Requesting scripts for data generation

[g1910]

Hi,

Thank you for making the code available for TRN. I am having a very hard time reproducing the results using features pretrained on Kinetics. Also I am noticing huge variation in performance based on minor changes in the data generation code? Could you please share the script you used to generate the training and test split for training TRN? Without the actual code to generate data, there are inconsistencies in the experiments and this hurts in making any progress. Please help!

Thanks
Gaurav