facebook/sapiens2-pose-0.8b
Full fine-tune vs facebook/sapiens2-pretrain-0.8b. Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Attention 288/288, Norms 129/129, MLP 128/128, Embeddings 3/3
Full fine-tune vs facebook/sapiens2-pretrain-0.8b. Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Attention 288/288, Norms 129/129, MLP 128/128, Embeddings 3/3