Conversation

David Martínez Martí

I wonder, does anyone how to fine-tune an open source text AI? I'm seeing the latest developments and I'm starting to get interested on trying stuff myself - but not sure where to start, or where is the community.

1
0
0
@deavid I think I started with GitHub repos some time back when I did some fine-tuning of models … and Medium articles. But it’s been a few months and unfortunately, I’ve already forgotten all the sources 😛

If you want (and haven’t found some yourself) I can try to dig those up …
1
0
0

@f if you could just suggest something easy to start with it would help. Some tutorial or demo that I can use to get the gist of it

1
0
0
@deavid I haven’t tried this particular one myself, but I’ve been meaning to try fine-tuning LLaMA or Alpaca and so this looks like a good candidate:

https://github.com/lxe/simple-llm-finetuner

This Reddit thread and the linked YouTube video might be helpful too:

https://www.reddit.com/r/deeplearning/comments/11ryc3s/how_to_finetune_llama_models_smaller_models_with/

Plus, that particular sub-reddit and other similar channels are a great place to find people who are willing to help — or at least willing to talk about things 🙂
1
1
0

@f I'll take a look. Thanks a lot for the pointers!

1
0
1