Using same dataset for training and evaluation in Full Pipeline run in Ai Centre?

I have used Data Manager to tag invoices. I have created an extraction that has created 4 files (2 folders, a split.csv file and a schema.json file). Is it possible to use the same dataset for training and evaluation when running a full pipeline run in AI Centre?

If not, what is the best way to split the data files extracted from Data Manager? Can I use the same dataset for both (See below).


If you extract the dataset from data manager without specifying that this is a test dataset only then it should be fine using same dataset for both input and evaluation. The split.csv file will be used to make sure that training and test set are indeed separated and that you are getting accurate results from the model.