if not os.path.exists(exp_dir):
os.makedirs(exp_dir)
callbacks_list = []
# Model checkpoint
if CHECKPOINTS:
callbacks_list.append(callbacks.checkpoints(exp_dir))
# Early stopping
if EARLY_STOP:
callbacks_list.append(callbacks.early_stopping(patience=7))
# Tensorboard
if TENSORBOARD:
callbacks_list.append(callbacks.tensorboard(exp_dir))
# Save best model
# ----------------
if SAVE_BEST:
callbacks_list.append(callbacks.save_best(exp_dir))
model.fit(
x=train_dataset,
epochs=1000,
steps_per_epoch=len(train_flow),
validation_data=validation_dataset,
validation_steps=len(validation_flow),
callbacks=callbacks_list,
)
def main(_):
## hyperparams
hps = tf.contrib.training.HParams(img_height=FLAGS.img_height,
img_width=FLAGS.img_width,
img_channels=FLAGS.img_channels,
discrete_outputs=FLAGS.discrete_outputs,
batch_size=FLAGS.batch_size,
episode_len=FLAGS.episode_len,
memory_size=FLAGS.memory_size,
code_size=FLAGS.code_size,
opt_iters=FLAGS.opt_iters,
enc_blocks=FLAGS.enc_blocks,
dec_blocks=FLAGS.dec_blocks,
num_filters=FLAGS.num_filters,
trainable_memory=FLAGS.trainable_memory,
use_bn=FLAGS.use_bn,
use_ddi=FLAGS.use_ddi,
lr=FLAGS.lr,
epochs=FLAGS.epochs)
## dataset
ds_train, ds_val, ds_test = data.get_dataset(name=FLAGS.dataset, hps=hps)
## model and session
model = VBMC.VariationalBayesianMemory(hps)
sess = tf.Session()
## tensorboard
train_writer = tf.summary.FileWriter(FLAGS.summaries_dir + '/train',
sess.graph)
test_writer = tf.summary.FileWriter(FLAGS.summaries_dir + '/test')
## checkpointing
saver = tf.train.Saver()
## init op
init_op = tf.global_variables_initializer()
_ = sess.run(init_op)
## restoring
if FLAGS.load_checkpoint != '' and os.path.exists(FLAGS.load_checkpoint):
saver.restore(sess, FLAGS.load_checkpoint)
else:
print('load checkpoint "{}" does not exist.'.format(
FLAGS.load_checkpoint))
print('continue anyway? [y/N]')
yn = input('> ')
if yn.strip().lower() != 'y':
print('program exiting.')
return
## helper functions for the various modes supported by this application
mode_to_routine = {
'train': routines.train,
'eval': routines.evaluate,
'generate': routines.generate,
'copy': routines.copy
}
routine = mode_to_routine[FLAGS.mode]
## rather than pass around tons of arguments,
# just use callbacks to perform the required functionality
if FLAGS.mode == 'train':
checkpoint_dir = FLAGS.checkpoint_dir
callbacks = {
'tensorboard': calls.tensorboard(train_writer),
'checkpointing': calls.checkpointing(sess, saver, checkpoint_dir)
}
routines.train(ds_train,
ds_val,
sess,
model,
callbacks,
epochs=hps.epochs)
elif FLAGS.mode == 'eval':
routines.evaluate(ds_test, sess, model)
else:
output_dir = FLAGS.output_dir
callbacks = {
'save_png': calls.save_png(output_dir),
'save_gif': calls.save_gif(output_dir)
}
routine(ds_test, sess, model, callbacks)
def main(_):
## hyperparams
hps = tf.contrib.training.HParams(
batch_size=FLAGS.batch_size,
img_height=FLAGS.img_height,
img_width=FLAGS.img_width,
img_channels=FLAGS.img_channels,
num_timesteps=FLAGS.num_timesteps,
z_dim=FLAGS.z_dim,
encoder_hidden_dim=FLAGS.encoder_hidden_dim,
decoder_hidden_dim=FLAGS.decoder_hidden_dim,
read_dim=FLAGS.read_dim,
write_dim=FLAGS.write_dim,
init_scale=FLAGS.init_scale,
forget_bias=FLAGS.forget_bias,
lr=FLAGS.lr,
epochs=FLAGS.epochs)
## dataset
ds_train, ds_test = get_dataset(name=FLAGS.dataset, hps=hps)
## model and session
model = DRAW(hps)
sess = tf.Session()
## tensorboard
train_writer = tf.summary.FileWriter(FLAGS.summaries_dir + '/train',
sess.graph)
test_writer = tf.summary.FileWriter(FLAGS.summaries_dir + '/test')
## checkpointing
saver = tf.train.Saver()
## init op
init_op = tf.global_variables_initializer()
_ = sess.run(init_op)
## restoring
if FLAGS.load_checkpoint != '':
saver.restore(sess, FLAGS.load_checkpoint)
## helper functions for the various modes supported by this application
mode_to_routine = {
'train': routines.train,
'eval': routines.evaluate,
'generate': routines.generate,
'reconstruct': routines.reconstruct,
'generate_gif': routines.generate_gif,
'reconstruct_gif': routines.reconstruct_gif
}
routine = mode_to_routine[FLAGS.mode]
## rather than pass around tons of arguments,
# just use callbacks to perform the required functionality
if FLAGS.mode == 'train':
checkpoint_dir = FLAGS.checkpoint_dir
checkpoint_frequency = FLAGS.checkpoint_frequency
callbacks = {
'tensorboard':
calls.tensorboard(train_writer),
'checkpointing':
calls.checkpointing(sess, saver, checkpoint_dir,
checkpoint_frequency)
}
routine(ds_train, sess, model, callbacks)
elif FLAGS.mode == 'eval':
callbacks = {}
routine(ds_test, sess, model, callbacks)
else:
output_dir = FLAGS.output_dir
callbacks = {
'save_png': calls.save_png(output_dir),
'save_gif': calls.save_gif(output_dir)
}
routine(ds_test, sess, model, callbacks)
