batch_size=args.batch_size,
charset=charset_base,
max_text_length=max_text_length,
predict=args.test)
model = HTRModel(architecture=args.arch,
input_size=input_size,
vocab_size=dtgen.tokenizer.vocab_size)
model.compile(learning_rate=0.001)
model.load_checkpoint(target=target_path)
if args.train:
model.summary(output_path, "summary.txt")
callbacks = model.get_callbacks(logdir=output_path,
checkpoint=target_path,
verbose=1)
start_time = time.time()
h = model.fit(x=dtgen.next_train_batch(),
epochs=args.epochs,
steps_per_epoch=dtgen.steps['train'],
validation_data=dtgen.next_valid_batch(),
validation_steps=dtgen.steps['valid'],
callbacks=callbacks,
shuffle=True,
verbose=1)
total_time = time.time() - start_time
loss = h.history['loss']
val_loss = h.history['val_loss']
network_func = getattr(architecture, args.arch)
ioo = network_func(input_size=input_size,
output_size=(dtgen.tokenizer.vocab_size + 1),
learning_rate=0.001)
model = HTRModel(inputs=ioo[0], outputs=ioo[1])
model.compile(optimizer=ioo[2])
checkpoint = "checkpoint_weights.hdf5"
model.load_checkpoint(target=os.path.join(output_path, checkpoint))
if args.train:
model.summary(output_path, "summary.txt")
callbacks = model.get_callbacks(logdir=output_path,
hdf5=checkpoint,
verbose=1)
start_time = time.time()
h = model.fit_generator(generator=dtgen.next_train_batch(),
epochs=args.epochs,
steps_per_epoch=dtgen.train_steps,
validation_data=dtgen.next_valid_batch(),
validation_steps=dtgen.valid_steps,
callbacks=callbacks,
shuffle=True,
verbose=1)
total_time = time.time() - start_time
loss = h.history['loss']
val_loss = h.history['val_loss']