meshllm/Llama-3.2-1B-Instruct-Q4_K_M-layers
Quantization (fewer layers) vs unsloth/Llama-3.2-1B-Instruct-GGUF. meshllm/Llama-3.2-1B-Instruct-Q4_K_M-layers matches part of unsloth/Llama-3.2-1B-Instruct-GGUF (10 of its 147 tensors; the rest of unsloth/Llama-3.2-1B-Instruct-GGUF has no counterpart). On the shared part: meshllm/Llama-3.2-1B-Instruct-Q4_K_M-layers is a GGUF quantization (Q4_K/Q6_K) of the same weights: 3 tensors match exactly and 7 within quantization error.
Tensors changed: other 0/1, Norms 0/2, MLP 0/3, Attention 0/4