llama3.1-gutenberg-8B-Heretic

A decensored version of nbeerbower/llama3.1-gutenberg-8B, made using Heretic v1.1.0

llama3.1-gutenberg-8B-Heretic Original model (llama3.1-gutenberg-8B)
Refusals 3/100 97/100
KL divergence 0.0615 0 (by definition)

Heretic Abliteration Parameters

Parameter Value
direction_index 13.98
attn.o_proj.max_weight 1.34
attn.o_proj.max_weight_position 19.54
attn.o_proj.min_weight 1.33
attn.o_proj.min_weight_distance 12.82
mlp.down_proj.max_weight 1.22
mlp.down_proj.max_weight_position 24.24
mlp.down_proj.min_weight 1.06
mlp.down_proj.min_weight_distance 11.39


llama3.1-gutenberg-8B

VAGOsolutions/Llama-3.1-SauerkrautLM-8b-Instruct finetuned on jondurbin/gutenberg-dpo-v0.1.

Method

Finetuned using 2x RTX 4060 for 3 epochs.

Fine-tune Llama 3 with ORPO

Downloads last month
5
Safetensors
Model size
8B params
Tensor type
F16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for ChiKoi7/llama3.1-gutenberg-8B-Heretic

Finetuned
(5)
this model
Quantizations
2 models

Dataset used to train ChiKoi7/llama3.1-gutenberg-8B-Heretic