facebook/sapiens2-seg-0.4b

Full fine-tune vs facebook/sapiens2-pretrain-0.4b. Fine-tune: 100% of the weights changed, across 24 of 24 layers.

Tensors changed: Attention 216/216, Norms 97/97, MLP 96/96, Embeddings 2/3

On the Hub · Downloads on Model Pulse