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
I think your results have been really close to what their paper showed. It may be helpful to adjust the box threshold for different tasks and add token span for text prompts. Nevertheless, I also got a result similar with yours. (:
hi,i eval the gdino model,but get a bad result:
celebrity:0.3
color:9.9
landmark:14.8
logo:7.9
material:8.5
negation:17.3
position:27.6
relationship:9.2
平均:11.94
can you gvie me a eval code? thanks!
The text was updated successfully, but these errors were encountered: