On: Uncensor any LLM with abliteration

[hammock]

Can you share the link?

[jcims → hammock]

Here are the models - https://huggingface.co/collections/failspy/abliterated-v3-66...

[msoad → hammock]

https://colab.research.google.com/drive/1VYm3hOcvCpbGiqKZb14...

[hammock → msoad]

Thanks. Forgive me I'm not a coder, what's the easiest way to use/run this?

[pelagicAustral → hammock]

Easiest way to test the one referenced on the post (neuraldaredevil-8b-abliterat-psq) is to simply deploy to HF Endpoints: https://ui.endpoints.huggingface.co/new?repository=mlabonne/...

[DonsDiscountGas → hammock]

If you've got a Google account you can run it on Colab (probably need to copy it to your account first)

[IncreasePosts → hammock]

Download ollama and import the model listed at the end of the article.

[Wheaties466 → hammock]

this is a jupyter notebook. so you'll need to download that.