Nexesenex/Llama_3.x_70b_Calme-L3.3_abliterated_fusion_v2
Partial fine-tune vs MaziyarPanahi/calme-2.3-llama3.1-70b. Partial change: 87% of the weights changed (attention, lm_head, mlp, norms), in 80 of 80 layers.
Tensors changed: Embeddings 0/1, Norms 161/161, MLP 215/240, Attention 259/320, lm_head 1/1
1 derivatives checked
- mradermacher/Llama_3.x_70b_Calme-L3.3_fusion_v2-GGUF: Quantization