This GitHub repository integrates Reflection Llama-3.1 70B, the world's top open-source large language model (LLM), in a Streamlit-based user interface. The LLM is trained using a new technique called Reflection-Tuning, which enables it to detect and correct mistakes in its own reasoning.
MIT License