First of all, thanks for sharing this great work!
I have tried to reproduce the performance of UCF-101-24 using your code, but it failed.
(100 epoch, [email protected] : 90.02 vs 95.1 (in the paper) / [email protected] : 84.62 vs 89.1 (in the paper)
In addition, the current version of the code is different from what you described in your README.md, so I'm having trouble reproducing the experiment.
It would be grateful if you could tell me how to run the experiments of UCF-101-24 based on the current version of the code.