Skip to content

Latest commit

 

History

History
24 lines (18 loc) · 590 Bytes

README.md

File metadata and controls

24 lines (18 loc) · 590 Bytes

Llama with ExecuTorch.js

This is an example showing how to run Llama 3.2 with ExecuTorch.js, the code is not production ready.

How to use

Download model:

$ npm install -g @frost-beta/huggingface
$ huggingface download frost-beta/llama3.2-instruct-executorch-kv-xnnpack-bf16

Run this script:

$ npm install
$ node chat.mjs llama3.2-instruct-executorch-kv-xnnpack-bf16
You> Who are you?
Assistant> I'm an artificial intelligence model known as Llama. Llama stands for "Large Language Model Meta AI."
You>