llm

llm copied to clipboard

Optionally add a subcommand to pull it.

Legally, the model is only available to academics and researchers: https://github.com/facebookresearch/llama/blob/main/README.md#llama

That being said, you may want to look at the PRs of that repository to get a better understanding of the current state of affairs. Once you have the model, you'll need to convert it to ggml format yourself.

It can't be distributed by setzer due to the original license.

that would be nice to have in the readme

There's also a torrent for the pre quantized version. Not gonna link it here but you can find it pretty easily @leiserfg

As @philpax already pointed out, the only legal source is facebook, and you have to go through their channels. But it is well known that the weights leaked a few days ago.

I added a note on the README, to better document this :+1: I'll also leave the issue open for a few days to avoid potential duplicates.

Not sure what is kosher or not to say here, but several projects offer helpers to download the sources. I found the Dalai (LLaMA) project especially helpful as it bundles together both original model fetching and quantize, for all sizes. There's even a Dockerfile in a PR.

Hopefully i've not said too much for this repo.

The files also seem to be available on Huggingface

I'm going to close this as I think everything that can be said has been said 😅