How can I get the test accuracy of action recognition benchmark?

[Hou9612]

Thanks for this wonderful work!
I can't find the ground truth file of test dataset in this repository and the competition EPIC-KITCHENS-55 Action Recognition has ended, so how can I get the test accuracy of action recognition benchmark? I want to get the test accuracy of my model both on epic-kitchens-55 and epic-kitchens-100 datasets.

[dimadamen]

You can evaluate on the test sets through submitting to the leaderboards. The leaderboard fr EPIC-KITCHENS-55 Action Recognition is available here:
https://competitions.codalab.org/competitions/20115

[AntonyPapadakis]

Hello there! Thank you for creating and maintaining a great project!
In my case I am working with a subset ,i.e., some of the kitchens, of the epic-kitchens-55 dataset. I think I need ground truth test action labels, since I am not using the entire dataset to perform my experiments. Basically, I want to calculate accuracy on certain kitchens and not the whole dataset, so I guess the leaderboard option will not do. How can I get the the test accuracy in my case using a subset of the published test data ?