train and validation data split

[maryawwm]

hello

I found it interesting to make a captioning model in another language and want to try it.
I've downloaded the dataset from the website and based on your instruction on repository put the dataset in image&caption directory and the json file in annotation directory.
and put all the images in training, but after training the model after first epoch I got error :

FileNotFoundError: [Errno 2] No such file or directory: 'image&caption/validation/7325.png'

I don't know how to spilt the images into train and validation data and I didn't find about it on dataset website and the repo.

It would be great to help me how to fix this to run the code.

[lxt-007]

The dataset contains all the photos. I think you could find 7325.png in the dataset. Do you know how to test the model?