Amy Unruh is a developer relations engineer for the Google Cloud Platform, where she focuses on machine learning and data analytics as well as other Cloud Platform technologies. Amy has an academic background in CS/AI and has also worked at several startups, done industrial R&D, and published a book on App Engine.
Session: Scaling Machine Learning on Google Cloud Platform
Recent advances in Machine Learning, in both hardware and software, have made ML more powerful and accessible. But it can be a challenge to scale out, as ML architectures grow more complex and datasets become larger. Support for distributed training with accelerators becomes increasingly essential, as does the need to scalably manage and serve trained models.
Google Cloud Platform (GCP) has a range of products and tools that can help you do machine learning at scale. They span the spectrum from pre-trained ML models that you can access with a REST API; to ways to tune trained models with your own data, including the recently-announced AutoML; to support for scalably training and serving ML models (TensorFlow and other) via a managed service (Cloud ML Engine) or “DIY” (e.g. Kubeflow on Google Kubernetes Engine).
This talk will give an overview of the different ways that you can do machine learning on GCP, and show some demos.