Hi,
I’ve created a notebook for you to illustrate this: Transformers-Tutorials/Fine_tuning_BERT_(and_friends)_for_multi_label_text_classification.ipynb at master · NielsRogge/Transformers-Tutorials · GitHub
Actually, there’s no need for a MultilabelTrainer anymore, as you can just set the problem_type of the model’s configuration to “multi_label_classification”.