I noticed three different training epochs mentioned in the paper—50, 75, and 108. Could you clarify the exact number of epochs used to train the microsoft/conditional-detr-resnet-50 model?
· Sign up or log in to comment
