finex666/Qwen3.8-27B-Abliterated-IQ4-MIX-MTP-GGUF
Quantization vs windowsxp811203/Qwen3.8-27B-Abliterated. finex666/Qwen3.8-27B-Abliterated-IQ4-MIX-MTP-GGUF is a GGUF quantization (IQ4_XS/IQ3_S) of the same weights: 161 tensors match exactly and 258 within quantization error.
Tensors changed: Norms 0/161, MLP 0/192, Attention 0/64, Embeddings 0/1, lm_head 0/1