byteshape/Qwen3.8-27B-GGUF
Quantization vs Qwen/Qwen3.8-27B. byteshape/Qwen3.8-27B-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
1 derivatives checked
- islamsidratul/Qwen3.8-27B-ByteShape-IQ4_XS-ASCII-GGUF: Same weights, repacked