docling-project/granite-docling-2stage-258m
Full fine-tune vs ibm-granite/granite-docling-258M. Fine-tune: 100% of the weights changed, across 30 of 30 layers.
Tensors changed: other 1/1, Norms 111/111, MLP 138/138, Attention 216/216, Embeddings 3/3
Full fine-tune vs ibm-granite/granite-docling-258M. Fine-tune: 100% of the weights changed, across 30 of 30 layers.
Tensors changed: other 1/1, Norms 111/111, MLP 138/138, Attention 216/216, Embeddings 3/3