Fastai Transformers - Fine Tuning With Custom Data Error:TextEncodeInput must be Union[TextInputSequence, Tuple[InputSequence, InputSequence]]

Hey guys, I was following the fastai transformers tutorial on a custom dataset and ran into this error when trying to create my dataset

TypeError: TextEncodeInput must be Union[TextInputSequence, Tuple[InputSequence, InputSequence]]

The notebook I am working off is in this repository along with my custom dataset.

Has anyone seen this error before? Are the docs out of date? If anyone has any ideas let me know, I would really love to fine tune on this dataset :slight_smile:

Most likely this is happening because of None values in the all_texts list.
https://stackoverflow.com/a/63870843 explains the same and worked for me.

Bless you for finding this. I will give it a shot tonight <3