mixbits/Qwen3.8-27B-NVFP4-MTP-VL-GGUF
Quantization vs Qwen/Qwen3.8-27B. mixbits/Qwen3.8-27B-NVFP4-MTP-VL-GGUF is a GGUF quantization (T40/Q5_K) of the same weights: 161 tensors match exactly and 2 within quantization error.
Tensors changed: Norms 0/161, Embeddings 0/1, lm_head 0/1