MartinV/clement-mk-gemma-12b
Full fine-tune vs google/gemma-4-12B. MartinV/clement-mk-gemma-12b is a GGUF (Q5_K_M) of different weights. Fine-tune: 100% of the weights changed, across 48 of 48 layers.
Tensors changed: Embeddings 1/1, Norms 0/289, MLP 144/144, Attention 184/184