Farama-Foundation/Metaworld
An open-source benchmark suite of 50+ robotics manipulation environments for evaluating multi-task and meta reinforcement learning algorithms.
Not currently ranked — collecting fresh signals.
star history
Meta-World provides a collection of continuous control robotic manipulation tasks designed to benchmark multi-task and meta-learning reinforcement learning algorithms. Built on Mujoco for physics simulation and following the Gymnasium API standard, it offers standardized benchmarks (MT1, MT10, MT50, ML1, ML10, ML45) to evaluate different aspects of RL algorithms including task generalization, few-shot learning, and transfer capabilities. The repository is maintained by the Farama-Foundation.
Frequently asked
- What is Farama-Foundation/Metaworld?
- An open-source benchmark suite of 50+ robotics manipulation environments for evaluating multi-task and meta reinforcement learning algorithms.
- Is Metaworld open source?
- Yes — Farama-Foundation/Metaworld is open source, released under the MIT license.
- What language is Metaworld written in?
- Farama-Foundation/Metaworld is primarily written in Python.
- How popular is Metaworld?
- Farama-Foundation/Metaworld has 1.9k stars on GitHub.
- Where can I find Metaworld?
- Farama-Foundation/Metaworld is on GitHub at https://github.com/Farama-Foundation/Metaworld.