p1atdev/dart-v1-sft
Full fine-tune vs p1atdev/dart-v1-base. Fine-tune: 98% of the weights changed, across 12 of 12 layers.
Tensors changed: Embeddings 1/2, Norms 25/50, MLP 48/48, Attention 96/96
Full fine-tune vs p1atdev/dart-v1-base. Fine-tune: 98% of the weights changed, across 12 of 12 layers.
Tensors changed: Embeddings 1/2, Norms 25/50, MLP 48/48, Attention 96/96