gpt4all
Training assistant chatbot with large scale data distillation from GPT-3.5-Turbo.
The readme describes the GPT4All project, which includes a demo, data, and code to train a language model based on LLaMa with around 800k GPT-3.5-Turbo generations. The repo includes instructions on downloading and running the quantized model on different operating systems, as well as reproducing the trained LoRa weights. The readme provides examples of prompts and corresponding model outputs, and includes a citation for use of the repository in downstream projects.