Sherpa

An implementation of llama.cpp that tries to recreate a chatbot

This app is a demo of the llama.cpp model.

You can find the source code of this app on GitHub

It was made on Flutter using an implementation of ggerganov/llama.cpp recompiled to work on mobiles.

It tries to recreate a chatbot that is offline, and works like OpenAI's ChatGPT.

You can fine tune it with preprompts.

Meta provides the 7B llama model for research. You need to download it on your device, rename it ggml-model.bin and put it in your download folder to run this app.

The LLaMA models are officially distributed by Meta and will never be provided by us.