google/gemma-4-12B-it-qat-q4_0-unquantized
Full fine-tune vs google/gemma-4-12B-it. Fine-tune: 100% of the weights changed, across 48 of 48 layers.
Tensors changed: Embeddings 5/12, Norms 193/289, other 46/48, MLP 144/144, Attention 184/184
50 derivatives checked
- google/gemma-4-12B-it-qat-w4a16-ct: Quantization
- google/gemma-4-12B-it-qat-q4_0-gguf: Quantization
- unsloth/gemma-4-12B-it-qat-GGUF: Not comparable yet
- lmstudio-community/gemma-4-12B-it-QAT-GGUF: Quantization
- cyankiwi/gemma-4-12B-it-qat-AWQ-INT4: Quantization
- huihui-ai/Huihui-gemma-4-12B-it-qat-q4_0-unquantized-abliterated-GGUF: Abliterated
- mlx-community/gemma-4-12B-it-qat-OptiQ-4bit: Quantization
- dahara1/gemma-4-12B-it-qat-UD-japanese-imatrix: Quantization
- prithivMLmods/gemma-4-12B-it-qat-GGUF: Same weights, converted
- mradermacher/gemma-4-12B-it-qat-q4_0-unquantized-i1-GGUF: Quantization
- JMingo/gemma-4-12B-it-qat-GGUF: Same weights, converted
- unsloth/gemma-4-12B-it-qat-w4a16: Quantization
- huihui-ai/Huihui-gemma-4-12B-it-qat-q4_0-unquantized-abliterated: Abliterated
- FreedomAISVR/Gemma-4-12B-it-QAT-NVFP4-GGUF: Quantization
- wepiqx/gemma-4-12B-it-SHQ6-GGUF: Quantization
- Mer0vin8ian/gemma-4-12B-it-qat-GGUF: Not comparable yet
- Jhustle44/gemma-4-12B-it-qat-GGUF: Not comparable yet
- Amod1arc/gemma-4-12B-it-qat-GGUF: Not comparable yet
- OliviaRossi/Gemma-4-12B-it-QAT-Q5_K_L-MTP-GGUF: Quantization
- OsaurusAI/gemma-4-12B-it-qat-MXFP4: Quantization
- FreedomAISVR/Gemma-4-12B-it-QAT-MXFP4-GGUF: Quantization
- jmarceno/Elixir-Gemma-4-12B: LoRA-style
- mradermacher/gemma-4-12B-it-qat-q4_0-unquantized-GGUF: Quantization
- FastFlowLM/gemma-4-12B-it-qat-q4_0-gguf: Quantization
- llmfan46/gemma-4-12B-it-qat-q4_0-unquantized-uncensored-heretic: Abliterated
- DevQuasar/google.gemma-4-12B-it-qat-q4_0-unquantized-GGUF: Quantization
- goodolclint/gemma-4-12B-it-qat-q4_0-mlx: Quantization
- npario/gemma-4-12B-it-qat-OptiQ-4bit: Quantization
- coder3101/gemma-4-12B-it-qat-q4_0-unquantized-heretic: Abliterated
- Grunzig/gemma-4-12B-it-qat-oQ4e-fp16-mtp: Quantization
7 other repos hold byte-for-byte identical weight files.