google/gemma-3-12b-it-qat-q4_0-unquantized
Full fine-tune vs google/gemma-3-12b-it. Fine-tune: 97% of the weights changed, across 48 of 48 layers.
Tensors changed: Embeddings 1/4, Norms 288/400, MLP 144/252, Attention 192/408, other 1/1
14 derivatives checked
- unsloth/gemma-3-12b-it-qat-GGUF: Same weights, converted (fewer layers)
- bartowski/google_gemma-3-12b-it-qat-GGUF: Same weights, converted (fewer layers)
- unsloth/gemma-3-12b-it-qat-bnb-4bit: Quantization
- mradermacher/gemma-3-12b-it-qat-q4_0-unquantized-i1-GGUF: Quantization (fewer layers)
- mradermacher/gemma-3-12b-it-qat-q4_0-unquantized-GGUF: Quantization (fewer layers)
- Uynaity/gemma-3-12b-it-qat-GGUF: Quantization (fewer layers)
- mlabonne/gemma-3-12b-it-qat-abliterated-GGUF: Quantization (fewer layers)
- DmitryDB/Gemma-3-12B-IT-QAT-Heretic-MOE-v1: Full fine-tune (fewer layers)
- unsloth/gemma-3-12b-it-qat: Identical copy
- mlabonne/gemma-3-12b-it-qat-abliterated: LoRA-style (fewer layers)
- soob3123/amoral-gemma3-12B-v2-qat: LoRA-style
- aifeifei798/QiMing-Socratic-12B: LoRA-style
- soob3123/GrayLine-Gemma3-12B: LoRA-style
- ConicCat/Gemma-3-12B-Fornax-QAT-CoT: LoRA-style
6 other repos hold byte-for-byte identical weight files.