After downloading the videos we use the command "! Python -m backend.parser train --model / models" to train. The 1/5000 epoch starts giving us all the IDs of the videos on which to train but then it does not continue and goes in loop. I am attaching a photo to better clarify the doubt.
