StableVicuna is the first large-scale open-source chatbot trained throughReinforcement Learning from Human Feedback (RLHF)launched by StabilityAI, the company behind Stable Diffusion. StableVicuna is a further instruction fine-tuned and RLHF-trained version of Vicuna v0 13b, and it is an instruction fine-tuned LLaMA 13 billion model.
StableVicuna
The first large-scale open-source chatbot trained through RLHF
