lsxi77777/Wat3R
Full fine-tune vs facebook/VGGT-1B. Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: other 229/229, Norms 408/408, Attention 208/208, MLP 212/212, Embeddings 345/346
Full fine-tune vs facebook/VGGT-1B. Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: other 229/229, Norms 408/408, Attention 208/208, MLP 212/212, Embeddings 345/346