[Adapt] [Seminar] Multi-Task Deep Neural Networks for Natural Language Understanding

[郭子路]

Hi, Adapters, 
This week I will introduce a paper from Microsoft published in 31st Jan 2019 "Multi-Task Deep Neural Networks for Natural Language Understanding"
For many NLP task-specific tasks,  To label the task-specific data, we need very professional knowledge.  So the lack of labeled training data is a huge problem for these tasks. To solve the problem, pre-training and multi-task learning are proposed. 
After Google proposed BERT in 2018, Microsoft combined BERT with a multi-task learning method to create MT-DNN. I'll introduce the architecture and training procedure in detail this week in my talk.


See you then!


Time: 17:00, April, 10
Venue: SEIEE 3-414
Best, 
Zilu Guo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs.sjtu.edu.cn/pipermail/adapt/attachments/20190410/06bbe805/attachment.html>