本文整理汇总了Python中baselines.common.cmd_util.mujoco_arg_parser方法的典型用法代码示例。如果您正苦于以下问题：Python cmd_util.mujoco_arg_parser方法的具体用法？Python cmd_util.mujoco_arg_parser怎么用？Python cmd_util.mujoco_arg_parser使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类baselines.common.cmd_util的用法示例。
在下文中一共展示了cmd_util.mujoco_arg_parser方法的7个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: main
# 需要导入模块: from baselines.common import cmd_util [as 别名]
# 或者: from baselines.common.cmd_util import mujoco_arg_parser [as 别名]
def main():
logger.configure()
parser = mujoco_arg_parser()
parser.add_argument('--model-path', default=os.path.join(logger.get_dir(), 'humanoid_policy'))
parser.set_defaults(num_timesteps=int(2e7))
args = parser.parse_args()
if not args.play:
# train the model
train(num_timesteps=args.num_timesteps, seed=args.seed, model_path=args.model_path)
else:
# construct the model object, load pre-trained model and render
pi = train(num_timesteps=1, seed=args.seed)
U.load_state(args.model_path)
env = make_mujoco_env('Humanoid-v2', seed=0)
ob = env.reset()
while True:
action = pi.act(stochastic=False, ob=ob)[0]
ob, _, done, _ = env.step(action)
env.render()
if done:
ob = env.reset() 示例2: main
# 需要导入模块: from baselines.common import cmd_util [as 别名]
# 或者: from baselines.common.cmd_util import mujoco_arg_parser [as 别名]
def main():
logger.configure()
parser = mujoco_arg_parser()
parser.add_argument('--model-path', default=os.path.join(logger.get_dir(), 'humanoid_policy'))
parser.set_defaults(num_timesteps=int(2e7))
args = parser.parse_args()
if not args.play:
# train the model
train(num_timesteps=args.num_timesteps, seed=args.seed, model_path=args.model_path)
else:
# construct the model object, load pre-trained model and render
pi = train(num_timesteps=1, seed=args.seed)
U.load_state(args.model_path)
env = make_mujoco_env('Humanoid-v2', seed=0)
ob = env.reset()
while True:
action = pi.act(stochastic=False, ob=ob)[0]
ob, _, done, _ = env.step(action)
env.render()
if done:
ob = env.reset() 示例3: main
# 需要导入模块: from baselines.common import cmd_util [as 别名]
# 或者: from baselines.common.cmd_util import mujoco_arg_parser [as 别名]
def main():
parser = mujoco_arg_parser()
parser.add_argument('--lr', type=float, default=3e-4, help="Learning rate")
parser.add_argument('--sil-update', type=float, default=10, help="Number of updates per iteration")
parser.add_argument('--sil-value', type=float, default=0.01, help="Weight for value update")
parser.add_argument('--sil-alpha', type=float, default=0.6, help="Alpha for prioritized replay")
parser.add_argument('--sil-beta', type=float, default=0.1, help="Beta for prioritized replay")
args = parser.parse_args()
logger.configure()
model, env = train(args.env, num_timesteps=args.num_timesteps, seed=args.seed,
lr=args.lr,
sil_update=args.sil_update, sil_value=args.sil_value,
sil_alpha=args.sil_alpha, sil_beta=args.sil_beta)
if args.play:
logger.log("Running trained model")
obs = np.zeros((env.num_envs,) + env.observation_space.shape)
obs[:] = env.reset()
while True:
actions = model.step(obs)[0]
obs[:] = env.step(actions)[0]
env.render() 示例4: main
# 需要导入模块: from baselines.common import cmd_util [as 别名]
# 或者: from baselines.common.cmd_util import mujoco_arg_parser [as 别名]
def main():
logger.configure()
parser = mujoco_arg_parser()
parser.add_argument('--model-path', default=os.path.join(logger.get_dir(), 'humanoid_policy'))
parser.set_defaults(num_timesteps=int(5e7))
args = parser.parse_args()
if not args.play:
# train the model
train(num_timesteps=args.num_timesteps, seed=args.seed, model_path=args.model_path)
else:
# construct the model object, load pre-trained model and render
pi = train(num_timesteps=1, seed=args.seed)
U.load_state(args.model_path)
env = make_mujoco_env('Humanoid-v2', seed=0)
ob = env.reset()
while True:
action = pi.act(stochastic=False, ob=ob)[0]
ob, _, done, _ = env.step(action)
env.render()
if done:
ob = env.reset() 示例5: main
# 需要导入模块: from baselines.common import cmd_util [as 别名]
# 或者: from baselines.common.cmd_util import mujoco_arg_parser [as 别名]
def main():
args = mujoco_arg_parser().parse_args()
logger.configure()
train(args.env, num_timesteps=args.num_timesteps, seed=args.seed) 示例6: main
# 需要导入模块: from baselines.common import cmd_util [as 别名]
# 或者: from baselines.common.cmd_util import mujoco_arg_parser [as 别名]
def main():
args = mujoco_arg_parser().parse_args()
train(args.env, num_timesteps=args.num_timesteps, seed=args.seed) 示例7: main
# 需要导入模块: from baselines.common import cmd_util [as 别名]
# 或者: from baselines.common.cmd_util import mujoco_arg_parser [as 别名]
def main():
args = mujoco_arg_parser().parse_args()
logger.configure()
model, env = train(args.env, num_timesteps=args.num_timesteps, seed=args.seed)
if args.play:
logger.log("Running trained model")
obs = np.zeros((env.num_envs,) + env.observation_space.shape)
obs[:] = env.reset()
while True:
actions = model.step(obs)[0]
obs[:] = env.step(actions)[0]
env.render()