What's the resolution of videos when eval on GoPro dataset in your paper? And any advice about the resolution when doing video debluring?

[double-fire-0]

In Table 2 of your paper, you report the result of your model on GoPro dataset. At what resolution did you get this result?

And can you give some advice about the resolution when doing video debluring?

From this issue #12 , it seems that this mode is hard to run at a high resolution.
And 180p is blurry in application, how can we apply this model on a higher resolution (eg.720p)
Is it possible for us to train this model on a lower resolution and inference on a higher resulution?