演讲将分享如下主题：

1. TensorFlow 开发工作与最新进展。
2. Tensorflow 1.0 与 1.1，主要介绍 GPU 性能优化与内置的图编译框架 Xla。
3. Google 神经机器翻译系统。
4. 一些研究主题：AutoML（自动学习模型架构），机器学习在医疗保健方面的应用。

In his talk, Yuefeng will cover below topics:

1. If you missed the days to watch the Google I/O 2017  in real-time, Yuefeng will summarize the major announcements and developments for you.
2. Tensorflow 1.0 and 1.1, especially for the performance optimization on GPU and build-in graph compilation framework Xla.
3. Google neural machine translation system.
4. Some research topics:  AutoML(automatically learn the model architecture), deep learning on health-care.