about the test dataset

[Juice113]

Thanks for your work!
in the paper, "For training purposes, 95% of the samples are randomly selected to construct the training dataset, while the remaining 5% form the test dataset for validation and analysis."
I wonder which are selected as the test set? I didn't find those test data in the dataset.
looking forward to your reply.