args = parser.parse_args()
verbose = args.verbose
threads = args.threads
seed = args.seed
if seed is None:
seed = np.random.randint(0, (2**32) - 1)
logger.info("Using seed {}".format(seed))
np.random.seed(seed)
rand.seed(seed)
logger.info("Creating MDPs")
logger.info("----------")
envs = [{
'env': environments.get_large_rewarding_frozen_lake_environment(),
'name': 'large_frozen_lake',
'readable_name': 'Frozen Lake (18x20)',
}, {
'env':
environments.get_large_rewarding_frozen_lake_slipper_environment(),
'name':
'large_frozen_lake',
'readable_name':
'Frozen Lake (18x20)',
}, {
'env': environments.get_windy_cliff_walking_environment(),
'name': 'cliff_walking_windy',
'readable_name': 'Cliff Walking (4x12)',
}, {
'env': environments.get_cliff_walking_environment(),
logger.info("----------")
envs = [{
'env':
environments.get_rewarding_frozen_lake_environment(
ENV_REWARDS['small_lake']['step_rew'],
ENV_REWARDS['small_lake']['hole_rew'],
ENV_REWARDS['small_lake']['goal_rew']),
'name':
'frozen_lake',
'readable_name':
'Frozen Lake (8x8)',
}, {
'env':
environments.get_large_rewarding_frozen_lake_environment(
ENV_REWARDS['large_lake']['step_rew'],
ENV_REWARDS['large_lake']['hole_rew'],
ENV_REWARDS['large_lake']['goal_rew']),
'name':
'large_frozen_lake',
'readable_name':
'Frozen Lake (15x15)',
}, {
'env':
environments.get_windy_cliff_walking_environment(
ENV_REWARDS['cliff_walking']['step_rew'],
ENV_REWARDS['cliff_walking']['fall_rew'],
ENV_REWARDS['cliff_walking']['goal_rew']),
'name':
'cliff_walking',
'readable_name':
'Cliff Walking (4x12)',