end='')
end = time.time()
print('\repochs: {:4d}, loss_batch(R=L1):{:.4f}'.format(
ep, np.mean(loss_it)),
end='')
if isAdv:
print(', loss_gen: {:.4f}, loss_dis: {:.4f}'.format(
np.mean(loss_gen), np.mean(loss_dis)),
end='')
print(' in {:3.2f}s'.format(end - start))
# Save model each {opts.save_epoch} epochs
if ep % opts.save_epoch == 0:
print('saving model at epoch {:4d}'.format(ep))
model.save_model(ep)
if __name__ == '__main__':
# Get parameters
opts = baseOpt().parse()
# Build model, and run test
model, isAdv = setModel(opts.model, opts)
# Loading a model
if opts.load_epoch > 0:
print('Loading model at epoch {:d}'.format(opts.load_epoch))
model.load_model(opts.load_epoch)
train_op(model, opts, isAdv)
#ambnt_batch = [ambnt_imgs[it]]
flash_batch = [flash_imgs[it]]
flash_file = file_list[it]
# Set inputs of the model and run
model.set_inputs(flash_batch, None)
# For the DeepFlash model
if opts.model == 'DeepFlash':
flash_bf_batch = [flash_bf_imgs[it]]
# Setting input and target images filtered
model.set_filtered_inputs(flash_bf_batch, None)
model.forward()
saveimg(results_path, flash_file, model.fake_Y, opts.out_act)
print('\riter:{:4d}/{:4d}'.format(it + 1, len(flash_imgs)), end='')
print('\rTesting [{:4d}/{:4d}]: check the results on "{}"'.format(
it + 1, len(flash_imgs), results_path))
if __name__ == "__main__":
# Get parameters
opts = baseOpt().parse()
# Build model, load, and run test
print('Testing {} model '.format(opts.model))
model, _ = setModel(opts, False)
model.load_model(opts.load_epoch)
test_op(model, opts)