Hacker Read

kristianp | karma 11941 | avg karma 3.44 · 2023-10-11 23:59:11

The base model is uncensored. From a glance at your link, that is about uncensoring the conversational training set.

nomel | karma 8158 | avg karma 2.06 · 2023-10-12 20:47:36

I don't think so: https://erichartford.com/uncensored-models#heading-whats-an-...

> Most of these models (for example, Alpaca, Vicuna, WizardLM, MPT-7B-Chat, Wizard-Vicuna, GPT4-X-Vicuna) have some sort of embedded alignment

> The reason these models are aligned is that they are trained with data that was generated by ChatGPT, which itself is aligned by an alignment team at OpenAI.

reply

kristianp | karma 11941 | avg karma 3.44 · 2023-10-12 22:19:36

Most of those are fine-tunes of the base model. The fine-tuning data is 'aligned'. The uncensored fine-tune training data is edited to remove the "I can't help you with that" responses.

nomel | karma 8158 | avg karma 2.06 · 2023-10-13 17:24:09

Yes, as stated in my earlier comment: there's an alignment tax, and then almost certainly an un-alignment tax, on top of that, compared to the raw, unaligned/uncensored, models.