Popular repositories Loading

1. verl verl Public

   verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework

   Python 21.6k 4k

2. verl-recipe verl-recipe Public

   A set of examples based on verl for end-to-end RL training recipes.

   Python 279 123

3. uni-agent uni-agent Public

   A unified framework for building, running, and training general agents at scale.

   Python 255 26

4. verl-omni verl-omni Public

   RL training framework for diffusion and omni-modality models

   Python 234 32

5. vexact vexact Public

   verl Zero-Mismatch Dense/MoE HuggingFace Rollout

   Python 44 5

6. rl-insight rl-insight Public

   Provide performance insight capabilities for RL frameworks.

   Python 30 22