Pinned Loading

1. OpenRLHF/OpenRLHF OpenRLHF/OpenRLHF Public

   An Easy-to-use, Scalable and High-performance RLHF Framework based on Ray (PPO & GRPO & REINFORCE++ & LoRA & vLLM & RFT)

   Python 6.7k 654

2. volcengine/verl volcengine/verl Public

   verl: Volcano Engine Reinforcement Learning for LLMs

   Python 8k 933

3. THUDM/SWE-Dev THUDM/SWE-Dev Public

   SWE-Dev is an open-source SWE agent with a scalable test case construction pipeline. This pipeline synthesizes test cases through a two-step process: generating Gherkin descriptions and correspondi…

   Python 25

4. Generalization-of-Transformers Generalization-of-Transformers Public

   [ICLR'25] Understanding the Generalization of In-Context Learning in Transformers: An Empirical Study

   Python 3