Question related to the assigned directories with TensorFlow

I have watched the video “Training with the cats vs. dogs dataset”, but it still bothers me a bit how each file is assigned into subdirectories(Ie., cats, dogs), like is there a naming rule or something that allows the files to be transferred into the correct directories?
Please give me some information/intuition and maybe examples about this matter, thank you.

are you stuck in creating the root directories?

probably the below comment should help you


I completed the assignment long ago, but what I don’t understand is how each file finds its way to the correct subdirectory(cats, dogs to be specific) after splitting into training/validation directories. I am asking about intuition or some explanation to get the idea more clearly for future problems. :smile:

more than intuition, knowing to write the right path is based on one’s understanding about his/her folders, and making sure to follow that your programming tools is using the local directory where it will look for your dataset.

If you had read that comment, you would have understood.

Thank you I finally saw it.