should_keep_run = query_yes_no("Should the run '{}' be kept?".format(
os.path.basename(run.base_path)),
default="yes")
if not should_keep_run:
shutil.rmtree(run.base_path)
if __name__ == "__main__":
# get arguments from console
arguments = parse_arguments()
# initialize new run
run = Run(run_id=None)
run.open()
run.set_config_value(arguments.input_style_file, "files", "input_style")
run.set_config_value(arguments.input_map_file, "files", "input_map")
run.set_config_value(arguments.output_map_file, "files", "output_map")
run.set_config_value(arguments.output_content_file, "files",
"output_content")
run.set_config_value(arguments.content_weight, "training",
"content_weight")
run.set_config_value(arguments.content_layers, "training",
"content_layers")
run.set_config_value(arguments.style_layers, "training", "style_layers")
run.set_config_value(arguments.style_weight, "training", "style_weight")
run.set_config_value(arguments.map_channel_weight, "training",
"map_channel_weight")
run.set_config_value(arguments.num_phases, "training", "num_phases")
run.set_config_value(arguments.device, "training", "device")
run.set_config_value(arguments.save_interval, "output", "save_interval")
argument_list.add_step_log_interval_argument(
"Interval how often step information should be displayed [steps].",
default=10)
argument_list.add_tf_verbosity_argument("Tensorflow verbosity.",
default="info")
argument_list.add_tf_min_log_level_argument(
"Tensorflow minimum log level.", default=3)
arguments = argument_list.parse()
# get dataset
dataset = get_dataset(arguments.dataset)
# initialize new run
run = Run(run_id=None)
run.open()
run.set_config_value(arguments.model, "model", "name")
run.set_config_value(arguments.dataset, "model", "dataset")
run.set_config_value(arguments.dataset_split, "model", "dataset_split")
run.set_config_value(dataset.num_classes + 1, "model", "num_classes")
run.set_config_value(arguments.random_seed, "training", "random_seed")
run.set_config_value(arguments.op_random_seed, "training",
"op_random_seed")
run.set_config_value(arguments.num_parallel_calls, "training",
"num_parallel_calls")
run.set_config_value(arguments.prefetch_buffer_size, "training",
"prefetch_buffer_size")
run.set_config_value(arguments.shuffle_buffer_size, "training",
"shuffle_buffer_size")
run.set_config_value(arguments.batch_size, "training", "batch_size")
run.set_config_value(arguments.num_steps, "training", "num_steps")
run.set_config_value(arguments.learning_rate, "training", "learning_rate")