Methods to keep text local and score on the cloud


(Nathan Boright) #1

We have a use case where we want to execute textcat models in the cloud, but the clear text cannot go to the cloud as it contains P.I.I and other sensitive information. Would extracting token vectors from doc objects and sending those to the cloud for the models to score possible? I know I can extract the vectors, but haven’t found a way to run the models apart from the doc objects…

Thank You