leeroy-jankins/leeroy
Quantization vs unsloth/Llama-3.2-1B-Instruct-GGUF. leeroy-jankins/leeroy is a GGUF quantization (Q4_K_M) of the same weights: 34 tensors match exactly and 113 within quantization error.
Tensors changed: other 0/2, Norms 0/33, MLP 0/48, Attention 0/64