Tensorflow 1 and 2 Step Savers, And Base Classes

README.md

@@ -19,8 +19,19 @@ def create_graph(step: TensorflowV1ModelStep):
 
     tf.Variable(np.random.rand(), name='weight')
     tf.Variable(np.random.rand(), name='bias')
+
+    return tf.add(tf.multiply(step['data_inputs'], step['weight']), step['bias'])
+
+"""
+# Note: you can also return a tuple containing two elements : tensor for training (fit), tensor for inference (transform)
+def create_graph(step: TensorflowV1ModelStep)
+    # ...
+    decoder_outputs_training = create_training_decoder(step, encoder_state, decoder_cell)
+    decoder_outputs_inference = create_inference_decoder(step, encoder_state, decoder_cell)
+
+    return decoder_outputs_training, decoder_outputs_inference
+"""
 
-    tf.add(tf.multiply(step['data_inputs'], step['weight']), step['bias'], name='output')
 
 def create_loss(step: TensorflowV1ModelStep):
     return tf.reduce_sum(tf.pow(step['output'] - step['expected_outputs'], 2)) / (2 * N_SAMPLES)
@@ -32,7 +43,7 @@ model_step = TensorflowV1ModelStep(
     create_grah=create_graph,
     create_loss=create_loss,
     create_optimizer=create_optimizer,
-    has_expected_outputs=False
+    has_expected_outputs=True
 ).set_hyperparams(HyperparameterSamples({
     'learning_rate': 0.01
 })).set_hyperparams_space(HyperparameterSpace({
@@ -51,8 +62,8 @@ def create_model(step: Tensorflow2ModelStep):
 def create_optimizer(step: Tensorflow2ModelStep):
     return tf.keras.optimizers.Adam(0.1)
 
-def create_loss(step: Tensorflow2ModelStep, expected_outputs, actual_outputs):
-    return tf.reduce_mean(tf.abs(actual_outputs - expected_outputs))
+def create_loss(step: Tensorflow2ModelStep, expected_outputs, predicted_outputs):
+    return tf.reduce_mean(tf.abs(predicted_outputs - expected_outputs))
 
 model_step = Tensorflow2ModelStep(
     create_model=create_model,

neuraxle_tensorflow/tensorflow_v1.py

@@ -61,7 +61,7 @@ def __init__(
         self.variable_scope = variable_scope
         self.has_expected_outputs = has_expected_outputs
         self.create_feed_dict = create_feed_dict
-        self.loss = []
+        self.losses = []
         self.print_loss = print_loss
         if print_func is None:
             print_func = print
@@ -147,9 +147,9 @@ def fit_model(self, data_inputs, expected_outputs=None) -> BaseStep:
             feed_dict.update(additional_feed_dict_arguments)
 
         results = self.session.run([self['optimizer'], self['loss']], feed_dict=feed_dict)
-        self.loss.append(results[1])
+        self.losses.append(results[1])
         if self.print_loss:
-            self.print_func(self.loss[-1])
+            self.print_func(self.losses[-1])
 
         return self
 
@@ -164,9 +164,7 @@ def transform_model(self, data_inputs):
         :param data_inputs:
         :return:
         """
-        inference_output_name = 'output'
-        if len(self['inference_output'].get_shape().as_list()) > 0:
-            inference_output_name = 'inference_output'
+        inference_output_name = self._get_inference_output_name()
 
         feed_dict = {
             self['data_inputs']: data_inputs
@@ -176,6 +174,19 @@ def transform_model(self, data_inputs):
 
         return results
 
+    def _get_inference_output_name(self):
+        """
+        Return the output tensor name for inference (transform).
+        In create_graph, the user can return a tuple of two elements : the output tensor for training, and the output tensor for inference.
+
+        :return:
+        """
+        inference_output_name = 'output'
+        if len(self['inference_output'].get_shape().as_list()) > 0:
+            inference_output_name = 'inference_output'
+
+        return inference_output_name
+
     def __getitem__(self, item):
         """
         Get a graph tensor by name using get item.

requirements.txt

@@ -1 +1 @@
-Neuraxle
+neuraxle>=0.3.0

setup.py

@@ -19,7 +19,7 @@
 
 from setuptools import setup, find_packages
 
-from neuraxle import __version__ as _VERSION
+from neuraxle_tensorflow import __version__ as _VERSION
 
 with open('README.md') as _f:
     _README = _f.read()
@@ -29,6 +29,7 @@
     version=_VERSION,
     description='TensorFlow steps, savers, and utilities for Neuraxle. Neuraxle is a Machine Learning (ML) library for building neat pipelines, providing the right '
                 'abstractions to both ease research, development, and deployment of your ML applications.',
+    long_description_content_type='text/markdown',
     long_description=_README,
     classifiers=[
         "Development Status :: 3 - Alpha",
@@ -79,16 +80,16 @@
         "Topic :: Utilities",
         "Typing :: Typed"
     ],
-    url='https://github.com/Neuraxio/Neuraxle',
-    download_url='https://github.com/Neuraxio/Neuraxle/tarball/{}'.format(
+    url='https://github.com/Neuraxio/Neuraxle-Tensorflow',
+    download_url='https://github.com/Neuraxio/Neuraxle-Tensorflow/tarball/{}'.format(
         _VERSION),
     author='Neuraxio Inc.',
     author_email='guillaume.chevalier@neuraxio.com',
-    packages=find_packages(include=['neuraxle-tensorflow*']),
+    packages=find_packages(include=['neuraxle_tensorflow*']),
     test_suite="testing",
     setup_requires=["pytest-runner"],
     install_requires=[
-        'neuraxle>=0.2.1'
+        'neuraxle>=0.3.0'
     ],
     tests_require=["pytest", "pytest-cov"],
     include_package_data=True,

testing/test_tensorflow_v1.py

@@ -19,7 +19,7 @@ def create_graph(step: TensorflowV1ModelStep):
     tf.Variable(np.random.rand(), name='weight')
     tf.Variable(np.random.rand(), name='bias')
 
-    tf.add(tf.multiply(step['data_inputs'], step['weight']), step['bias'], name='output')
+    return tf.add(tf.multiply(step['data_inputs'], step['weight']), step['bias'])
 
 
 def create_loss(step: TensorflowV1ModelStep):