Machine Intelligence at Google Scale

The biggest challenge of Deep Learning technology is the scalability. As long as using single GPU server, you have to wait for hours or days to get the result of your work. This doesn't scale for production service, so you need distributed training on the cloud eventually, or take advantage of pre-trained models. Google has been building infrastructure for training the large scale neural network on the cloud for years, and started to share the technology with external developers. In this session, we present pre-trained ML services such as Cloud Vision API and Speech API that works without any training. In addition, we introduce Cloud AutoML, which helps customizing our pre-trained models with your data. Also, we look at how TensorFlow and Cloud Machine Learning can accelerate custom model training with Google's distributed training infrastructure.