netcat420/MFANNv0.16.10
LoRA-style vs MaziyarPanahi/Llama-3-8B-Instruct-v0.4. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of gate_proj, k_proj, o_proj, q_proj, v_proj changed, in 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/65, MLP 31/96, Attention 127/128
1 derivatives checked
- mradermacher/MFANNv0.16.10-GGUF: Same weights, converted