google/gemma-4-E2B-it-qat-q4_0-unquantized
Full fine-tune vs google/gemma-4-E2B-it. Fine-tune: 91% of the weights changed, across 35 of 35 layers.
Tensors changed: MLP 105/585, Norms 177/433, other 104/240, Attention 100/687, Embeddings 3/6
34 derivatives checked
- google/gemma-4-E2B-it-qat-q4_0-gguf: Quantization (fewer layers)
- google/gemma-4-E2B-it-qat-w4a16-ct: Quantization
- unsloth/gemma-4-E2B-it-qat-GGUF: Not comparable yet
- huihui-ai/Huihui-gemma-4-E2B-it-qat-q4_0-unquantized-abliterated-GGUF: Abliterated (fewer layers)
- lmstudio-community/gemma-4-E2B-it-QAT-GGUF: Quantization (fewer layers)
- prithivMLmods/gemma-4-E2B-it-qat-GGUF: Same weights, converted (fewer layers)
- mradermacher/gemma-4-E2B-it-qat-q4_0-unquantized-i1-GGUF: Quantization (fewer layers)
- FreedomAISVR/Gemma-4-E2B-it-QAT-NVFP4-GGUF: Quantization (fewer layers)
- mlx-community/gemma-4-e2b-it-qat-OptiQ-4bit: Quantization
- JMingo/gemma-4-E2B-it-qat-GGUF: Same weights, converted (fewer layers)
- Luigi/gemma-4-E2B-meeting-agent-zh-GGUF: Quantization (fewer layers)
- DevQuasar/google.gemma-4-E2B-it-qat-q4_0-unquantized-GGUF: Quantization (fewer layers)
- mewse/gemma-4-E2B-home-assistant: LoRA-style
- rohitsoni1010/gemma-4-E2B-it-qat-q4_0-unquantized: LoRA-style
- mradermacher/gemma-4-E2B-it-qat-q4_0-unquantized-GGUF: Same weights, converted (fewer layers)
- hanemay/smartsightCoach: Full fine-tune
- FreedomAISVR/Gemma-4-E2B-it-QAT-MXFP4-GGUF: Quantization (fewer layers)
- alekshandru/gemma-4-E2B-it-qat-GGUF-mobile: Not comparable yet
- asincole/gemma-4-E2B-it-qat-q4_0-gguf: Quantization (fewer layers)
- nwtgck/gemma-4-E2B-it-qat-q4_0-gguf: Quantization (fewer layers)
- unsloth/gemma-4-E2B-it-qat-mobile: Quantization
- TheStageAI/gemma-4-E2B-it-qat-GGUF: Quantization (fewer layers)
- roman220220/gemma-4-E2B-it-qat-phone-GGUF: Quantization (fewer layers)
- xbill9/gemma-4-E2B-it-qat-q4_0-exact-gguf: Quantization (fewer layers)
- hung2903/gemma-4-E2B-vaccine-xai-merged-qat: Quantization (fewer layers)
- Dzluck/Gemma-4-E2B-it-qat-unquantized-abliterated-GGUF: Abliterated (fewer layers)
- coder3101/gemma-4-E2B-it-qat-q4_0-unquantized-heretic: Abliterated
- craquehouse/gemma-4-E2B-it-qat-oQ4e-fp16-text-only: Quantization
- MichaelAnthony/gemma4-e2b-Snowfox-GGUF: Quantization (fewer layers)
- erenyeager-1/gemma-4-E2B-it-qat-GGUF: Full fine-tune (fewer layers)
4 other repos hold byte-for-byte identical weight files.