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
Hi, thanks for your impressive work! I noticed that the text clip embeddings are calculated but don't use for any loss in your code. Could you please explain the reason for that? Does the supervision of the distance between the rendered image and the text embedding degrade the performance?
The text was updated successfully, but these errors were encountered:
Hi, thanks for your impressive work! I noticed that the text clip embeddings are calculated but don't use for any loss in your code. Could you please explain the reason for that? Does the supervision of the distance between the rendered image and the text embedding degrade the performance?
The text was updated successfully, but these errors were encountered: