#132132nd of 206 in ML frameworks and libraries
xuance
Deep reinforcement learning library with single- and multi-agent algorithms
- Stars
- 1.1k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Jul 2026
- Language
- Python
1,045 on 21 Apr 20261,115 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
XuanCe is a Python library of deep reinforcement learning implementations covering DQN variants, policy-gradient and actor-critic methods, model-based (DreamerV2/V3), contrastive, and multi-agent algorithms such as QMIX, MAPPO and MADDPG. It runs on PyTorch, TensorFlow2 or MindSpore and works with Gymnasium and PettingZoo environments, including Atari, MuJoCo, SMAC and MPE. It is installed with pip and documented at xuance.org.
Who it is for: RL researchers and students running DRL and MARL experiments
Strengths
- Covers DRL, model-based, contrastive and MARL algorithms in one codebase
- Supports PyTorch, TensorFlow2 and MindSpore backends
- Works with many environments: Atari, MuJoCo, Box2D, MiniGrid, SMAC, MPE, drones
- Supports parallel environments, multi-GPU training, and TensorBoard or wandb logging
Weaknesses
- Reinforcement learning only; not for LLM or generative model work
- Each environment family needs its own extra packages
- Recommended Python 3.8 is old and may conflict with newer tooling
What it needs
- GPU optional
- Needs PyTorch, TensorFlow, MindSpore, Gymnasium, PettingZoo
What changed
From the release notes. No flag means the note does not say so, not that the upgrade is safe.
Also in ML frameworks and libraries
See all 206| Rank | Project | Score |
|---|---|---|
| 1 | transformersPython library defining open models for text, vision, audio and multimodal tasks | 94 out of 100 |
| 2 | ultralyticsPython library and CLI for training and running YOLO vision models | 94 out of 100 |
| 3 | pytorchPython tensor library with GPU support and tape-based autograd | 93 out of 100 |
| #4 | pytorch-image-modelsPyTorch library of image models, pretrained weights and training scripts | 93 out of 100 |
| #5 | supervisionPython toolkit for loading, annotating and counting computer vision detections | 92 out of 100 |