About fine-tuning

[darwonsamal]

I would love to use train this on the bert-large corpus and then fine-tune it on the BC5CDR-chem corpus for NER and then use it to predict on unlabelled raw text. Any help/advice/instructions would be highly appreciated.

This is what the BC5CDR-chem corpus looks like, is this the correct format for this project?

drug O
withdrawal O
. O

OBJECTIVES O
: O
The O
United O
Kingdom O
Parkinson O
' O
s O
Disease O
Research O
Group O
( O
UKPDRG O
) O

[ranjeetds]

@darwonsamal This is correct format.