bakhtin/Qwen3.8-27B-ED-5bpw-GGUF
Quantization vs ggml-org/Qwen3.8-27B-GGUF. bakhtin/Qwen3.8-27B-ED-5bpw-GGUF is a GGUF quantization (IQ4_XS/Q5_K) 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