nvidia/C-RADIOv4-1D-H
Full fine-tune vs nvidia/C-RADIOv4-H. Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: other 1/2, Norms 96/98, Attention 96/96, MLP 96/96, Embeddings 2/2
Full fine-tune vs nvidia/C-RADIOv4-H. Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: other 1/2, Norms 96/98, Attention 96/96, MLP 96/96, Embeddings 2/2