In 2019 we’ve noticed a tremendous shift of client demand on model creation to model deployment and monitoring (CI/CD). This indicates a further step towards maturity of wider AI adoption within enterprises. In this talk we’ll introduce you to the latest developments in the most widely used DeepLearning framework TensorFlow 2.x. We show how Kubernetes and Kubeflow Pipelines work and how Open Data Hub provides a Open Source powered platform for all data science tasks. Finally, we show you an end to end project example of a product using all of those components in harmony.
Key Takeaways
- Understand the nature of a trained machine learning weights
- Overview of model interchange formats like TensorFlow SavedModel, PMML, ONNX
- Understand quantization
- Learn how TensorFlow Service, TensorFlow Light, TensorFlow.JS, Fabric for DeepLearning and KNative on Kubernetes work
Add comment