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 uni-perceiver V2, Sk denote the sampling ratio of each task.
When using unmixed sampling, you only sample one task per iteration, so does the value of K in LOSS equal to 1 ?
If that so, does Sk make a real difference?
The text was updated successfully, but these errors were encountered:
In uni-perceiver V2, Sk denote the sampling ratio of each task.
When using unmixed sampling, you only sample one task per iteration, so does the value of K in LOSS equal to 1 ?
If that so, does Sk make a real difference?
The text was updated successfully, but these errors were encountered: