ThomasTheMaker/gm3-270m-TinyGSM-reasoning

Full fine-tune vs unsloth/gemma-3-270m-it. 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

1 derivatives checked

• mradermacher/gm3-270m-TinyGSM-reasoning-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse