ikostrikov/jaxrl
JAX (Flax) implementations of deep reinforcement learning algorithms including Soft Actor Critic, DDPG, and AWAC for continuous action spaces.
Not currently ranked — collecting fresh signals.
star history
This repository provides clean, minimal implementations of reinforcement learning algorithms written in JAX and Flax. It includes Soft Actor Critic with learnable temperature, Advantage Weighted Actor Critic, Deep Deterministic Policy Gradient with clipped double Q-learning, randomized ensemble double Q-learning, and behavioral cloning. The goal is to offer simple implementations that researchers can build upon for RL research in continuous control environments.
Frequently asked
- What is ikostrikov/jaxrl?
- JAX (Flax) implementations of deep reinforcement learning algorithms including Soft Actor Critic, DDPG, and AWAC for continuous action spaces.
- Is jaxrl open source?
- Yes — ikostrikov/jaxrl is open source, released under the MIT license.
- What language is jaxrl written in?
- ikostrikov/jaxrl is primarily written in Jupyter Notebook.
- How popular is jaxrl?
- ikostrikov/jaxrl has 757 stars on GitHub.
- Where can I find jaxrl?
- ikostrikov/jaxrl is on GitHub at https://github.com/ikostrikov/jaxrl.