search

stanfordmlgroup / chexpert-labeler

CheXpert NLP tool to extract observations from radiology reports.
MIT License
340 stars 79 forks source link

How to run the labeler on GPU. #31

Closed ashwanikumar04 closed 3 years ago

ashwanikumar04 commented 3 years ago

Hello:

When running the labeler on a laptop, it is taking too much time for a few thousand items. Is it possible to run the same on GPU?

Thanks

jirvin16 commented 3 years ago

Hello:

When running the labeler on a laptop, it is taking too much time for a few thousand items. Is it possible to run the same on GPU?

Thanks

Unfortunately the labeler cannot be run on GPU.

ashwanikumar04 commented 3 years ago

Thanks for the reply.

binerone commented 3 years ago

One approach I am using is to split the text into multiple parts, with each individual sample being 500 or smaller, this will reduce the memory requirements. Then using multi-threaded processing, such as 32, will greatly reduce the processing time.