The model did not return a loss from the inputs, only the following keys: logits. For reference, the inputs it received are input_values

Intermediate
10

In my case the perfect solution was to have a column named “label”. My data set had the column “target”, “label_name” and so on. Seems like the column names made the issue.

I did toxic_encoded = toxic_encoded.rename_column("target", "label") and it started training.

2 Likes