tensorflow_estimator/python/estimator/canned/linear_testing_utils.py [731:820]: - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - def _assert_checkpoint(self, expected_global_step, expected_age_weight=None, expected_bias=None): shapes = { name: shape for (name, shape) in tf.train.list_variables(self._model_dir) } self.assertEqual([], shapes[tf.compat.v1.GraphKeys.GLOBAL_STEP]) self.assertEqual( expected_global_step, tf.train.load_variable(self._model_dir, tf.compat.v1.GraphKeys.GLOBAL_STEP)) self.assertEqual([1, 1], shapes[AGE_WEIGHT_NAME]) if expected_age_weight is not None: self.assertEqual(expected_age_weight, tf.train.load_variable(self._model_dir, AGE_WEIGHT_NAME)) self.assertEqual([1], shapes[BIAS_NAME]) if expected_bias is not None: self.assertEqual(expected_bias, tf.train.load_variable(self._model_dir, BIAS_NAME)) def testFromScratchWithDefaultOptimizer(self): # Create LinearRegressor. label = 5. age = 17 linear_regressor = self._linear_regressor_fn( feature_columns=(self._fc_lib.numeric_column('age'),), model_dir=self._model_dir) # Train for a few steps, and validate final checkpoint. num_steps = 10 linear_regressor.train( input_fn=lambda: ({ 'age': ((age,),) }, ((label,),)), steps=num_steps) self._assert_checkpoint(num_steps) def testTrainWithOneDimLabel(self): label_dimension = 1 batch_size = 20 feature_columns = [self._fc_lib.numeric_column('age', shape=(1,))] est = self._linear_regressor_fn( feature_columns=feature_columns, label_dimension=label_dimension, model_dir=self._model_dir) data_rank_1 = np.linspace(0., 2., batch_size, dtype=np.float32) self.assertEqual((batch_size,), data_rank_1.shape) train_input_fn = numpy_io.numpy_input_fn( x={'age': data_rank_1}, y=data_rank_1, batch_size=batch_size, num_epochs=None, shuffle=True) est.train(train_input_fn, steps=200) self._assert_checkpoint(200) def testTrainWithOneDimWeight(self): label_dimension = 1 batch_size = 20 feature_columns = [self._fc_lib.numeric_column('age', shape=(1,))] est = self._linear_regressor_fn( feature_columns=feature_columns, label_dimension=label_dimension, weight_column='w', model_dir=self._model_dir) data_rank_1 = np.linspace(0., 2., batch_size, dtype=np.float32) self.assertEqual((batch_size,), data_rank_1.shape) train_input_fn = numpy_io.numpy_input_fn( x={ 'age': data_rank_1, 'w': data_rank_1 }, y=data_rank_1, batch_size=batch_size, num_epochs=None, shuffle=True) est.train(train_input_fn, steps=200) self._assert_checkpoint(200) def testFromScratch(self): # Create LinearRegressor. label = 5. age = 17 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - tensorflow_estimator/python/estimator/canned/v1/linear_testing_utils_v1.py [847:936]: - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - def _assert_checkpoint(self, expected_global_step, expected_age_weight=None, expected_bias=None): shapes = { name: shape for (name, shape) in tf.train.list_variables(self._model_dir) } self.assertEqual([], shapes[tf.compat.v1.GraphKeys.GLOBAL_STEP]) self.assertEqual( expected_global_step, tf.train.load_variable(self._model_dir, tf.compat.v1.GraphKeys.GLOBAL_STEP)) self.assertEqual([1, 1], shapes[AGE_WEIGHT_NAME]) if expected_age_weight is not None: self.assertEqual(expected_age_weight, tf.train.load_variable(self._model_dir, AGE_WEIGHT_NAME)) self.assertEqual([1], shapes[BIAS_NAME]) if expected_bias is not None: self.assertEqual(expected_bias, tf.train.load_variable(self._model_dir, BIAS_NAME)) def testFromScratchWithDefaultOptimizer(self): # Create LinearRegressor. label = 5. age = 17 linear_regressor = self._linear_regressor_fn( feature_columns=(self._fc_lib.numeric_column('age'),), model_dir=self._model_dir) # Train for a few steps, and validate final checkpoint. num_steps = 10 linear_regressor.train( input_fn=lambda: ({ 'age': ((age,),) }, ((label,),)), steps=num_steps) self._assert_checkpoint(num_steps) def testTrainWithOneDimLabel(self): label_dimension = 1 batch_size = 20 feature_columns = [self._fc_lib.numeric_column('age', shape=(1,))] est = self._linear_regressor_fn( feature_columns=feature_columns, label_dimension=label_dimension, model_dir=self._model_dir) data_rank_1 = np.linspace(0., 2., batch_size, dtype=np.float32) self.assertEqual((batch_size,), data_rank_1.shape) train_input_fn = numpy_io.numpy_input_fn( x={'age': data_rank_1}, y=data_rank_1, batch_size=batch_size, num_epochs=None, shuffle=True) est.train(train_input_fn, steps=200) self._assert_checkpoint(200) def testTrainWithOneDimWeight(self): label_dimension = 1 batch_size = 20 feature_columns = [self._fc_lib.numeric_column('age', shape=(1,))] est = self._linear_regressor_fn( feature_columns=feature_columns, label_dimension=label_dimension, weight_column='w', model_dir=self._model_dir) data_rank_1 = np.linspace(0., 2., batch_size, dtype=np.float32) self.assertEqual((batch_size,), data_rank_1.shape) train_input_fn = numpy_io.numpy_input_fn( x={ 'age': data_rank_1, 'w': data_rank_1 }, y=data_rank_1, batch_size=batch_size, num_epochs=None, shuffle=True) est.train(train_input_fn, steps=200) self._assert_checkpoint(200) def testFromScratch(self): # Create LinearRegressor. label = 5. age = 17 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -