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
During inference, we output a single category and then aggregate all categories to obtain multi-category results. However, it is also possible to specify multiple categories in the instructions, but doing so will lead to a performance decrease because we used single-category training.
Very insightful work!
I would like to ask, during training, the ADE dataset is trained on a single category, what about during inference?
The text was updated successfully, but these errors were encountered: