Data folder structure

[avnishks]

Hi,

Thank you for making your code public. I am trying to run your single_model_finetune code on just the Stanford dog datasets. That dataset has separate folders for each of the 120 classes.

Could you please clarify how is your data folder is structured and how are you dividing your data into training, validation and testing sets?

[fenglf]

split training, validation and testing sets with 0.8, 0.1, 0.1 randomly.
training and validation sets keep 120 classes folder, while testing sets with 1 folder including all testing image data.