werty1248/Mistral-Nemo-NT-Ko-12B-dpo
LoRA-style vs werty1248/Mistral-Nemo-NT-Ko-12B-sft. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 40 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/81, MLP 120/120, Attention 160/160, lm_head 0/1
2 derivatives checked
- mradermacher/Mistral-Nemo-NT-Ko-12B-dpo-i1-GGUF: Quantization
- mradermacher/Mistral-Nemo-NT-Ko-12B-dpo-GGUF: Quantization