Tetratype/Gemma-4-26B-A4B-it-heretic-antislop
Partial fine-tune vs coder3101/gemma-4-26B-A4B-it-heretic. Partial change: 18% of the weights changed (attention, embeddings, mlp), in 5 of 30 layers.
Tensors changed: Embeddings 1/4, MLP 25/231, Norms 0/433, other 0/122, Attention 5/223
2 derivatives checked
- mradermacher/Gemma-4-26B-A4B-it-heretic-antislop-i1-GGUF: Quantization
- mradermacher/Gemma-4-26B-A4B-it-heretic-antislop-GGUF: Quantization