Kavli Affiliate: Zeeshan Ahmed | First 5 Authors: Scott Novotney, Sreeparna Mukherjee, Zeeshan Ahmed, Andreas Stolcke, | Summary: We propose a framework to modularize the training of neural language models that use diverse forms of sentence-external context (including metadata) by eliminating the need to jointly train sentence-external and within-sentence encoders. Our approach, contextual universal embeddings […]
Continue.. CUE Vectors: Modular Training of Language Models Conditioned on Diverse Contextual Signals