Re: [leak] facebook llama most trained research language model

[Undescribed Horrific Abuse, One Victim & Survivor of Many]

some people at stanford finetuned the smallest of these models to
follow instructions like openai’s models

their code is at https://github.com/tatsu-lab/stanford_alpaca . the
finetuned model itself is not shared but people are replicating it
using the code. a web demo is at https://crfm.stanford.edu/alpaca/ .