Question for Evaluation

[lbg030]

Hi, I have question for evaluation.
I checked the evaluation and found that the evaluation is being conducted with the image downscaled.
When I tried to do it with a high resolution, it couldn't be executed, but was the performance published in the paper also evaluated with a downscaled resolution?

Thank you.

[lbg030]

And I tried the original extension of Replica, but I got an error on corr.py .
How can I evaluate with the original scale?