Emanon14/NONAMEmix_v1
Full fine-tune vs Laxhar/noobai-XL-0.75. Fine-tune: 100% of the weights changed, across 12 of 12 layers.
Tensors changed: Embeddings 590/590, other 659/662, Attention 716/716, Norms 546/546
Full fine-tune vs Laxhar/noobai-XL-0.75. Fine-tune: 100% of the weights changed, across 12 of 12 layers.
Tensors changed: Embeddings 590/590, other 659/662, Attention 716/716, Norms 546/546