Hi, I just notice the csv split files which author provided in google_drive, the test.csv and val.csv is same, total 24777 samples. This split should be ssv2 offical validation split, so the provided acc in paper is base on official validation? What other methods are based on which split, are they all base on validation? Waiting for reply, thanks!
Hi, I just notice the csv split files which author provided in google_drive, the test.csv and val.csv is same, total 24777 samples. This split should be ssv2 offical validation split, so the provided acc in paper is base on official validation? What other methods are based on which split, are they all base on validation? Waiting for reply, thanks!