return optimizer_tasks, optimizer_results, optimizer_agent, starting_weights
def make_rollout_agent(ob_size, action_size, starting_weights, args):
rollout_agent = rollout_master_agent.parallel_rollout_master_agent(
args, ob_size, action_size
)
rollout_agent.set_policy_weights(starting_weights)
return rollout_agent
if __name__ == '__main__':
# get the configuration
logger.info('New environments available : {}'.format(
register.get_name_list()))
args = get_config()
# args.use_nervenet = 0
if args.write_log:
logger.set_file_handler(
path=args.output_dir,
prefix='mujoco_' + args.task, time_str=args.time_id
)
if args.task in dm_control_util.DM_ENV_INFO:
args.dm = 1
# optional visdom plotting
if args.viz:
viz_item = ['avg_reward', 'entropy', 'kl', 'surr_loss',
import gym
import environments.register as register
for env_name in register.get_name_list():
print(env_name)
env = gym.make(env_name)
env.reset()
for _ in range(10):
env.render()
action = env.action_space.sample()
state, reward, done, info = env.step(action)
# print(state.shape, action.shape, reward, done, info)
"""
ALL_ENV_LIST = [
WalkersHopperone-v1
WalkersHalfhumanoidone-v1
WalkersHalfcheetahone-v1
WalkersFullcheetahone-v1
WalkersOstrichone-v1
WalkersHoppertwo-v1
WalkersHalfhumanoidtwo-v1
WalkersHalfcheetahtwo-v1
WalkersFullcheetahtwo-v1
WalkersOstrichtwo-v1
WalkersHopperthree-v1
WalkersHalfhumanoidthree-v1
WalkersHalfcheetahthree-v1
WalkersFullcheetahthree-v1
WalkersOstrichthree-v1
WalkersHopperfour-v1