Google AI Blog: MultiModel: Multi-Task Machine Learning Across Domains

MultiModel: Multi-Task Machine Learning Across Domains

Wednesday, June 21, 2017

Posted by Łukasz Kaiser, Senior Research Scientist, Google Brain Team and Aidan N. Gomez, Researcher, Department of Computer Science Machine Learning Group, University of TorontoMultiModelGoogle’s Multilingual Neural Machine Translation System

MultiModel architecture: small modality-specific sub-networks work with a shared encoder, I/O mixer and decoder. Each petal represents a modality, transforming to and from the internal representation.

Tensor2TensorAcknowledgementsGoogle Brain team
1 The 8 tasks were: (1) speech recognition (WSJ corpus), (2) image classification (ImageNet), (3) image captioning (MS COCO), (4) parsing (Penn Treebank), (5) English-German translation, (6) German-English translation, (7) English-French translation, (8) French-English translation (all using WMT data-sets).^↩

Blog

MultiModel: Multi-Task Machine Learning Across Domains

No comments :

Post a Comment

Labels

Archive

Feed