Search
 
loader loading subjects...
cover
Inference Time of a CamemBERT Deep Learning Model for Sentiment Analysis of COVID Vaccines on Twitter
Guillaume Guerdoux, Théophile Tiffet, Cedric Bousquet
269 - 270
10.3233/SHTI220714
Research Article
Studies in Health Technology and Informatics
Volume 295: Advances in Informatics, Management and Technology in Healthcare
Abstract

In previous work, we implemented a deep learning model with CamemBERT and PyTorch, and built a microservices architecture using the TorchServe serving library. Without TorchServe, inference time was three times faster when the model was loaded once in memory compared when the model was loaded each time. The preloaded model without TorchServe presented comparable inference time with the TorchServe instance. However, using a PyTorch preloaded model in a web application without TorchServe would necessitate to implement functionalities already present in TorchServe.

Download
Creative Commons License