Open-source codebase for EfficientZero, from "Mastering Atari Games with Limited Data" at NeurIPS 2021.
EfficientZero requires python3 (>=3.6) and pytorch (>=1.8.0) with the development headers.
We recommend to use torch amp (--amp_type torch_amp) to accelerate training.
Before starting training, you need to build the c++/cython style external packages. (GCC version 7.5+ is required.)
cd core/ctree
bash make.sh
The distributed framework of this codebase is built on ray.
As for other packages required for this codebase, please run pip install -r requirements.txt.
- Train:
python main.py --env BreakoutNoFrameskip-v4 --case atari --opr train --amp_type torch_amp --num_gpus 1 --num_cpus 10 --cpu_actor 1 --gpu_actor 1 --force - Test:
python main.py --env BreakoutNoFrameskip-v4 --case atari --opr test --amp_type torch_amp --num_gpus 1 --load_model --model_path model.p \
We provide train.sh and test.sh for training and evaluation.
- Train:
- With 4 GPUs (3090):
bash train.sh
- With 4 GPUs (3090):
- Test:
bash test.sh
| Required Arguments | Description |
|---|---|
--env |
Name of the environment |
--case {atari} |
It's used for switching between different domains(default: atari) |
--opr {train,test} |
select the operation to be performed |
--amp_type {torch_amp,none} |
use torch amp for acceleration |
| Other Arguments | Description |
|---|---|
--force |
will rewrite the result directory |
--num_gpus 4 |
how many GPUs are available |
--num_cpus 96 |
how many CPUs are available |
--cpu_actor 14 |
how many cpu workers |
--gpu_actor 20 |
how many gpu workers |
--seed 0 |
the seed |
--use_priority |
use priority in replay buffer sampling |
--use_max_priority |
use the max priority for the newly collectted data |
--amp_type 'torch_amp' |
use torch amp for acceleration |
--info 'EZ-V0' |
some tags for you experiments |
--p_mcts_num 8 |
set the parallel number of envs in self-play |
--revisit_policy_search_rate 0.99 |
set the rate of reanalyzing policies |
--use_root_value |
use root values in value targets (require more GPU actors) |
--render |
render in evaluation |
--save_video |
save videos for evaluation |
The architecture of the training pipeline is shown as follows:
- To use a smaller model, you can choose smaller dim of the projection layers (Eg: 256/64) and the LSTM hidden layer (Eg: 64) in the config.
- For GPUs with 10G memory instead of 20G memory, you can allocate 0.25 gpu for each GPU maker (
@ray.remote(num_gpus=0.25)) incore/reanalyze_worker.py.
If you wan to apply EfficientZero to a new environment like mujoco. Here are the steps for registration:
- Follow the directory
config/atariand create dir for the env atconfig/mujoco. - Implement your
MujocoConfig(BaseConfig)class and implement the models as well as your environment wrapper. - Register the case at
main.py.
Evaluation with 32 seeds for 3 different runs (different seeds).
If you find this repo useful, please cite our paper:
@inproceedings{ye2021mastering,
title={Mastering Atari Games with Limited Data},
author={Weirui Ye, and Shaohuai Liu, and Thanard Kurutach, and Pieter Abbeel, and Yang Gao},
booktitle={NeurIPS},
year={2021}
}
If you have any question or want to use the code, please contact ywr20@mails.tsinghua.edu.cn .
We appreciate the following github repos a lot for their valuable code base implementations:
https://github.com/koulanurag/muzero-pytorch