AxionLab-official/DogeAI-v1.5-Coder
Full fine-tune vs google/gemma-3-270m. Fine-tune: 100% of the weights changed, across 18 of 18 layers.
Tensors changed: Embeddings 1/1, Norms 59/109, MLP 54/54, Attention 72/72
3 derivatives checked
- WithinUsAI/Gemma3-Prompt.Coder.Uncensored.270m-GGUF: Full fine-tune
- mradermacher/DogeAI-v1.5-Coder-GGUF: Same weights, converted
- WithinUsAI/Gemma3-Prompt.Coder.it.Uncensored-270m: Full fine-tune