bgg1996/Melinoe-Gemma4-31B-VL
Full fine-tune vs google/gemma-4-31B-it. Fine-tune: 98% of the weights changed, across 60 of 60 layers.
Tensors changed: Embeddings 1/4, Norms 214/523, other 0/62, MLP 180/261, Attention 230/338
4 derivatives checked
- mradermacher/Melinoe-Gemma4-31B-VL-i1-GGUF: Quantization (fewer layers)
- bgg1996/Melinoe-Gemma4-31B-VL-heretic: Abliterated
- mradermacher/Melinoe-Gemma4-31B-VL-GGUF: Quantization (fewer layers)
- bgg1996/Melinoe-Gemma4-31B-VL-GGUF: Same weights, converted (fewer layers)