fearlessdots/Llama-3-Alpha-Centauri-v0.1

Text Generation • 8B • Updated May 25, 2024 • 7 • 10

Note This is the first model of the series. My main objective here was to get a good model (Llama-3) that was uncensored (https://huggingface.co/failspy/Meta-Llama-3-8B-Instruct-abliterated-v3) and fine-tune it to, at least try, to remove as much of the remaining censorship as possible. The result looks cool. Now I plan to use this uncensored model as base for the next models, which will be trained on diverse topics.