allenai/MolmoPoint-GUI-8B
Full fine-tune vs allenai/MolmoPoint-8B. Fine-tune: 93% of the weights changed, across 36 of 36 layers.
Tensors changed: Norms 246/246, MLP 172/172, Attention 272/272, Embeddings 8/10, other 19/19
Full fine-tune vs allenai/MolmoPoint-8B. Fine-tune: 93% of the weights changed, across 36 of 36 layers.
Tensors changed: Norms 246/246, MLP 172/172, Attention 272/272, Embeddings 8/10, other 19/19