soob3123/amoral-gemma3-4B-v1
LoRA-style vs google/gemma-3-4b-it. 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 34 of 34 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/4, Norms 0/316, MLP 102/210, Attention 136/352, other 0/1
5 derivatives checked
- bartowski/soob3123_amoral-gemma3-4B-GGUF: Same weights, converted (fewer layers)
- mradermacher/amoral-gemma3-4B-i1-GGUF: Quantization (fewer layers)
- soob3123/amoral-gemma3-4B-v1-gguf: Same weights, converted (fewer layers)
- mradermacher/amoral-gemma3-4B-GGUF: Same weights, converted (fewer layers)
- MaziyarPanahi/amoral-gemma3-4B-GGUF: Same weights, converted (fewer layers)