runanywhere/qwen3_8_27b_HNPU
Quantization vs unsloth/Qwen3.8-27B-GGUF. runanywhere/qwen3_8_27b_HNPU is a GGUF quantization (IQ1_M) of the same weights: 353 tensors match exactly and 498 within quantization error.
Tensors changed: Norms 0/209, other 0/289, Attention 0/160, MLP 0/192, lm_head 0/1