If you appreciate my work, cryptocurrency addresses are at the end of every model readme!
-
qwen3.5-uncensored
Qwen3.5-2B with refusal behavior largely removed using abliteration. No fine-tuning was involved. Computed in nothink mode, please use /set nothink in ollama for best results.
844 Pulls 1 Tag Updated 1 week ago
-
llama3.2-3b-uncensored
Built with Llama. This model is Meta's Llama 3.2 3B Instruct with refusal behavior largely removed using abliteration, aka fining the "refusal direction" in the model's activations and projects it out of the weights. No fine-tuning was involved.
770 Pulls 1 Tag Updated 6 days ago