2 Comments

How do we prepare our own data sets? How should they enter the transfer model? thank you

Expand full comment

With Transformers package, things actually were easier.

For preparation, just treat the dataset like usual. You can make them as a DataFrame object with the words and labels.

To enter the transfer model, make sure the text is a string object and put them in the list. The rest you could follow in the example above.

Expand full comment