about

https://huggingface.co/cognitivecomputations/Dolphin3.0-Llama3.2-1B abliterated with https://github.com/Orion-zhen/abliteration

The abliteration was done over the whole model, not layer by layer.

Dolphin being already quite uncensored, the lesser refusals should unleash it properly.

Downloads last month
33
Safetensors
Model size
1.24B params
Tensor type
BF16
·
FP16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Nexesenex/Dolphin3.0-Llama3.1-1B-abliterated

Finetuned
(1)
this model
Merges
13 models
Quantizations
3 models