ncbi-nlp/bluebert

Evaluation on MedNLI dataset

passing2961 opened this issue · 1 comments

passing2961 commented

Hi,

I have one issues for evaluating my pretrained model on the MedNLI dataset.

The provided MedNLI dataset is jsonl format, but the code MedNLIProcessor in run_bluebert.py read the tsv format dataset.

Could you tell me how to figure it out?

Best,
Thanks

passing2961 commented

I figure it out by using create_mednli_bert.py source code.

Thanks