File size: 1,972 Bytes
079c32c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 |
from easydict import EasyDict
ant_td3_default_config = dict(
exp_name='multi_mujoco_ant_2x4_td3',
env=dict(
scenario='Ant-v2',
agent_conf="2x4d",
agent_obsk=2,
add_agent_id=False,
episode_limit=1000,
collector_env_num=8,
evaluator_env_num=8,
n_evaluator_episode=8,
stop_value=6000,
),
policy=dict(
cuda=True,
random_collect_size=0,
multi_agent=True,
model=dict(
agent_obs_shape=54,
global_obs_shape=111,
action_shape=4,
action_space='regression',
actor_head_hidden_size=256,
critic_head_hidden_size=256,
twin_critic=True,
),
learn=dict(
update_per_collect=10,
batch_size=256,
learning_rate_actor=1e-3,
learning_rate_critic=1e-3,
target_theta=0.005,
discount_factor=0.99,
actor_update_freq=2,
noise=True,
),
collect=dict(
n_sample=400,
noise_sigma=0.1,
),
eval=dict(evaluator=dict(eval_freq=500, )),
other=dict(replay_buffer=dict(replay_buffer_size=100000, ), ),
),
)
ant_td3_default_config = EasyDict(ant_td3_default_config)
main_config = ant_td3_default_config
ant_td3_default_create_config = dict(
env=dict(
type='mujoco_multi',
import_names=['dizoo.multiagent_mujoco.envs.multi_mujoco_env'],
),
env_manager=dict(type='subprocess'),
policy=dict(type='td3'),
replay_buffer=dict(type='naive', ),
)
ant_td3_default_create_config = EasyDict(ant_td3_default_create_config)
create_config = ant_td3_default_create_config
if __name__ == '__main__':
# or you can enter `ding -m serial -c ant_matd3_config.py -s 0`
from ding.entry.serial_entry import serial_pipeline
serial_pipeline((main_config, create_config), seed=0, max_env_step=int(1e7))
|