OpenRLHF is an easy-to-use, scalable, and high-performance framework for Reinforcement Learning with Human Feedback (RLHF). It supports various training techniques and model architectures.
Features
- Implements Proximal Policy Optimization (PPO) for training
- Supports Iterative Direct Preference Optimization (DPO)
- Provides Low-Rank Adaptation (LoRA) for efficient fine-tuning
- Includes RingAttention and Retrieval-augmented Fine-Tuning (RFT)
- Scales to large models with high performance
- Offers comprehensive documentation and examples
Categories
Machine Learning, Reinforcement Learning Frameworks, Reinforcement Learning Libraries, Reinforcement Learning AlgorithmsLicense
Apache License V2.0Follow OpenRLHF
Other Useful Business Software
Try Google Cloud Risk-Free With $300 in Credit
Use your credit across every product. Compute, storage, AI, analytics. When it runs out, 20+ products stay free. You only pay when you choose to.
Rate This Project
Login To Rate This Project
User Reviews
Be the first to post a review of OpenRLHF!