def main(args):
"""Run training and validation.
1. Build graphs
1.1 Training graph to run on multiple GPUs
1.2 Validation graph to run on multiple GPUs
2. Configure sessions
2.1 Train
2.2 Validate
3. Main loop
3.1 Train
3.2 Write summary
3.3 Save model
3.4 Validate model
Author:
Ashley Gritzman
"""
# Set reproduciable random seed
tf.set_random_seed(1234)
# Directories
train_dir, train_summary_dir = conf.setup_train_directories()
# Logger
conf.setup_logger(logger_dir=train_dir, name="logger_train.txt")
# Hyperparameters
conf.load_or_save_hyperparams(train_dir)
# Get dataset hyperparameters
logger.info('Using dataset: {}'.format(FLAGS.dataset))
dataset_size_train = conf.get_dataset_size_train(FLAGS.dataset)
dataset_size_val = conf.get_dataset_size_validate(FLAGS.dataset)
build_arch = conf.get_dataset_architecture(FLAGS.dataset)
num_classes = conf.get_num_classes(FLAGS.dataset)
create_inputs_train = conf.get_create_inputs(FLAGS.dataset, mode="train")
create_inputs_val = conf.get_create_inputs(FLAGS.dataset, mode="validate")
#*****************************************************************************
# 1. BUILD GRAPHS
#*****************************************************************************
#----------------------------------------------------------------------------
# GRAPH - TRAIN
#----------------------------------------------------------------------------
logger.info('BUILD TRAIN GRAPH')
g_train = tf.Graph()
with g_train.as_default(), tf.device('/cpu:0'):
# Get global_step
global_step = tf.train.get_or_create_global_step()
# Get batches per epoch
num_batches_per_epoch = int(dataset_size_train / FLAGS.batch_size)
# In response to a question on OpenReview, Hinton et al. wrote the
# following:
# "We use an exponential decay with learning rate: 3e-3, decay_steps: 20000, # decay rate: 0.96."
# https://openreview.net/forum?id=HJWLfGWRb¬eId=ryxTPFDe2X
lrn_rate = tf.train.exponential_decay(learning_rate=FLAGS.lrn_rate,
global_step=global_step,
decay_steps=20000,
decay_rate=0.96)
tf.summary.scalar('learning_rate', lrn_rate)
opt = tf.train.AdamOptimizer(learning_rate=lrn_rate)
# Get batch from data queue. Batch size is FLAGS.batch_size, which is then
# divided across multiple GPUs
input_dict = create_inputs_train()
batch_x = input_dict['image']
batch_labels = input_dict['label']
# AG 03/10/2018: Split batch for multi gpu implementation
# Each split is of size FLAGS.batch_size / FLAGS.num_gpus
# See: https://github.com/naturomics/CapsNet-Tensorflow/blob/master/
# dist_version/distributed_train.py
splits_x = tf.split(axis=0,
num_or_size_splits=FLAGS.num_gpus,
value=batch_x)
splits_labels = tf.split(axis=0,
num_or_size_splits=FLAGS.num_gpus,
value=batch_labels)
#--------------------------------------------------------------------------
# MULTI GPU - TRAIN
#--------------------------------------------------------------------------
# Calculate the gradients for each model tower
tower_grads = []
tower_losses = []
tower_logits = []
reuse_variables = None
for i in range(FLAGS.num_gpus):
with tf.device('/gpu:%d' % i):
with tf.name_scope('tower_%d' % i) as scope:
logger.info('TOWER %d' % i)
#with slim.arg_scope([slim.model_variable, slim.variable],
# device='/cpu:0'):
with slim.arg_scope([slim.variable], device='/cpu:0'):
loss, logits = tower_fn(
build_arch,
splits_x[i],
splits_labels[i],
scope,
num_classes,
reuse_variables=reuse_variables,
is_train=True)
# Don't reuse variable for first GPU, but do reuse for others
reuse_variables = True
# Compute gradients for one GPU
grads = opt.compute_gradients(loss)
# Keep track of the gradients across all towers.
tower_grads.append(grads)
# Keep track of losses and logits across for each tower
tower_logits.append(logits)
tower_losses.append(loss)
# Loss for each tower
tf.summary.scalar("loss", loss)
# We must calculate the mean of each gradient. Note that this is the
# synchronization point across all towers.
grad = average_gradients(tower_grads)
# See: https://stackoverflow.com/questions/40701712/how-to-check-nan-in-
# gradients-in-tensorflow-when-updating
grad_check = ([
tf.check_numerics(g, message='Gradient NaN Found!')
for g, _ in grad if g is not None
] + [tf.check_numerics(loss, message='Loss NaN Found')])
# Apply the gradients to adjust the shared variables
with tf.control_dependencies(grad_check):
update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS)
with tf.control_dependencies(update_ops):
train_op = opt.apply_gradients(grad, global_step=global_step)
# Calculate mean loss
loss = tf.reduce_mean(tower_losses)
# Calculate accuracy
logits = tf.concat(tower_logits, axis=0)
acc = met.accuracy(logits, batch_labels)
# Prepare predictions and one-hot labels
probs = tf.nn.softmax(logits=logits)
labels_oh = tf.one_hot(batch_labels, num_classes)
# Group metrics together
# See: https://cs230-stanford.github.io/tensorflow-model.html
trn_metrics = {
'loss': loss,
'labels': batch_labels,
'labels_oh': labels_oh,
'logits': logits,
'probs': probs,
'acc': acc,
}
# Reset and read operations for streaming metrics go here
trn_reset = {}
trn_read = {}
# Logging
tf.summary.scalar('trn_loss', loss)
tf.summary.scalar('trn_acc', acc)
# Set Saver
# AG 26/09/2018: Save all variables including Adam so that we can continue
# training from where we left off
# max_to_keep=None should keep all checkpoints
saver = tf.train.Saver(tf.global_variables(), max_to_keep=None)
# Display number of parameters
train_params = np.sum([
np.prod(v.get_shape().as_list()) for v in tf.trainable_variables()
]).astype(np.int32)
logger.info('Trainable Parameters: {}'.format(train_params))
# Set summary op
trn_summary = tf.summary.merge_all()
#----------------------------------------------------------------------------
# GRAPH - VALIDATION
#----------------------------------------------------------------------------
logger.info('BUILD VALIDATION GRAPH')
g_val = tf.Graph()
with g_val.as_default():
# Get global_step
global_step = tf.train.get_or_create_global_step()
num_batches_val = int(dataset_size_val / FLAGS.batch_size *
FLAGS.val_prop)
# Get data
input_dict = create_inputs_val()
batch_x = input_dict['image']
batch_labels = input_dict['label']
# AG 10/12/2018: Split batch for multi gpu implementation
# Each split is of size FLAGS.batch_size / FLAGS.num_gpus
# See: https://github.com/naturomics/CapsNet-
# Tensorflow/blob/master/dist_version/distributed_train.py
splits_x = tf.split(axis=0,
num_or_size_splits=FLAGS.num_gpus,
value=batch_x)
splits_labels = tf.split(axis=0,
num_or_size_splits=FLAGS.num_gpus,
value=batch_labels)
#--------------------------------------------------------------------------
# MULTI GPU - VALIDATE
#--------------------------------------------------------------------------
# Calculate the logits for each model tower
tower_logits = []
reuse_variables = None
for i in range(FLAGS.num_gpus):
with tf.device('/gpu:%d' % i):
with tf.name_scope('tower_%d' % i) as scope:
with slim.arg_scope([slim.variable], device='/cpu:0'):
loss, logits = tower_fn(
build_arch,
splits_x[i],
splits_labels[i],
scope,
num_classes,
reuse_variables=reuse_variables,
is_train=False)
# Don't reuse variable for first GPU, but do reuse for others
reuse_variables = True
# Keep track of losses and logits across for each tower
tower_logits.append(logits)
# Loss for each tower
tf.summary.histogram("val_logits", logits)
# Combine logits from all towers
logits = tf.concat(tower_logits, axis=0)
# Calculate metrics
val_loss = mod.spread_loss(logits, batch_labels)
val_acc = met.accuracy(logits, batch_labels)
# Prepare predictions and one-hot labels
val_probs = tf.nn.softmax(logits=logits)
val_labels_oh = tf.one_hot(batch_labels, num_classes)
# Group metrics together
# See: https://cs230-stanford.github.io/tensorflow-model.html
val_metrics = {
'loss': val_loss,
'labels': batch_labels,
'labels_oh': val_labels_oh,
'logits': logits,
'probs': val_probs,
'acc': val_acc,
}
# Reset and read operations for streaming metrics go here
val_reset = {}
val_read = {}
tf.summary.scalar("val_loss", val_loss)
tf.summary.scalar("val_acc", val_acc)
# Saver
saver = tf.train.Saver(max_to_keep=None)
# Set summary op
val_summary = tf.summary.merge_all()
#****************************************************************************
# 2. SESSIONS
#****************************************************************************
#----- SESSION TRAIN -----#
# Session settings
sess_train = tf.Session(config=tf.ConfigProto(allow_soft_placement=True,
log_device_placement=False),
graph=g_train)
# Debugger
# AG 05/06/2018: Debugging using either command line or TensorBoard
if FLAGS.debugger is not None:
# sess = tf_debug.LocalCLIDebugWrapperSession(sess)
sess_train = tf_debug.TensorBoardDebugWrapperSession(
sess_train, FLAGS.debugger)
with g_train.as_default():
sess_train.run([
tf.global_variables_initializer(),
tf.local_variables_initializer()
])
# Restore previous checkpoint
# AG 26/09/2018: where should this go???
if FLAGS.load_dir is not None:
prev_step = load_training(saver, sess_train, FLAGS.load_dir)
else:
prev_step = 0
# Create summary writer, and write the train graph
summary_writer = tf.summary.FileWriter(train_summary_dir,
graph=sess_train.graph)
#----- SESSION VALIDATION -----#
sess_val = tf.Session(config=tf.ConfigProto(allow_soft_placement=True,
log_device_placement=False),
graph=g_val)
with g_val.as_default():
sess_val.run([
tf.local_variables_initializer(),
tf.global_variables_initializer()
])
#****************************************************************************
# 3. MAIN LOOP
#****************************************************************************
SUMMARY_FREQ = 100
SAVE_MODEL_FREQ = num_batches_per_epoch # 500
VAL_FREQ = num_batches_per_epoch # 500
PROFILE_FREQ = 5
for step in range(prev_step, FLAGS.epoch * num_batches_per_epoch + 1):
#for step in range(0,3):
# AG 23/05/2018: limit number of iterations for testing
# for step in range(100):
epoch_decimal = step / num_batches_per_epoch
epoch = int(np.floor(epoch_decimal))
# TF queue would pop batch until no file
try:
# TRAIN
with g_train.as_default():
# With profiling
if (FLAGS.profile is True) and ((step % PROFILE_FREQ) == 0):
logger.info("Train with Profiling")
run_options = tf.RunOptions(
trace_level=tf.RunOptions.FULL_TRACE)
run_metadata = tf.RunMetadata()
# Without profiling
else:
run_options = None
run_metadata = None
# Reset streaming metrics
if step % (num_batches_per_epoch / 4) == 1:
logger.info("Reset streaming metrics")
sess_train.run([trn_reset])
# MAIN RUN
tic = time.time()
train_op_v, trn_metrics_v, trn_summary_v = sess_train.run(
[train_op, trn_metrics, trn_summary],
options=run_options,
run_metadata=run_metadata)
toc = time.time()
# Read streaming metrics
trn_read_v = sess_train.run(trn_read)
# Write summary for profiling
if run_options is not None:
summary_writer.add_run_metadata(run_metadata,
'step{:d}'.format(step))
# Logging
logger.info('TRN' + ' e-{:d}'.format(epoch) +
' stp-{:d}'.format(step) +
' {:.2f}s'.format(toc - tic) +
' loss: {:.4f}'.format(trn_metrics_v['loss']) +
' acc: {:.2f}%'.format(trn_metrics_v['acc'] * 100))
except KeyboardInterrupt:
sess_train.close()
sess_val.close()
sys.exit()
except tf.errors.InvalidArgumentError as e:
logger.warning('%d iteration contains NaN gradients. Discard.' %
step)
logger.error(str(e))
continue
else:
# WRITE SUMMARY
if (step % SUMMARY_FREQ) == 0:
logger.info("Write Train Summary")
with g_train.as_default():
# Summaries from graph
summary_writer.add_summary(trn_summary_v, step)
# SAVE MODEL
if (step % SAVE_MODEL_FREQ) == 100:
logger.info("Save Model")
with g_train.as_default():
train_checkpoint_dir = train_dir + '/checkpoint'
if not os.path.exists(train_checkpoint_dir):
os.makedirs(train_checkpoint_dir)
# Save ckpt from train session
ckpt_path = os.path.join(train_checkpoint_dir,
'model.ckpt')
saver.save(sess_train, ckpt_path, global_step=step)
# VALIDATE MODEL
if (step % VAL_FREQ) == 100:
#----- Validation -----#
with g_val.as_default():
logger.info("Start Validation")
# Restore ckpt to val session
latest_ckpt = tf.train.latest_checkpoint(
train_checkpoint_dir)
saver.restore(sess_val, latest_ckpt)
# Reset accumulators
accuracy_sum = 0
loss_sum = 0
sess_val.run(val_reset)
for i in range(num_batches_val):
val_metrics_v, val_summary_str_v = sess_val.run(
[val_metrics, val_summary])
# Update
accuracy_sum += val_metrics_v['acc']
loss_sum += val_metrics_v['loss']
# Read
val_read_v = sess_val.run(val_read)
# Get checkpoint number
ckpt_num = re.split('-', latest_ckpt)[-1]
# Logging
logger.info('VAL ckpt-{}'.format(ckpt_num) +
' bch-{:d}'.format(i) +
' cum_acc: {:.2f}%'.format(accuracy_sum /
(i + 1) * 100) +
' cum_loss: {:.4f}'.format(loss_sum /
(i + 1)))
# Average across batches
ave_acc = accuracy_sum / num_batches_val
ave_loss = loss_sum / num_batches_val
logger.info('VAL ckpt-{}'.format(ckpt_num) +
' avg_acc: {:.2f}%'.format(ave_acc * 100) +
' avg_loss: {:.4f}'.format(ave_loss))
logger.info("Write Val Summary")
summary_val = tf.Summary()
summary_val.value.add(tag="val_acc", simple_value=ave_acc)
summary_val.value.add(tag="val_loss",
simple_value=ave_loss)
summary_writer.add_summary(summary_val, step)
# Close (main loop)
sess_train.close()
sess_val.close()
sys.exit()
def main(args):
# Set reproduciable random seed
tf.set_random_seed(1234)
# Directories
# Get name
split = FLAGS.load_dir.split('/')
if split[-1]:
name = split[-1]
else:
name = split[-2]
# Get parent directory
split = FLAGS.load_dir.split("/" + name)
parent_dir = split[0]
test_dir = '{}/{}/test'.format(parent_dir, name)
test_summary_dir = test_dir + '/summary'
# Clear the test log directory
if (FLAGS.reset is True) and os.path.exists(test_dir):
shutil.rmtree(test_dir)
if not os.path.exists(test_summary_dir):
os.makedirs(test_summary_dir)
# Logger
conf.setup_logger(logger_dir=test_dir, name="logger_test.txt")
logger.info("name: " + name)
logger.info("parent_dir: " + parent_dir)
logger.info("test_dir: " + test_dir)
# Load hyperparameters from train run
conf.load_or_save_hyperparams()
# Get dataset hyperparameters
logger.info('Using dataset: {}'.format(FLAGS.dataset))
# Dataset
dataset_size_test = conf.get_dataset_size_test(FLAGS.dataset)
num_classes = conf.get_num_classes(FLAGS.dataset)
create_inputs_test = conf.get_create_inputs(FLAGS.dataset, mode="test")
#----------------------------------------------------------------------------
# GRAPH - TEST
#----------------------------------------------------------------------------
logger.info('BUILD TEST GRAPH')
g_test = tf.Graph()
with g_test.as_default():
# Get global_step
global_step = tf.train.get_or_create_global_step()
num_batches_test = int(dataset_size_test / FLAGS.batch_size)
# Get data
input_dict = create_inputs_test()
batch_x = input_dict['image']
batch_labels = input_dict['label']
# AG 10/12/2018: Split batch for multi gpu implementation
# Each split is of size FLAGS.batch_size / FLAGS.num_gpus
# See: https://github.com/naturomics/CapsNet-
# Tensorflow/blob/master/dist_version/distributed_train.py
splits_x = tf.split(
axis=0,
num_or_size_splits=FLAGS.num_gpus,
value=batch_x)
splits_labels = tf.split(
axis=0,
num_or_size_splits=FLAGS.num_gpus,
value=batch_labels)
# Build architecture
build_arch = conf.get_dataset_architecture(FLAGS.dataset)
# for baseline
#build_arch = conf.get_dataset_architecture('baseline')
#--------------------------------------------------------------------------
# MULTI GPU - TEST
#--------------------------------------------------------------------------
# Calculate the logits for each model tower
tower_logits = []
reuse_variables = None
for i in range(FLAGS.num_gpus):
with tf.device('/gpu:%d' % i):
with tf.name_scope('tower_%d' % i) as scope:
with slim.arg_scope([slim.variable], device='/cpu:0'):
loss, logits = tower_fn(
build_arch,
splits_x[i],
splits_labels[i],
scope,
num_classes,
reuse_variables=reuse_variables,
is_train=False)
# Don't reuse variable for first GPU, but do reuse for others
reuse_variables = True
# Keep track of losses and logits across for each tower
tower_logits.append(logits)
# Loss for each tower
tf.summary.histogram("test_logits", logits)
# Combine logits from all towers
logits = tf.concat(tower_logits, axis=0)
# Calculate metrics
test_loss = mod.spread_loss(logits, batch_labels)
test_acc = met.accuracy(logits, batch_labels)
# Prepare predictions and one-hot labels
test_probs = tf.nn.softmax(logits=logits)
test_labels_oh = tf.one_hot(batch_labels, num_classes)
# Group metrics together
# See: https://cs230-stanford.github.io/tensorflow-model.html
test_metrics = {'loss' : test_loss,
'labels' : batch_labels,
'labels_oh' : test_labels_oh,
'logits' : logits,
'probs' : test_probs,
'acc' : test_acc,
}
# Reset and read operations for streaming metrics go here
test_reset = {}
test_read = {}
tf.summary.scalar("test_loss", test_loss)
tf.summary.scalar("test_acc", test_acc)
# Saver
saver = tf.train.Saver(max_to_keep=None)
# Set summary op
test_summary = tf.summary.merge_all()
#--------------------------------------------------------------------------
# SESSION - TEST
#--------------------------------------------------------------------------
#sess_test = tf.Session(
# config=tf.ConfigProto(allow_soft_placement=True,
# log_device_placement=False),
# graph=g_test)
# Perry: added in for RTX 2070 incompatibility workaround
config = tf.ConfigProto(allow_soft_placement=True, log_device_placement=False)
config.gpu_options.allow_growth = True
sess_test = tf.Session(config=config, graph=g_test)
#sess_test.run(tf.local_variables_initializer())
#sess_test.run(tf.global_variables_initializer())
summary_writer = tf.summary.FileWriter(
test_summary_dir,
graph=sess_test.graph)
ckpts_to_test = []
load_dir_chechpoint = os.path.join(FLAGS.load_dir, "train", "checkpoint")
# Evaluate the latest ckpt in dir
if FLAGS.ckpt_name is None:
latest_ckpt = tf.train.latest_checkpoint(load_dir_chechpoint)
ckpts_to_test.append(latest_ckpt)
# Evaluate all ckpts in dir
elif FLAGS.ckpt_name == "all":
# Get list of files in firectory and sort by date created
filenames = os.listdir(load_dir_chechpoint)
regex = re.compile(r'.*.index')
filenames = filter(regex.search, filenames)
data_ckpts = (os.path.join(load_dir_chechpoint, fn) for fn in filenames)
data_ckpts = ((os.stat(path), path) for path in data_ckpts)
# regular files, insert creation date
data_ckpts = ((stat[ST_CTIME], path) for stat, path in data_ckpts
if S_ISREG(stat[ST_MODE]))
data_ckpts= sorted(data_ckpts)
# remove ".index"
ckpts_to_test = [path[:-6] for ctime, path in data_ckpts]
# Evaluate ckpt specified by name
else:
ckpt_name = os.path.join(load_dir_chechpoint, FLAGS.ckpt_name)
ckpts_to_test.append(ckpt_name)
#--------------------------------------------------------------------------
# MAIN LOOP
#--------------------------------------------------------------------------
# Run testing on checkpoints
for ckpt in ckpts_to_test:
saver.restore(sess_test, ckpt)
# Reset accumulators
accuracy_sum = 0
loss_sum = 0
sess_test.run(test_reset)
for i in range(num_batches_test):
test_metrics_v, test_summary_str_v = sess_test.run(
[test_metrics, test_summary])
# Update
accuracy_sum += test_metrics_v['acc']
loss_sum += test_metrics_v['loss']
ckpt_num = re.split('-', ckpt)[-1]
logger.info('TEST ckpt-{}'.format(ckpt_num)
+ ' bch-{:d}'.format(i)
+ ' cum_acc: {:.2f}%'.format(accuracy_sum/(i+1)*100)
+ ' cum_loss: {:.4f}'.format(loss_sum/(i+1))
)
ave_acc = accuracy_sum / num_batches_test
ave_loss = loss_sum / num_batches_test
logger.info('TEST ckpt-{}'.format(ckpt_num)
+ ' avg_acc: {:.2f}%'.format(ave_acc*100)
+ ' avg_loss: {:.4f}'.format(ave_loss))
logger.info("Write Test Summary")
summary_test = tf.Summary()
summary_test.value.add(tag="test_acc", simple_value=ave_acc)
summary_test.value.add(tag="test_loss", simple_value=ave_loss)
summary_writer.add_summary(summary_test, ckpt_num)