daihao42 / coknight

A distributed Actor-Learner framework based on torchbeast

A distributed Actor-Learner framework based on torchbeast

combined with GRPC, for Atari Games training

Docker Env: pytorch/pytorch:1.7.0-cuda11.0-cudnn8-runtime

TODO

• Multiple servers support

• Hyperparameter tuning

Paper citation

If you used this code for your experiments or found it helpful, consider citing the following paper:

@article{DAI2022108,
  title = {Towards scalable and efficient Deep-RL in edge computing: A game-based partition approach},
  journal = {Journal of Parallel and Distributed Computing},
  volume = {168},
  pages = {108-119},
  year = {2022},
  issn = {0743-7315},
  doi = {https://doi.org/10.1016/j.jpdc.2022.06.006},
  author = {Hao Dai and Jiashu Wu and Yang Wang and Chengzhong Xu}
}