File size: 3,311 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 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 |
from easydict import EasyDict
collector_env_num = 8
evaluator_env_num = 8
pong_r2d2_config = dict(
exp_name='pong_r2d2_seed0',
env=dict(
collector_env_num=collector_env_num,
evaluator_env_num=evaluator_env_num,
n_evaluator_episode=evaluator_env_num,
stop_value=20,
env_id='PongNoFrameskip-v4',
#'ALE/Pong-v5' is available. But special setting is needed after gym make.
frame_stack=4,
),
policy=dict(
cuda=True,
priority=True,
priority_IS_weight=True,
model=dict(
obs_shape=[4, 84, 84],
action_shape=6,
encoder_hidden_size_list=[128, 128, 512],
),
discount_factor=0.997,
nstep=5,
burnin_step=20,
# (int) the whole sequence length to unroll the RNN network minus
# the timesteps of burnin part,
# i.e., <the whole sequence length> = <unroll_len> = <burnin_step> + <learn_unroll_len>
learn_unroll_len=80,
learn=dict(
# according to the R2D2 paper, actor parameter update interval is 400
# environment timesteps, and in per collect phase, we collect <n_sample> sequence
# samples, the length of each sequence sample is <burnin_step> + <learn_unroll_len>,
# e.g. if n_sample=32, <sequence length> is 100, thus 32*100/400=8,
# we will set update_per_collect=8 in most environments.
update_per_collect=8,
batch_size=64,
learning_rate=0.0005,
target_update_theta=0.001,
),
collect=dict(
# NOTE: It is important that set key traj_len_inf=True here,
# to make sure self._traj_len=INF in serial_sample_collector.py.
# In R2D2 policy, for each collect_env, we want to collect data of length self._traj_len=INF
# unless the episode enters the 'done' state.
# In each collect phase, we collect a total of <n_sample> sequence samples.
n_sample=32,
traj_len_inf=True,
env_num=collector_env_num,
),
eval=dict(env_num=evaluator_env_num, ),
other=dict(
eps=dict(
type='exp',
start=0.95,
end=0.05,
decay=1e5,
),
replay_buffer=dict(
replay_buffer_size=10000,
# (Float type) How much prioritization is used: 0 means no prioritization while 1 means full prioritization
alpha=0.6,
# (Float type) How much correction is used: 0 means no correction while 1 means full correction
beta=0.4,
)
),
),
)
pong_r2d2_config = EasyDict(pong_r2d2_config)
main_config = pong_r2d2_config
pong_r2d2_create_config = dict(
env=dict(
type='atari',
import_names=['dizoo.atari.envs.atari_env'],
),
env_manager=dict(type='subprocess'),
policy=dict(type='r2d2'),
)
pong_r2d2_create_config = EasyDict(pong_r2d2_create_config)
create_config = pong_r2d2_create_config
if __name__ == "__main__":
# or you can enter `ding -m serial -c pong_r2d2_config.py -s 0`
from ding.entry import serial_pipeline
serial_pipeline((main_config, create_config), seed=0)
|