anthracite-org/magnum-v3-27b-kto-gguf
Full fine-tune vs IntervitensInc/gemma-2-27b-chatml. anthracite-org/magnum-v3-27b-kto-gguf is a GGUF (Q8_0) of different weights. Fine-tune: 100% of the weights changed, across 46 of 46 layers.
Tensors changed: Embeddings 1/1, Attention 184/184, Norms 163/185, MLP 138/138