rasyosef/Llama-3.2-3B-Instruct-DSpark

Full fine-tune vs unsloth/Llama-3.2-3B-Instruct. Fine-tune: 100% of the weights changed, across 5 of 5 layers.

Tensors changed: Norms 11/11, MLP 15/15, Attention 20/20

1 derivatives checked

• rasyosef/Llama-3.2-3B-Instruct-DSpark-GGUF: Partial fine-tune

On the Hub · Downloads on Model Pulse