sakamakismile/gemma-4-12B-coder-fable5-composer2.5-GGUF

Quantization vs yuxinlu1/gemma-4-12B-coder-fable5-composer2.5-v1-GGUF. sakamakismile/gemma-4-12B-coder-fable5-composer2.5-GGUF is a GGUF quantization (Q5_K_M) of the same weights: 338 tensors match exactly and 329 within quantization error.

Tensors changed: Norms 0/289, other 0/50, Attention 0/184, MLP 0/144

On the Hub · Downloads on Model Pulse