Open Tombery1 opened 2 years ago
You can get the sum of trainable weights from a model with sum(p.numel() for p in model.parameters() if p.requires_grad)
or using the torchinfo package to get a summary with that information.
Thank you, I want to know how many parameters in the original paper, to compare the results with another model.
@ahatamiz would you have the parameter count for your paper implementation? Thanks!
Hi @ahatamiz, besides the parameters, I see the FLOPs of Swin UNETR reported in the paper is 394.84G, would you release the code to calculate this? Thank you.
Hi, I want to know, what is the number of parameters of model Swin UNETR?