Modular Deep Reinforcement Learning framework in PyTorch.
|Installation||How to install SLM Lab|
|Gitter||SLM Lab user chatroom|
SLM Lab implements a number of canonical RL algorithms with reusable modular components and class-inheritance, with commitment to code quality and performance.
The benchmark results also include complete spec files to enable full reproducibility using SLM Lab.
Below shows the latest benchmark status. See benchmark results here.
|Double-DQN, Dueling-DQN, PER-DQN||:white_check_mark:|
|A2C, A3C (N-step & GAE)||:white_check_mark:|
|SIL (A2C, PPO)|
SLM Lab integrates with multiple environment offerings:
- OpenAI gym
- OpenAI Roboschool
- VizDoom (credit: joelouismarino)
- Unity environments with prebuilt binaries
Contributions are welcome to integrate more environments!
Metrics and Experimentation
To facilitate better RL development, SLM Lab also comes with prebuilt metrics and experimentation framework:
- every run generates metrics, graphs and data for analysis, as well as spec for reproducibility
- scalable hyperparameter search using Ray tune