You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In our V1 (hourglass decoder), we did not enclose normal prediction tasks. If you need that, you can directly add a head to predict the normal, and follow ViT-models losses for training.
I want to train the ConvNext model, but the training script that is given only works with RAFTDecoder.
How to train the ConvNext model?
The text was updated successfully, but these errors were encountered: