Trinity-RFT is a general-purpose, flexible and scalable framework designed for reinforcement fine-tuning (RFT) of large language models (LLM).
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).