Source code for l2rpn_baselines.utils.cli_eval

# Copyright (c) 2020, RTE (https://www.rte-france.com)
# See AUTHORS.txt
# This Source Code Form is subject to the terms of the Mozilla Public License, version 2.0.
# If a copy of the Mozilla Public License, version 2.0 was not distributed with this file,
# you can obtain one at http://mozilla.org/MPL/2.0/.
# SPDX-License-Identifier: MPL-2.0
# This file is part of L2RPN Baselines, L2RPN Baselines a repository to host baselines for l2rpn competitions.

import argparse
from l2rpn_baselines.utils.str2bool import str2bool


# TODO add possibilitiy to add new CLI when calling the function
[docs]def cli_eval(): """ some usefull command line arguments (CLI) for the evaluation of the baseline.""" parser = argparse.ArgumentParser(description="Train baseline DDQN") parser.add_argument("--env_name", default="l2rpn_case14_sandbox", help="Name of the l2rpn dataset environment") parser.add_argument("--load_path", default=".", help="The path where the trained model has been stored") parser.add_argument("--name", default="baseline", help="The name of the trained baseline") parser.add_argument("--logs_dir", required=False, default="./logs-eval", type=str, help="Path to output logs directory") parser.add_argument("--nb_episode", required=False, default=1, type=int, help="Number of episodes to evaluate") parser.add_argument("--nb_process", required=False, default=1, type=int, help="Number of cores to use") parser.add_argument("--max_steps", required=False, default=8000, type=int, help="Maximum number of steps per scenario") parser.add_argument("--verbose", type=str2bool, nargs='?', const=True, default=False, help="Enable verbose runner mode..") parser.add_argument("--save_gif", type=str2bool, nargs='?', const=True, default=False, help="Save the gif as \"epidose.gif\" in the episode path module.") return parser