Novaciano/Gemma3-HADES-1B
LoRA-style vs Novaciano/Gemma3-Persephone-1B. 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 25 of 26 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/157, MLP 75/78, Attention 100/104
3 derivatives checked
- mradermacher/Gemma3-HADES-1B-i1-GGUF: Quantization
- mradermacher/Gemma3-HADES-1B-GGUF: Same weights, converted
- Novaciano/Prophesy-1B: LoRA-style