SLM Lab

CircleCI Maintainability Test Coverage

Modular Deep Reinforcement Learning framework in PyTorch.

ddqn_beamrider ddqn_breakout ddqn_pong
BeamRider Breakout Pong
ddqn_qbert ddqn_seaquest ddqn_spaceinvaders
Qbert Seaquest SpaceInvaders
References
Installation How to install SLM Lab
Documentation Usage documentation
Benchmark Benchmark results
Gitter SLM Lab user chatroom

Features

Algorithms

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.

Algorithm\Benchmark Atari Roboschool
SARSA -
DQN, distributed-DQN :white_check_mark:
Double-DQN, Dueling-DQN, PER-DQN :white_check_mark:
REINFORCE -
A2C, A3C (N-step & GAE) :white_check_mark:
PPO, distributed-PPO :white_check_mark:
SIL (A2C, PPO)

Environments

SLM Lab integrates with multiple environment offerings:

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

results matching ""

    No results matching ""