6 ms·
This is basically an ad for fine-tuning as a service. Can anyone offer an example of a free public-facing LLM which has been fine-tuned by adding much specific
by Animats 3y ago
This is basically an ad for fine-tuning as a service.
Can anyone offer an example of a free public-facing LLM which has been fine-tuned by adding much specific info about some narrow area? Say, one that knows all the PR about some car brand or fandom? Somebody must have tried that by now.
- treprinum 3y agoThe only known to me is LLaMA 2 to remove censorship using LoRA.
- Animats 3y agoHm. Why aren't those things all over the place?
- treprinum 3y agoThey kinda are, HuggingFace hub is already full of them. Most of them were released like 3-4 days ago so that might be a reason why you don't see them. I already put one of them into production.
- nickpsecurity 3y agoThe people that make them often post them or comment in the sub-Reddit below. I've linked to a search for uncensored models in there. Usernames of main people involved are faldore and The-Bloke. Most had Wizard Uncensored in their names. https://www.reddit.com/r/LocalLLaMA/search/?q=uncensored&restrict_sr=1 https://www.reddit.com/r/LocalLLaMA/search/?q=uncensored&res... People have posted comparisons showing it works pretty well. Some foundational models have enough censorship built in that they still try to resist it. You have to prompt those cleverly to do what you want them to do.
- fnordpiglet 3y agoThere are tons of them for other models, often Alpaca or LLaMA. This is a popular tool for finetuning Alpaca: https://github.com/tloen/alpaca-lora https://github.com/tloen/alpaca-lora
- treprinum 3y agoLLaMA or Alpaca aren't free outside research though, so it's better to focus on LLaMA 2.
- fnordpiglet 3y agoThere are lots of guides out there, but most these days tend to be selling something under the covers. Here’s one a quick Kagi found on llama 2: https://towardsdatascience.com/fine-tune-your-own-llama-2-model-in-a-colab-notebook-df9823a04a32 https://towardsdatascience.com/fine-tune-your-own-llama-2-mo...
- joshka 3y agoTake a look at https://replicate.com/blog/fine-tune-llama-to-speak-like-homer-simpson https://replicate.com/blog/fine-tune-llama-to-speak-like-hom... (and the related blog posts) for a good example of this.
- Animats 3y agoWell, OK. They put in 61,000 lines, the entire first 12 seasons of the Simpsons's. Lines such as {'previous': 'Marge Simpson: Ooo, careful, Homer.', 'character': 'Homer Simpson', 'line': "There's no time to be careful."} and then used it to generate more Simpsons dialogue. With a data set so well matched to the goal, the algorithm barely matters. Remember that recent paper, "Copy is all you need."?[1] This is the ideal case for that approach. I'm thinking more in terms of loading in the detailed product descriptions and maybe manuals from a catalog, then letting users ask questions about how to do things with the products. [1] https://news.ycombinator.com/item?id=36758233 https://news.ycombinator.com/item?id=36758233
- muttled 3y agoNot exactly what you're asking, but I fine-tuned llama 65b on my texts and it knows a LOT about me. When I ask it about specific situations where I've talked a lot about the subject, it brings in other seemingly unrelated details I don't think would show up in a vector search: https://airic.serveo.net https://airic.serveo.net (I scrubbed PII from it and it's only public-facing for a few days)