Platform for AI - Supports Model Training with Fully Managed Flink Resources
Apr 30 2022
Platform for AIMachine Learning Platform for AI now supports fully managed Flink resources. Users can purchase fully managed Flink resources, associate them with a workspace, and then use them to train distributed models in large-scale scenarios.
Content
Target customer: customers who own Flink resources and want to train machine learning models. Feature released: Designer of Machine Learning Platform for AI allows you to create AI pipelines by using drag-and-drop methods in the visualized console and PyAlink scripts. This allows you to use Flink computing resources to perform end-to-end data preprocessing, feature engineering, model training, and inference. You can store data in MaxCompute or Object Storage Service (OSS) and perform hybrid scheduling for computing nodes based on MaxCompute to meet your requirements.
