Nexesenex/Llama_3.x_70b_L3.3-calme-2.3_abliterated_fusion
Full fine-tune vs MaziyarPanahi/calme-2.3-llama3.1-70b. Fine-tune: 100% of the weights changed, across 80 of 80 layers.
Tensors changed: Embeddings 1/1, Norms 150/161, MLP 240/240, Attention 320/320, lm_head 1/1