How do you use dataloader to train an image classification model with data from different folders?

fastai
1

I trying to train an image classification model. I have three folders with source data. Also, I have label’s csv files for each folder. How do I use the dataloader to train my model with the data in the three folders?

Thank you for your help,
Henry

1 Like
2

I found the solution. You created a data frame from each folder’s csv files with the labels, then concatenate all the dataframes into one, and finally pass the concatenated dataframe to the dataloader.

2 Likes