groxaxo/QwenPaw-Flash-9B-heretic-MTP-GPTQPro-INT4-g64
Quantization vs SC117/QwenPaw-Flash-9B-heretic-MTP-GGUF. groxaxo/QwenPaw-Flash-9B-heretic-MTP-GPTQPro-INT4-g64 is a GGUF quantization (BF16/Q4_0) of the same weights: 242 tensors match exactly and 200 within quantization error.
Tensors changed: lm_head 0/1, other 0/146, Norms 0/112, Attention 0/84, MLP 0/99