test_batch_accuracy, batch_loss, test_summaries, test_step = sess.run(
[tensors['accuracy_test'], tensors['loss_test'], tensors['summary_test_op'],
tensors['global_step']],
feed_dict={tensors['image_place']: test_batch_data,
tensors['label_place']: test_batch_label})
test_accuracy += test_batch_accuracy
########################################
########## Write summaries #############
########################################
# Write the summaries
test_summary_writer.add_summary(test_summaries, global_step=test_step)
# # Write the specific summaries for training phase.
# train_summary_writer.add_summary(train_image_summary, global_step=training_step)
#################################################
########## Plot the progressive bar #############
#################################################
progress = float(batch_num + 1) / total_batch_test
progress_bar.print_progress(progress, epoch_num=1, loss=batch_loss)
######################################################################
########## Calculate the accuracy for the whole test set #############
######################################################################
test_accuracy_total = test_accuracy / float(total_batch_test)
print("Testing Accuracy={:.5f}".format(test_accuracy_total))
def train(**keywords):
"""
This function run the session whether in training or evaluation mode.
NOTE: **keywords is defined in order to make the code easily changable.
WARNING: All the arguments for the **keywords must be defined when calling this function.
**keywords:
:param sess: The default session.
:param saver: The saver operator to save and load the model weights.
:param tensors: The tensors dictionary defined by the graph.
:param data: The data structure.
:param train_dir: The training dir which is a reference for saving the logs and model checkpoints.
:param finetuning: If fine tuning should be done or random initialization is needed.
:param num_epochs: Number of epochs for training.
:param online_test: If the testing is done while training.
:param checkpoint_dir: The directory of the checkpoints.
:param batch_size: The training batch size.
:return:
Run the session.
"""
# The prefix for checkpoint files
checkpoint_prefix = 'model'
###################################################################
########## Defining the summary writers for train/test ###########
###################################################################
train_summary_dir = os.path.join(keywords['train_dir'], "summaries", "train")
train_summary_writer = tf.summary.FileWriter(train_summary_dir)
train_summary_writer.add_graph(keywords['sess'].graph)
test_summary_dir = os.path.join(keywords['train_dir'], "summaries", "test")
test_summary_writer = tf.summary.FileWriter(test_summary_dir)
test_summary_writer.add_graph(keywords['sess'].graph)
# If fie-tuning flag in 'True' the model will be restored.
if keywords['finetuning']:
keywords['saver'].restore(keywords['sess'], os.path.join(keywords['checkpoint_dir'], checkpoint_prefix))
print("Model restored for fine-tuning...")
###################################################################
########## Run the training and loop over the batches #############
###################################################################
for epoch in range(keywords['num_epochs']):
total_batch_training = int(keywords['data'].train.images.shape[0] / keywords['batch_size'])
# go through the batches
for batch_num in range(total_batch_training):
#################################################
########## Get the training batches #############
#################################################
start_idx = batch_num * keywords['batch_size']
end_idx = (batch_num + 1) * keywords['batch_size']
# Fit training using batch data
train_batch_data, train_batch_label = keywords['data'].train.images[start_idx:end_idx], keywords[
'data'].train.labels[
start_idx:end_idx]
########################################
########## Run the session #############
########################################
# Run optimization op (backprop) and Calculate batch loss and accuracy
# When the tensor tensors['global_step'] is evaluated, it will be incremented by one.
batch_loss, _, train_summaries, training_step = keywords['sess'].run(
[keywords['tensors']['cost'], keywords['tensors']['train_op'], keywords['tensors']['summary_train_op'],
keywords['tensors']['global_step']],
feed_dict={keywords['tensors']['image_place']: train_batch_data,
keywords['tensors']['label_place']: train_batch_label,
keywords['tensors']['dropout_param']: 0.5})
########################################
########## Write summaries #############
########################################
# Write the summaries
train_summary_writer.add_summary(train_summaries, global_step=training_step)
# # Write the specific summaries for training phase.
# train_summary_writer.add_summary(train_image_summary, global_step=training_step)
#################################################
########## Plot the progressive bar #############
#################################################
progress = float(batch_num + 1) / total_batch_training
progress_bar.print_progress(progress, epoch_num=epoch + 1, loss=batch_loss)
# ################################################################
# ############ Summaries per epoch of training ###################
# ################################################################
summary_epoch_train_op = keywords['tensors']['summary_epoch_train_op']
train_epoch_summaries = keywords['sess'].run(summary_epoch_train_op,
feed_dict={keywords['tensors']['image_place']: train_batch_data,
keywords['tensors']['label_place']: train_batch_label,
keywords['tensors']['dropout_param']: 1.0})
# Put the summaries to the train summary writer.
train_summary_writer.add_summary(train_epoch_summaries, global_step=training_step)
#####################################################
########## Evaluation on the test data #############
#####################################################
if keywords['online_test']:
# WARNING: In this evaluation the whole test data is fed. In case the test data is huge this implementation
# may lead to memory error. In presense of large testing samples, batch evaluation on testing is
# recommended as in the training phase.
test_accuracy_epoch, test_summaries = keywords['sess'].run(
[keywords['tensors']['accuracy'], keywords['tensors']['summary_test_op']],
feed_dict={keywords['tensors']['image_place']: keywords['data'].test.images,
keywords['tensors'][
'label_place']: keywords['data'].test.labels,
keywords['tensors'][
'dropout_param']: 1.})
print("Epoch " + str(epoch + 1) + ", Testing Accuracy= " + \
"{:.5f}".format(test_accuracy_epoch))
###########################################################
########## Write the summaries for test phase #############
###########################################################
# Returning the value of global_step if necessary
current_step = tf.train.global_step(keywords['sess'], keywords['tensors']['global_step'])
# Add the couter of global step for proper scaling between train and test summuries.
test_summary_writer.add_summary(test_summaries, global_step=current_step)
###########################################################
############ Saving the model checkpoint ##################
###########################################################
# # The model will be saved when the training is done.
# Create the path for saving the checkpoints.
if not os.path.exists(keywords['checkpoint_dir']):
os.makedirs(keywords['checkpoint_dir'])
# save the model
save_path = keywords['saver'].save(keywords['sess'], os.path.join(keywords['checkpoint_dir'], checkpoint_prefix))
print("Model saved in file: %s" % save_path)
def train(**keywords):
"""
This function run the session whether in training or evaluation mode.
NOTE: **keywords is defined in order to make the code easily changable.
WARNING: All the arguments for the **keywords must be defined when calling this function.
**keywords:
:param sess: The default session.
:param saver: The saver operator to save and load the model weights.
:param tensors: The tensors dictionary defined by the graph.
:param data: The data structure.
:param train_dir: The training dir which is a reference for saving the logs and model checkpoints.
:param finetuning: If fine tuning should be done or random initialization is needed.
:param num_epochs: Number of epochs for training.
:param online_test: If the testing is done while training.
:param checkpoint_dir: The directory of the checkpoints.
:param batch_size: The training batch size.
:return:
Run the session.
"""
# The prefix for checkpoint files
checkpoint_prefix = 'model'
###################################################################
########## Defining the summary writers for train/test ###########
###################################################################
train_summary_dir = os.path.join(keywords['train_dir'], "summaries",
"train")
train_summary_writer = tf.summary.FileWriter(train_summary_dir)
train_summary_writer.add_graph(keywords['sess'].graph)
test_summary_dir = os.path.join(keywords['train_dir'], "summaries", "test")
test_summary_writer = tf.summary.FileWriter(test_summary_dir)
test_summary_writer.add_graph(keywords['sess'].graph)
# If fie-tuning flag in 'True' the model will be restored.
if keywords['finetuning']:
keywords['saver'].restore(
keywords['sess'],
os.path.join(keywords['checkpoint_dir'], checkpoint_prefix))
print("Model restored for fine-tuning...")
###################################################################
########## Run the training and loop over the batches #############
###################################################################
for epoch in range(keywords['num_epochs']):
total_batch_training = int(keywords['data'].train.images.shape[0] /
keywords['batch_size'])
# go through the batches
for batch_num in range(total_batch_training):
#################################################
########## Get the training batches #############
#################################################
start_idx = batch_num * keywords['batch_size']
end_idx = (batch_num + 1) * keywords['batch_size']
# Fit training using batch data
train_batch_data, train_batch_label = keywords[
'data'].train.images[start_idx:end_idx], keywords[
'data'].train.labels[start_idx:end_idx]
########################################
########## Run the session #############
########################################
# Run optimization op (backprop) and Calculate batch loss and accuracy
# When the tensor tensors['global_step'] is evaluated, it will be incremented by one.
batch_loss, _, train_summaries, training_step = keywords[
'sess'].run(
[
keywords['tensors']['cost'],
keywords['tensors']['train_op'],
keywords['tensors']['summary_train_op'],
keywords['tensors']['global_step']
],
feed_dict={
keywords['tensors']['image_place']: train_batch_data,
keywords['tensors']['label_place']: train_batch_label,
keywords['tensors']['dropout_param']: 0.5
})
########################################
########## Write summaries #############
########################################
# Write the summaries
train_summary_writer.add_summary(train_summaries,
global_step=training_step)
# # Write the specific summaries for training phase.
# train_summary_writer.add_summary(train_image_summary, global_step=training_step)
#################################################
########## Plot the progressive bar #############
#################################################
progress = float(batch_num + 1) / total_batch_training
progress_bar.print_progress(progress,
epoch_num=epoch + 1,
loss=batch_loss)
# ################################################################
# ############ Summaries per epoch of training ###################
# ################################################################
summary_epoch_train_op = keywords['tensors']['summary_epoch_train_op']
train_epoch_summaries = keywords['sess'].run(
summary_epoch_train_op,
feed_dict={
keywords['tensors']['image_place']: train_batch_data,
keywords['tensors']['label_place']: train_batch_label,
keywords['tensors']['dropout_param']: 1.0
})
# Put the summaries to the train summary writer.
train_summary_writer.add_summary(train_epoch_summaries,
global_step=training_step)
#####################################################
########## Evaluation on the test data #############
#####################################################
if keywords['online_test']:
# WARNING: In this evaluation the whole test data is fed. In case the test data is huge this implementation
# may lead to memory error. In presense of large testing samples, batch evaluation on testing is
# recommended as in the training phase.
test_accuracy_epoch, test_summaries = keywords['sess'].run(
[
keywords['tensors']['accuracy'],
keywords['tensors']['summary_test_op']
],
feed_dict={
keywords['tensors']['image_place']:
keywords['data'].test.images,
keywords['tensors']['label_place']:
keywords['data'].test.labels,
keywords['tensors']['dropout_param']: 1.
})
print("Epoch " + str(epoch + 1) + ", Testing Accuracy= " + \
"{:.5f}".format(test_accuracy_epoch))
###########################################################
########## Write the summaries for test phase #############
###########################################################
# Returning the value of global_step if necessary
current_step = tf.train.global_step(
keywords['sess'], keywords['tensors']['global_step'])
# Add the couter of global step for proper scaling between train and test summuries.
test_summary_writer.add_summary(test_summaries,
global_step=current_step)
###########################################################
############ Saving the model checkpoint ##################
###########################################################
# # The model will be saved when the training is done.
# Create the path for saving the checkpoints.
if not os.path.exists(keywords['checkpoint_dir']):
os.makedirs(keywords['checkpoint_dir'])
# save the model
save_path = keywords['saver'].save(
keywords['sess'],
os.path.join(keywords['checkpoint_dir'], checkpoint_prefix))
print("Model saved in file: %s" % save_path)
[tensors['accuracy_test'], tensors['loss_test'], tensors['summary_test_op'],
tensors['global_step']],
feed_dict={tensors['image_place']: test_batch_data,
tensors['label_place']: test_batch_label})
test_accuracy += test_batch_accuracy
########################################
########## Write summaries #############
########################################
# Write the summaries
test_summary_writer.add_summary(test_summaries, global_step=test_step)
# # Write the specific summaries for training phase.
# train_summary_writer.add_summary(train_image_summary, global_step=training_step)
#################################################
########## Plot the progressive bar #############
#################################################
progress = float(batch_num + 1) / total_batch_test
progress_bar.print_progress(progress, epoch_num=1, loss=batch_loss)
######################################################################
########## Calculate the accuracy for the whole test set #############
######################################################################
test_accuracy_total = test_accuracy / float(total_batch_test)
print("Testing Accuracy= " + \
"{:.5f}".format(test_accuracy_total))