search

AndriyMulyar / bert_document_classification

architectures and pre-trained models for long document classification.
154 stars 47 forks source link

Document embedding #2

Closed zahra-rou closed 4 years ago

zahra-rou commented 4 years ago

Is it possible to produce an embedding or a fixed-size vector for each document?

AndriyMulyar commented 4 years ago

Yes but it will likely not be very useful unless finetuned via some supervised task.

zahra-rou commented 4 years ago

Thanks for your answer. So is it possible to get the vector for each document before fine-tuning the model?