theprint/Llama3.2-3B-Explained
LoRA-style vs meta-llama/Llama-3.2-3B-Instruct. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of k_proj, o_proj, q_proj, v_proj changed, in 28 of 28 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/57, MLP 0/84, Attention 112/112
3 derivatives checked
- mradermacher/Llama3.2-3B-Explained-i1-GGUF: Quantization
- theprint/Llama3.2-3B-Explained-GGUF: LoRA-style
- mradermacher/Llama3.2-3B-Explained-GGUF: LoRA-style