Gryphe/Gemma-4-12B-StyleTune
Same weights, repacked vs google/gemma-4-12B-it. Same weights in different files (re-saved, re-sharded or new metadata), plus 1 new tensors.
Tensors changed: Embeddings 0/12, Norms 0/289, other 0/48, MLP 0/144, Attention 0/184
7 derivatives checked
- mradermacher/Gemma-4-12B-StyleTune-i1-GGUF: Quantization
- mradermacher/Gemma-4-12B-StyleTune-GGUF: Quantization
- alexisStacksCode/Gemma-4-12B-StyleTune-QAT-GGUF: Full fine-tune
- motionsilse/Gemma-4-12B-QAT-Heretic-StyleTune-GGUF: Full fine-tune
- Wiself/gemma-4-12B-Styletune-Voice: Same weights, repacked (fewer layers)
- Wiself/gemma-4-12B-QAT-Styletune-Voice: Full fine-tune
- Retreatcost/Limn-Alpha-12B: LoRA-style
1 other repos hold byte-for-byte identical weight files.