DireDreadlord/GemCod-Jade-270M
Full fine-tune vs google/gemma-3-270m-it. Fine-tune: 100% of the weights changed, across 18 of 18 layers.
Tensors changed: Norms 23/109, MLP 54/54, Attention 72/72
2 derivatives checked
- mradermacher/GemCod-Jade-270M-GGUF: Same weights, converted
- DireDreadlord/GemCod-Topaz-270M: LoRA-style