Force is a library for deep reinforcement learning (RL) research, built on PyTorch and OpenAI Gym. It is under active development. Basic features at present:
- Implementations of various deep RL algorithms (DDPG, TD3, SAC, FQE)
- Tools for managing configuration and launching jobs on the Slurm scheduler
The name Force was originally derived from the word reinforcement, but, by a fun coincidence, it is also related to one of my advisors via Newton's second law.
I am using Python 3.8 and PyTorch 1.4.0. I think any Python 3.6+ would be fine, although I haven't verified.