We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
810d29a
There was an error while loading. Please reload this page.
The archive contains the weights for models trained 4 epochs on ImageNet with image size 256x256, batch size 8 and L1 or L2 loss.