ji-farthing/gemma-4-qat-q4_0-MTP-assistants-ik-llama-GGUF
Quantization vs google/gemma-4-12B-it-qat-q4_0-unquantized-assistant. ji-farthing/gemma-4-qat-q4_0-MTP-assistants-ik-llama-GGUF is a GGUF quantization (Q4_0) of the same weights: 21 tensors match exactly and 21 within quantization error.
Tensors changed: Embeddings 0/1, Norms 0/21, MLP 0/12, Attention 0/8