chrishoertnagl/SmolLM2-360M-Rickified-GRPO

Full fine-tune vs qgallouedec/SmolLM2-360M-Rickified. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: Embeddings 1/1, Norms 65/65, MLP 96/96, Attention 128/128

1 derivatives checked