cdiamond/Qwen3.8-27B-iMatrix-NVFP4-MTP-GGUF

Quantization vs Qwen/Qwen3.8-27B. cdiamond/Qwen3.8-27B-iMatrix-NVFP4-MTP-GGUF is a GGUF quantization (T40/Q5_K) of the same weights: 161 tensors match exactly and 90 within quantization error.

Tensors changed: Norms 0/161, Attention 0/64, Embeddings 0/1, MLP 0/24, lm_head 0/1

On the Hub · Downloads on Model Pulse