facebookresearch/ReAgent
Facebook/Meta's open-source platform for training and deploying reinforcement learning agents in production environments.
Not currently ranked — collecting fresh signals.
star history
ReAgent is an end-to-end reinforcement learning platform built with Python and PyTorch. It provides workflows for training popular deep RL algorithms including DQN, Double DQN, Dueling DQN, C51, and QR-DQN variants. The platform handles the full ML lifecycle from data preprocessing and feature transformation through distributed training and counterfactual policy evaluation, to optimized TorchScript model serving.
Frequently asked
- What is facebookresearch/ReAgent?
- Facebook/Meta's open-source platform for training and deploying reinforcement learning agents in production environments.
- Is ReAgent open source?
- Yes — facebookresearch/ReAgent is open source, released under the BSD-3-Clause license.
- What language is ReAgent written in?
- facebookresearch/ReAgent is primarily written in Python.
- How popular is ReAgent?
- facebookresearch/ReAgent has 3.7k stars on GitHub.
- Where can I find ReAgent?
- facebookresearch/ReAgent is on GitHub at https://github.com/facebookresearch/ReAgent.