jpetrina/Swift-Qwen3.8-27B-IQ4_XS-pure-GGUF
Quantization vs ukisai/Swift-Qwen3.8-27B-GGUF. jpetrina/Swift-Qwen3.8-27B-IQ4_XS-pure-GGUF is a GGUF quantization (IQ4_XS/F32) of the same weights: 360 tensors match exactly and 506 within quantization error.
Tensors changed: Norms 0/216, other 0/290, Attention 0/164, MLP 0/195, lm_head 0/1