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, I'm trying to use SCST on a diffusion model for image captioning. It uses Cross Entropy Loss but it's trained on global features (and not local ones), is there any way reinforcement learning is going to work on a model like this?
The text was updated successfully, but these errors were encountered:
Hi, I'm trying to use SCST on a diffusion model for image captioning. It uses Cross Entropy Loss but it's trained on global features (and not local ones), is there any way reinforcement learning is going to work on a model like this?
The text was updated successfully, but these errors were encountered: