Token classification is not working

[TekeshwarHirwani]

I have tried to run script of token classification and it showed an error that this script only works for fast tokenizer, I wanted to use biogpt for NER task, so any one have any idea how to fine-tune the model for that please let me know

[renqianluo]

Hi, you can format the task to a sequence-to-sequence format which is more convenient for the generative model. For NER task, you can follow the format in this paper We have tested it on BioGPT and it works well.

e.g., Input: Aspirin is tested in the experiment as the treatment to the patient. Output: Aspirin is tested in the experiment as the treatment to the patient.

[TekeshwarHirwani]

Thankyou for responding @renqianluo,

How can we perform few-shot learning in BioGPT for NER task ?