def testKeyValueOverrideBadKey(self): """Tests that overwriting with a bad key causes an exception.""" pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() configs = self._create_and_load_test_configs(pipeline_config) hparams = tf.contrib.training.HParams( **{"train_config.no_such_field": 10}) with self.assertRaises(ValueError): config_util.merge_external_params_with_configs(configs, hparams)
def testOverwriteBatchSizeWithBadValueType(self): """Tests that overwriting with a bad valuye type causes an exception.""" pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.train_config.batch_size = 2 configs = self._create_and_load_test_configs(pipeline_config) # Type should be an integer, but we're passing a string "10". hparams = tf.contrib.training.HParams( **{"train_config.batch_size": "10"}) with self.assertRaises(TypeError): config_util.merge_external_params_with_configs(configs, hparams)
def testNewFocalLossParameters(self): """Tests that the loss weight ratio is updated appropriately.""" original_alpha = 1.0 original_gamma = 1.0 new_alpha = 0.3 new_gamma = 2.0 hparams = tf.contrib.training.HParams(focal_loss_alpha=new_alpha, focal_loss_gamma=new_gamma) pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() classification_loss = pipeline_config.model.ssd.loss.classification_loss classification_loss.weighted_sigmoid_focal.alpha = original_alpha classification_loss.weighted_sigmoid_focal.gamma = original_gamma _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) configs = config_util.merge_external_params_with_configs( configs, hparams) classification_loss = configs["model"].ssd.loss.classification_loss self.assertAlmostEqual( new_alpha, classification_loss.weighted_sigmoid_focal.alpha) self.assertAlmostEqual( new_gamma, classification_loss.weighted_sigmoid_focal.gamma)
def testUpdateMaskTypeForAllInputConfigs(self): original_mask_type = input_reader_pb2.NUMERICAL_MASKS new_mask_type = input_reader_pb2.PNG_MASKS pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() train_config = pipeline_config.train_input_reader train_config.mask_type = original_mask_type eval_1 = pipeline_config.eval_input_reader.add() eval_1.mask_type = original_mask_type eval_1.name = "eval_1" eval_2 = pipeline_config.eval_input_reader.add() eval_2.mask_type = original_mask_type eval_2.name = "eval_2" _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"mask_type": new_mask_type} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) self.assertEqual(configs["train_input_config"].mask_type, new_mask_type) for eval_input_config in configs["eval_input_configs"]: self.assertEqual(eval_input_config.mask_type, new_mask_type)
def testOverwriteBatchSizeWithKeyValue(self): """Tests that batch size is overwritten based on key/value.""" pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.train_config.batch_size = 2 configs = self._create_and_load_test_configs(pipeline_config) hparams = tf.contrib.training.HParams( **{"train_config.batch_size": 10}) configs = config_util.merge_external_params_with_configs( configs, hparams) new_batch_size = configs["train_config"].batch_size self.assertEqual(10, new_batch_size)
def _get_configs_for_model(model_name): """Returns configurations for model.""" filename = get_pipeline_config_path(model_name) data_path = _get_data_path() label_map_path = _get_labelmap_path() configs = config_util.get_configs_from_pipeline_file(filename) override_dict = { 'train_input_path': data_path, 'eval_input_path': data_path, 'label_map_path': label_map_path } configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) return configs
def testUseMovingAverageForEval(self): use_moving_averages_orig = False pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.eval_config.use_moving_averages = use_moving_averages_orig _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"eval_with_moving_averages": True} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) self.assertEqual(True, configs["eval_config"].use_moving_averages)
def _get_configs_for_model(model_name): """Returns configurations for model.""" fname = os.path.join(tf.resource_loader.get_data_files_path(), 'samples/configs/' + model_name + '.config') label_map_path = os.path.join(tf.resource_loader.get_data_files_path(), 'data/pet_label_map.pbtxt') data_path = os.path.join(tf.resource_loader.get_data_files_path(), 'test_data/pets_examples.record') configs = config_util.get_configs_from_pipeline_file(fname) override_dict = { 'train_input_path': data_path, 'eval_input_path': data_path, 'label_map_path': label_map_path } return config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict)
def testNewBatchSizeWithClipping(self): """Tests that batch size is clipped to 1 from below.""" original_batch_size = 2 hparams = tf.contrib.training.HParams(batch_size=0.5) pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.train_config.batch_size = original_batch_size _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) configs = config_util.merge_external_params_with_configs( configs, hparams) new_batch_size = configs["train_config"].batch_size self.assertEqual(1, new_batch_size) # Clipped to 1.0.
def testNewBatchSize(self): """Tests that batch size is updated appropriately.""" original_batch_size = 2 hparams = tf.contrib.training.HParams(batch_size=16) pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.train_config.batch_size = original_batch_size _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) configs = config_util.merge_external_params_with_configs( configs, hparams) new_batch_size = configs["train_config"].batch_size self.assertEqual(16, new_batch_size)
def testTrainShuffle(self): """Tests that `train_shuffle` keyword arguments are applied correctly.""" original_shuffle = True desired_shuffle = False pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.train_input_reader.shuffle = original_shuffle _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"train_shuffle": desired_shuffle} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) train_shuffle = configs["train_input_config"].shuffle self.assertEqual(desired_shuffle, train_shuffle)
def testMergingKeywordArguments(self): """Tests that keyword arguments get merged as do hyperparameters.""" original_num_train_steps = 100 desired_num_train_steps = 10 pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.train_config.num_steps = original_num_train_steps _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"train_steps": desired_num_train_steps} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) train_steps = configs["train_config"].num_steps self.assertEqual(desired_num_train_steps, train_steps)
def testNewMomentumOptimizerValue(self): """Tests that new momentum value is updated appropriately.""" original_momentum_value = 0.4 hparams = tf.contrib.training.HParams(momentum_optimizer_value=1.1) pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() optimizer_config = pipeline_config.train_config.optimizer.rms_prop_optimizer optimizer_config.momentum_optimizer_value = original_momentum_value _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) configs = config_util.merge_external_params_with_configs( configs, hparams) optimizer_config = configs["train_config"].optimizer.rms_prop_optimizer new_momentum_value = optimizer_config.momentum_optimizer_value self.assertAlmostEqual(1.0, new_momentum_value) # Clipped to 1.0.
def testNewTrainInputPathList(self): """Tests that train input path can be overwritten with multiple files.""" original_train_path = ["path/to/data"] new_train_path = ["another/path/to/data", "yet/another/path/to/data"] pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() reader_config = pipeline_config.train_input_reader.tf_record_input_reader reader_config.input_path.extend(original_train_path) _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"train_input_path": new_train_path} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) reader_config = configs["train_input_config"].tf_record_input_reader final_path = reader_config.input_path self.assertEqual(new_train_path, final_path)
def testErrorOverwritingMultipleInputConfig(self): original_shuffle = False new_shuffle = True pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() eval_1 = pipeline_config.eval_input_reader.add() eval_1.shuffle = original_shuffle eval_1.name = "eval_1" eval_2 = pipeline_config.eval_input_reader.add() eval_2.shuffle = original_shuffle eval_2.name = "eval_2" _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"eval_shuffle": new_shuffle} with self.assertRaises(ValueError): configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict)
def testOverwriteAllEvalNumEpochs(self): original_num_epochs = 10 new_num_epochs = 1 pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.eval_input_reader.add( ).num_epochs = original_num_epochs pipeline_config.eval_input_reader.add( ).num_epochs = original_num_epochs _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"eval_num_epochs": new_num_epochs} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) for eval_input_config in configs["eval_input_configs"]: self.assertEqual(new_num_epochs, eval_input_config.num_epochs)
def testOverwriteAllEvalSampling(self): original_num_eval_examples = 1 new_num_eval_examples = 10 pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.eval_input_reader.add().sample_1_of_n_examples = ( original_num_eval_examples) pipeline_config.eval_input_reader.add().sample_1_of_n_examples = ( original_num_eval_examples) _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"sample_1_of_n_eval_examples": new_num_eval_examples} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) for eval_input_config in configs["eval_input_configs"]: self.assertEqual(new_num_eval_examples, eval_input_config.sample_1_of_n_examples)
def testOverWriteRetainOriginalImages(self): """Tests that `train_shuffle` keyword arguments are applied correctly.""" original_retain_original_images = True desired_retain_original_images = False pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.eval_config.retain_original_images = ( original_retain_original_images) _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = { "retain_original_images_in_eval": desired_retain_original_images } configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) retain_original_images = configs["eval_config"].retain_original_images self.assertEqual(desired_retain_original_images, retain_original_images)
def testNewMaskType(self): """Tests that mask type can be overwritten in input readers.""" original_mask_type = input_reader_pb2.NUMERICAL_MASKS new_mask_type = input_reader_pb2.PNG_MASKS pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() train_input_reader = pipeline_config.train_input_reader train_input_reader.mask_type = original_mask_type eval_input_reader = pipeline_config.eval_input_reader.add() eval_input_reader.mask_type = original_mask_type _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"mask_type": new_mask_type} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) self.assertEqual(new_mask_type, configs["train_input_config"].mask_type) self.assertEqual(new_mask_type, configs["eval_input_configs"][0].mask_type)
def testDontOverwriteEmptyLabelMapPath(self): """Tests that label map path will not by overwritten with empty string.""" original_label_map_path = "path/to/original/label_map" new_label_map_path = "" pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() train_input_reader = pipeline_config.train_input_reader train_input_reader.label_map_path = original_label_map_path eval_input_reader = pipeline_config.eval_input_reader.add() eval_input_reader.label_map_path = original_label_map_path _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"label_map_path": new_label_map_path} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) self.assertEqual(original_label_map_path, configs["train_input_config"].label_map_path) self.assertEqual(original_label_map_path, configs["eval_input_configs"][0].label_map_path)
def testGenericConfigOverride(self): """Tests generic config overrides for all top-level configs.""" # Set one parameter for each of the top-level pipeline configs: pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.model.ssd.num_classes = 1 pipeline_config.train_config.batch_size = 1 pipeline_config.eval_config.num_visualizations = 1 pipeline_config.train_input_reader.label_map_path = "/some/path" pipeline_config.eval_input_reader.add().label_map_path = "/some/path" pipeline_config.graph_rewriter.quantization.weight_bits = 1 pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") _write_config(pipeline_config, pipeline_config_path) # Override each of the parameters: configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) hparams = tf.contrib.training.HParams( **{ "model.ssd.num_classes": 2, "train_config.batch_size": 2, "train_input_config.label_map_path": "/some/other/path", "eval_config.num_visualizations": 2, "graph_rewriter_config.quantization.weight_bits": 2 }) configs = config_util.merge_external_params_with_configs( configs, hparams) # Ensure that the parameters have the overridden values: self.assertEqual(2, configs["model"].ssd.num_classes) self.assertEqual(2, configs["train_config"].batch_size) self.assertEqual("/some/other/path", configs["train_input_config"].label_map_path) self.assertEqual(2, configs["eval_config"].num_visualizations) self.assertEqual( 2, configs["graph_rewriter_config"].quantization.weight_bits)
def testNewLabelMapPath(self): """Tests that label map path can be overwritten in input readers.""" original_label_map_path = "path/to/original/label_map" new_label_map_path = "path//to/new/label_map" pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() train_input_reader = pipeline_config.train_input_reader train_input_reader.label_map_path = original_label_map_path eval_input_reader = pipeline_config.eval_input_reader.add() eval_input_reader.label_map_path = original_label_map_path _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) override_dict = {"label_map_path": new_label_map_path} configs = config_util.merge_external_params_with_configs( configs, kwargs_dict=override_dict) self.assertEqual(new_label_map_path, configs["train_input_config"].label_map_path) for eval_input_config in configs["eval_input_configs"]: self.assertEqual(new_label_map_path, eval_input_config.label_map_path)
def testNewClassificationLocalizationWeightRatio(self): """Tests that the loss weight ratio is updated appropriately.""" original_localization_weight = 0.1 original_classification_weight = 0.2 new_weight_ratio = 5.0 hparams = tf.contrib.training.HParams( classification_localization_weight_ratio=new_weight_ratio) pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.model.ssd.loss.localization_weight = ( original_localization_weight) pipeline_config.model.ssd.loss.classification_weight = ( original_classification_weight) _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) configs = config_util.merge_external_params_with_configs( configs, hparams) loss = configs["model"].ssd.loss self.assertAlmostEqual(1.0, loss.localization_weight) self.assertAlmostEqual(new_weight_ratio, loss.classification_weight)
def _assertOptimizerWithNewLearningRate(self, optimizer_name): """Asserts successful updating of all learning rate schemes.""" original_learning_rate = 0.7 learning_rate_scaling = 0.1 warmup_learning_rate = 0.07 hparams = tf.contrib.training.HParams(learning_rate=0.15) pipeline_config_path = os.path.join(self.get_temp_dir(), "pipeline.config") # Constant learning rate. pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() optimizer = getattr(pipeline_config.train_config.optimizer, optimizer_name) _update_optimizer_with_constant_learning_rate(optimizer, original_learning_rate) _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) configs = config_util.merge_external_params_with_configs( configs, hparams) optimizer = getattr(configs["train_config"].optimizer, optimizer_name) constant_lr = optimizer.learning_rate.constant_learning_rate self.assertAlmostEqual(hparams.learning_rate, constant_lr.learning_rate) # Exponential decay learning rate. pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() optimizer = getattr(pipeline_config.train_config.optimizer, optimizer_name) _update_optimizer_with_exponential_decay_learning_rate( optimizer, original_learning_rate) _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) configs = config_util.merge_external_params_with_configs( configs, hparams) optimizer = getattr(configs["train_config"].optimizer, optimizer_name) exponential_lr = optimizer.learning_rate.exponential_decay_learning_rate self.assertAlmostEqual(hparams.learning_rate, exponential_lr.initial_learning_rate) # Manual step learning rate. pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() optimizer = getattr(pipeline_config.train_config.optimizer, optimizer_name) _update_optimizer_with_manual_step_learning_rate( optimizer, original_learning_rate, learning_rate_scaling) _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) configs = config_util.merge_external_params_with_configs( configs, hparams) optimizer = getattr(configs["train_config"].optimizer, optimizer_name) manual_lr = optimizer.learning_rate.manual_step_learning_rate self.assertAlmostEqual(hparams.learning_rate, manual_lr.initial_learning_rate) for i, schedule in enumerate(manual_lr.schedule): self.assertAlmostEqual( hparams.learning_rate * learning_rate_scaling**i, schedule.learning_rate) # Cosine decay learning rate. pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() optimizer = getattr(pipeline_config.train_config.optimizer, optimizer_name) _update_optimizer_with_cosine_decay_learning_rate( optimizer, original_learning_rate, warmup_learning_rate) _write_config(pipeline_config, pipeline_config_path) configs = config_util.get_configs_from_pipeline_file( pipeline_config_path) configs = config_util.merge_external_params_with_configs( configs, hparams) optimizer = getattr(configs["train_config"].optimizer, optimizer_name) cosine_lr = optimizer.learning_rate.cosine_decay_learning_rate self.assertAlmostEqual(hparams.learning_rate, cosine_lr.learning_rate_base) warmup_scale_factor = warmup_learning_rate / original_learning_rate self.assertAlmostEqual(hparams.learning_rate * warmup_scale_factor, cosine_lr.warmup_learning_rate)