Vicuna-13B: open-source chatbot trained by fine-tuning LLaMA on ChatGPT data

Post by amrb » Tue, Apr 04, 2023, 12:40 AM UTC

Post by amrb » Tue, Apr 04, 2023, 10:41 AM UTC

axutio wrote: post #3

amrb wrote: post #2

Weight are not released

Delta weights have been released, can be automatically applied to the LLaMA weights: https://news.ycombinator.com/item?id=35430432