Rangle2/gemma-4-12B-it-uncensored-opus4.7-cot
LoRA-style vs google/gemma-4-12B-it. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, o_proj, q_proj, up_proj, v_proj changed, in 48 of 48 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/12, Norms 0/289, other 0/48, MLP 18/144, Attention 94/184
5 derivatives checked
- mradermacher/gemma-4-12B-it-uncensored-opus4.7-cot-i1-GGUF: Quantization
- mradermacher/gemma-4-12B-uncensored-opus4.7-cot-i1-GGUF: Quantization
- mradermacher/gemma-4-12B-uncensored-opus4.7-cot-GGUF: Quantization
- Rangle2/gemma-4-12B-it-uncensored-opus4.7-cot-GGUF: Same weights, converted
- mradermacher/gemma-4-12B-it-uncensored-opus4.7-cot-GGUF: Quantization