sj5430/Qwen2.5-Omni-3B-Q4_K_M-GGUF
Quantization vs Qwen/Qwen2.5-Omni-3B. sj5430/Qwen2.5-Omni-3B-Q4_K_M-GGUF is a GGUF quantization (Q4_K_M) of the same weights: 95 tensors match exactly and 134 within quantization error.
Tensors changed: Norms 0/38, MLP 0/57, Attention 0/133, lm_head 0/1