I have a training (gold) labelled dataset than consists of 10000 sentences. The task is to create a model that classifies correctly unseen data with B-I-O tags.

I have used a BERT and a GRU RNN model.

Now, I also have millions of sentences from the same field (law) but these have no label. I have tried to produce silver labels and re train the models on the full dataset but the results were not great. How can I use these unclassified data to improve a) a BERT model and b) the GRU model? Any ideas? Many thanks.

  • $\begingroup$ You can pretrain BERT and the GRU with your unlabeled data and then fine-tune them with the labeled data. $\endgroup$
    – noe
    Jun 22 at 7:20


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Browse other questions tagged or ask your own question.