I have a TextDataBunch created with validation split as 0.2. And have predicted using my learn object. However I want to bring my actual text(with no tokenization and numericilization applied), along with true label and predicted label in one dataframe. How can I do it? i tried below approaches
- data_clas.train_ds.to_df() – But threw error
- Tried to iterate over data_clas.train_ds but in vain (can only get numericalized text but not original text)
Any suggestions would be good !