(load_checkpoint.save_counter), 1) def test_checkpoint_saved_model_compatibility(self): model = self._create_trackable() input_value = constant_op.constant([[3.]]) expected_output = self.evaluate(model(input_value)) model.deferred_variable = variables_lib.Variable(5.) saved_model_dir = os.path.join(self.get_temp_dir(), "saved_model") saved_model_save.save(model, saved_model_dir) new_model = self._create_trackable() load_checkpoint = trackable_utils.Checkpoint(new_model) with self.assertRaisesRegex( errors_impl.NotFoundError, "Error when restoring from checkpoint or SavedModel"): load_checkpoint.restore(saved_model_dir + "no").expect_partial() load_checkpoint.restore(saved_model_dir).expect_partial() self.assertAllClose(expected_output, new_model(input_value)) new_model.deferred_variable = variables_lib.Variable(1.) self.assertEqual(self.evaluate(new_model.deferred_variable), 5) def test_deferred_dependency_avoids_reference_cycles(self): # Tests that there are no reference cycles when running garbage collection. # Python uses reference counts as the primary garbage collector, which will # not delete and finalize (__del__) objects in a cycle. The deletion is # eventually triggered by gc, which only runs when the garbage has reached # a certain threshold. delete_counter = 0 class TrackableWithDel(autotrackable.AutoTrackable): def __del__(self): nonlocal delete_counter delete_counter += 1 x = autotrackable.AutoTrackable() x.v = variables_lib.Variable(100.) x.has_del = TrackableWithDel() checkpoint = trackable_utils.Checkpoint(x) checkpoint_prefix = os.path.join(self.get_temp_dir(), "ckpt") save_path = checkpoint.save(checkpoint_prefix) self.assertEqual(delete_counter, 0) del checkpoint del x self.assertEqual(delete_counter, 1) no_v = autotrackable.AutoTrackable() no_v.has_del = TrackableWithDel() checkpoint = trackable_utils.Checkpoint(no_v) checkpoint.restore(save_path).expect_partial() del checkpoint del no_v self.assertEqual(delete_counter, 2) def test_defer_objects_with_values_only(self): # Tests that deferred dependencies are only added if the node in the # object graph has children or checkpointed values. root = autotrackable.AutoTrackable() root.branch_with_value = autotrackable.AutoTrackable() root.branch_with_value.v = variables_lib.Variable(5.0) root.branch_no_value = autotrackable.AutoTrackable() root.branch_no_value.child = autotrackable.AutoTrackable() root.v = variables_lib.Variable(1.0) checkpoint = trackable_utils.Checkpoint(model=root) checkpoint_prefix = os.path.join(self.get_temp_dir(), "ckpt") save_path = checkpoint.save(checkpoint_prefix) new_root = autotrackable.AutoTrackable() checkpoint = trackable_utils.Checkpoint(model=new_root) checkpoint.restore(save_path) # root should have two nodes with values/children (`branch-with_value`/`v`). self.assertLen(new_root._deferred_dependencies, 2) new_root.branch_no_value = autotrackable.AutoTrackable() self.assertLen(new_root._deferred_dependencies, 2) new_root.branch_with_value = autotrackable.AutoTrackable() self.assertLen(new_root._deferred_dependencies, 1) new_root.v = variables_lib.Variable(1.0) self.assertEmpty(new_root._deferred_dependencies, 1) def test_root_arg(self): root = autotrackable.AutoTrackable() root.v = variables_lib.Variable(1) w = variables_lib.Variable(2) y = variables_lib.Variable(3) root_ckpt = trackable_utils.Checkpoint(root=root, w=w, y=y) root2 = autotrackable.AutoTrackable() root2.w = variables_lib.Variable(4) v2 = variables_lib.Variable(5) z = variables_lib.Variable(6) root2_ckpt = trackable_utils.Checkpoint(root=root2, v=v2, z=z) root_save_path = root_ckpt.save(os.path.join(self.get_temp_dir(), "root_ckpt")) root2_save_path = root2_ckpt.save(os.path.join(self.get_temp_dir(), "root2_ckpt")) root_ckpt.restore(root2_save_path) root2_ckpt.restore(root_save_path) self.assertEqual(root.v.numpy(), 5) self.assertEqual(w.numpy(), 4) self.assertEqual(y.numpy(), 3) self.assertEqual(root2.w.numpy(), 2) self.assertEqual(v2.numpy(), 1) self.assertEqual(z.numpy(), 6) def test_weakref_root(self): root = autotrackable.AutoTrackable() root.v = variables_lib.Variable(1) ref = root.v.ref() ckpt = trackable_utils.Checkpoint(root=weakref.ref(root)) save_path = ckpt.save(os.path.join(self.get_temp_dir(), "ckpt")) root.v.assign(2) ckpt.restore(save_path) self.assertEqual(root.v.numpy(), 1) del root # Verifying if the variable is only referenced from `ref`. # We expect the reference counter to be 1, but `sys.getrefcount` reports # one higher reference counter because a temporary is created when we call # sys.getrefcount(). Hence check if the number returned is 2. # https://docs.python.org/3/library/sys.html#sys.getrefcount self.assertEqual(sys.getrefcount(ref.deref()), 2) def test_restore_incompatible_shape(self): v = variables_lib.Variable([1.0, 1.0]) w = variables_lib.Variable([1.0]) ckpt = trackable_utils.Checkpoint(v=v) save_path = ckpt.save(os.path.join(self.get_temp_dir(), "ckpt")) with self.assertRaisesRegex(ValueError, "incompatible tensor with shape"): trackable_utils.Checkpoint(v=w).restore(save_path) def test_save_restore_fspath(self): v = variables_lib.Variable(1.0) w = variables_lib.Variable(0.0) ckpt = trackable_utils.Checkpoint(v=v) prefix = pathlib.Path(self.get_temp_dir()) / "ckpt" save_path = ckpt.save(prefix) save_path = pathlib.Path(save_path) ckpt2 = trackable_utils.Checkpoint(v=w) ckpt2.restore(save_path) self.assertEqual(ckpt.v.numpy(), 1.0) def test_read_write_fspath(self): v = variables_lib.Variable(1.0) w = variables_lib.Variable(0.0) ckpt = trackable_utils.Checkpoint(v=v) prefix = pathlib.Path(self.get_temp_dir()) / "ckpt" save_path = ckpt.write(prefix) save_path = pathlib.Path(save_path) ckpt2 = trackable_utils.Checkpoint(v=w) ckpt2.read(save_path) self.assertEqual(ckpt.v.numpy(), 1.0) @test_util.run_deprecated_v1 def test_save_in_graph_but_no_session(self): v = variables_lib.Variable(1.0) ckpt = trackable_utils.Checkpoint(v=v) self.evaluate(v.initializer) prefix = pathlib.Path(self.get_temp_dir()) / "ckpt" with stack.default_session(None): with self.assertRaisesRegex(RuntimeError, "create a session"): ckpt.write(prefix) def test_ckpt_files_closed_after_restoration(self): if not psutil_import_succeeded: self.skipTest( "psutil is required to check that we've closed our files.") root = autotrackable.AutoTrackable() root.v = variables_lib.Variable(1) ckpt = trackable_utils.Checkpoint(root=root) save_path = ckpt.save(os.path.join(self.get_temp_dir(), "ckpt")) root2 = autotrackable.AutoTrackable() ckpt2 = trackable_utils.Checkpoint(root=root2) ckpt2.restore(save_path) proc = psutil.Process() for file in proc.open_files(): self.assertNotIn(save_path, file[0]) def testLookupCache(self): """Ensure that Checkpoint.restore passes cached dependencies to lookup.""" root = autotrackable.AutoTrackable() root.v1 = variables_lib.Variable(1) root.v2 = variables_lib.Variable(2) root.v3 = variables_lib.Variable(3) ckpt = trackable_utils.Checkpoint(model=root) save_path = ckpt.save(os.path.join(self.get_temp_dir(), "ckpt")) called_with_cache = [] class LookupOverride(autotrackable.AutoTrackable): def _lookup_dependency(self, name, cached_dependencies=None): if cached_dependencies is not None: called_with_cache.append(name) return super()._lookup_dependency(name, cached_dependencies) root2 = LookupOverride() ckpt2 = trackable_utils.Checkpoint(model=root2) ckpt2.restore(save_path) self.assertCountEqual(called_with_cache, ["v1", "v2", "v3"]) @parameterized.named_parameters( ("_enable_async_ckpt", True), ("_disable_async_ckpt", False) ) def testCallbackWithManager(self, enable_async_ckpt): """Tests experimental_write_callback with a checkpoint manager.""" # 1. Define checkpoint and manager accordingly v = variables_lib.Variable(1.) if enable_async_ckpt: ckpt = async_checkpoint_helper.AsyncCheckpointHelper( trackable_utils.Checkpoint, v=v ) else: ckpt = trackable_utils.Checkpoint(v=v) checkpoint_manager = checkpoint_management.CheckpointManager( checkpoint=ckpt, directory=os.path.join(self.get_temp_dir(), "ckpt"), max_to_keep=None, checkpoint_name="test-callback", ) # 2. Define 2 callbacks, which will be executed in order as stated in the # expected behavior of `CheckpointOptions.experimental_write_callbacks` testing_list = [] test_str = "callback 2 was here" # Define callback 1 that takes in 1 argument def my_callback_1(save_path): testing_list.append(save_path) # Define callback 2 that takes in 0 argument def my_callback_2(): testing_list.append(test_str) # 3. Save with `options` options = checkpoint_options.CheckpointOptions( experimental_write_callbacks=[my_callback_1, my_callback_2] ) save_path = checkpoint_manager.save(options=options) # 4. Assert results if enable_async_ckpt: checkpoint_manager.sync() # otherwise callbacks may not have finished # Ensure that user's options is not mutated by internal mechanisms. Here, # we would internally register a callback `_record_and_sweep_state()`. # Users should not have access to it, hence length still being 2. self.assertLen(options.experimental_write_callbacks, 2) # Ensure `_record_and_sweep_state()` executes and sets `_latest_checkpoint` self.assertEqual(save_path, checkpoint_manager._latest_checkpoint) # Ensure my_callback_1 is executed first self.assertEqual(save_path, testing_list[0]) # Ensure my_callback_2 is executed second self.assertEqual(test_str, testing_list[1]) # Ensure nothing else is written to `testing_list` self.assertLen(testing_list, 2) @parameterized.named_parameters( ("_async_ckpt_save", True, False, True), ("_async_ckpt_write", True, False, False), ("_regular_ckptV1_save", False, True, True), ("_regular_ckptV1_write", False, True, False), ("_regular_ckptV2_save", False, False, True), ("_regular_ckptV2_write", False, False, False) ) def testCallbackWithoutManager(self, enable_async_ckpt, use_v1, use_save): """Tests experimental_write_callback without using a checkpoint manager.""" # Note that the underlying checkpoint of `AsyncCheckpoint.save()` will call # `Checkpoint.save()`. # The underlying checkpoint of `AsyncCheckpoint.write()` will call # `Checkpoint.write()`. # 1. Define checkpoint instance accordingly v = variables_lib.Variable(1.) prefix = os.path.join(self.get_temp_dir(), "ckpt") if enable_async_ckpt: ckpt = async_checkpoint_helper.AsyncCheckpointHelper( trackable_utils.Checkpoint, v=v ) else: if use_v1: ckpt = trackable_utils.CheckpointV1(v=v) else: ckpt = trackable_utils.Checkpoint(v=v) # 2. Define 2 callbacks, which will be executed in order as stated in the # expected behavior of `CheckpointOptions.experimental_write_callbacks` testing_list = [] test_str = "callback 2 was here" # Define callback 1 that takes in 1 argument def my_callback_1(save_path): testing_list.append(save_path) # Define callback 2 that takes in 0 argument def my_callback_2(): testing_list.append(test_str) # 3. Save with `options` options = checkpoint_options.CheckpointOptions( experimental_write_callbacks=[my_callback_1, my_callback_2] ) if use_save: save_path = ckpt.save(prefix, options=options) else: save_path = ckpt.write(prefix, options=options) # 4. Assert results if enable_async_ckpt: ckpt.sync() # otherwise callbacks may not have finished # Ensure that user's options is not mutated by internal mechanisms. Here, # if we are using regular checkpoint's save(), we would internally register # a callback `_update_checkpoint_state_internal()`. Users should not have # access to it, hence length still being 2. self.assertLen(options.experimental_write_callbacks, 2) # Ensure my_callback_1 is executed first self.assertEqual(save_path, testing_list[0]) # Ensure my_callback_2 is executed second self.assertEqual(test_str, testing_list[1]) # Ensure nothing else is written to `testing_list` self.assertLen(testing_list, 2) def test_callback_argument_error(self): """Ensure passing in a callback with more than 1 argument raises error.""" # Define callback 1 that takes in 1 argument def my_callback_1(save_path): return save_path # Define callback 2 that takes in 2 arguments (would raise error) def my_callback_2(save_path, another_argument): return save_path, another_argument with self.assertRaises(AssertionError): _ = checkpoint_options.CheckpointOptions( experimental_write_callbacks=[my_callback_1, my_callback_2] ) def test_checkpoint_options_copyable(self): """Ensure that `CheckpointOptions` can be copied with `copy.deepcopy()`.""" def my_callback(save_path): return save_path def my_callback_2(save_path): return save_path + "some string" options_original = checkpoint_options.CheckpointOptions( experimental_io_device="CPU:0", enable_async=True, experimental_write_callbacks=[my_callback] ) options_copy = copy.copy(options_original) options_copy.enable_async = False options_copy.experimental_io_device = "CPU:1" options_copy.experimental_write_callbacks.append(my_callback_2) # Check that the original options instance is not affected self.assertEqual(options_original.experimental_io_device, "CPU:0") self.assertEqual(options_original.enable_async, True) self.assertLen(options_original.experimental_write_callbacks, 1) class SerializeToTensorTest(test.TestCase): @test_util.run_in_graph_and_eager_modes def test_keys_and_metadata(self): class MultiTensor(base.Trackable): def __init__(self, v1, v2): self.v1 = v1 self.v2 = v2 def _serialize_to_tensors(self): return {"v1": self.v1, "v2": self.v2} def _restore_from_tensors(self, restored_tensors): return control_flow_ops.group( self.v1.assign(restored_tensors["v1"]), self.v2.assign(restored_tensors["v2"])) root = MultiTensor(variables_lib.Variable(1), variables_lib.Variable(2)) child = MultiTensor(variables_lib.Variable(3), variables_lib.Variable(4)) ckpt = trackable_utils.Checkpoint(root=root, child=child) self.evaluate([root.v1.initializer, root.v2.initializer, child.v1.initializer, child.v2.initializer]) save_path = ckpt.save(os.path.join(self.get_temp_dir(), "ckpt")) # Check the checkpoint contents and metadata. reader = checkpoint_utils.load_checkpoint(save_path) object_proto = trackable_utils.object_metadata(save_path) root_attributes = object_proto.nodes[0].attributes self.assertLen(root_attributes, 2) self.assertDictEqual( {"v1": "/.ATTRIBUTES/v1", "v2": "/.ATTRIBUTES/v2"}, {attr.name: attr.checkpoint_key for attr in root_attributes}) self.assertEqual(1, reader.get_tensor("/.ATTRIBUTES/v1")) self.assertEqual(2, reader.get_tensor("/.ATTRIBUTES/v2")) child_attributes = object_proto.nodes[1].attributes self.assertLen(child_attributes, 2) self.assertDictEqual( {"v1": "child/.ATTRIBUTES/v1", "v2": "child/.ATTRIBUTES/v2"}, {attr.name: attr.checkpoint_key for attr in child_attributes}) self.assertEqual(3, reader.get_tensor("child/.ATTRIBUTES/v1")) self.assertEqual(4, reader.get_tensor("child/.ATTRIBUTES/v2")) # Try restoring the checkpoint. self.evaluate([root.v1.assign(0), root.v2.assign(0), child.v1.assign(0), child.v2.assign(0)]) ckpt.restore(save_path).assert_consumed().run_restore_ops() self.assertAllEqual([1, 2, 3, 4], self.evaluate([root.v1, root.v2, child.v1, child.v2])) @test_util.run_in_graph_and_eager_modes def test_reference_variable(self): # Test that refvariable is compatible with tf1 saver / tf2 checkpoint. with self.cached_session() as sess: root = autotrackable.AutoTrackable() root.v = variable_v1.VariableV1(5, use_resource=False) sess.run(root.v.initializer) ckpt = trackable_utils.Checkpoint(root) ckpt_path = os.path.join(self.get_temp_dir(), "ckpt") ckpt.write(ckpt_path) sess.run(root.v.assign(10)) saver = saver_lib.Saver(var_list=[root.v]) save_path = saver.save(sess, os.path.join(self.get_temp_dir(), "saver")) ckpt.read(ckpt_path).assert_consumed().run_restore_ops() self.assertEqual(5, sess.run(root.v)) saver.restore(sess, save_path) self.assertEqual(10, sess.run(root.v)) class TemplateTests(parameterized.TestCase, test.TestCase): @test_util.run_in_graph_and_eager_modes def test_trackable_save_restore_nested(self): def _inner_template(): v = variable_scope.get_variable( "v", shape=[1], initializer=init_ops.zeros_initializer()) return v def _outer_template(): first_inner = template.make_template("i1", _inner_template) second_inner = template.make_template("i2", _inner_template) v1 = first_inner() v2 = second_inner() v3 = second_inner() return (first_inner, second_inner), (v1, v2, v3) with variable_scope.variable_scope("ignored"): save_template = template.make_template("s1", _outer_template) save_root = trackable_utils.Checkpoint(my_template=save_template) (inner_template_one, inner_template_two), _ = save_template() self.evaluate(inner_template_one.variables[0].assign([20.])) self.evaluate(inner_template_two.variables[0].assign([25.])) checkpoint_directory = self.get_temp_dir() checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt") save_path = save_root.save(checkpoint_prefix) load_template = template.make_template("s2", _outer_template) load_root = trackable_utils.Checkpoint(my_template=load_template) status = load_root.restore(save_path) (inner_template_one, inner_template_two), (v1, v2, v3) = load_template() outer_template_dependencies = load_root.my_template._trackable_children() self.assertLen(outer_template_dependencies, 2) self.assertDictEqual({"i1": inner_template_one, "i2": inner_template_two}, outer_template_dependencies) self.assertLen(inner_template_one._trackable_children(), 1) self.assertIn("v", inner_template_one._trackable_children()) self.assertLen(inner_template_two._trackable_children(), 1) self.assertIn("v", inner_template_two._trackable_children()) status.assert_consumed().run_restore_ops() self.assertAllEqual([20.], self.evaluate(v1)) self.assertAllEqual([25.], self.evaluate(v2)) self.assertAllEqual([25.], self.evaluate(v3)) if __name__ == "__main__": ops.enable_eager_execution() test.main()