Train compute-intensive models with Azure Machine Learning
Large-scale machine-learning and deep-learning models require ample compute power. Learn when to choose GPU compute, and how different frameworks help you to make optimal use of GPU compute during preprocessing, model training, and deployment.
Prerequisites
Before starting this learning path, you should be familiar with the Azure Machine Learning service and training compute-intensive machine-learning and deep-learning models.
Achievement Code
Would you like to request an achievement code?
Modules in this learning path
Choose GPU compute in Azure Machine Learning when training compute-intensive models. To reduce the time needed to process the data, store your data efficiently and use a data manipulation library compatible with GPU compute.
Train compute-intensive models with GPU compute in Azure Machine Learning. By monitoring workloads, you can find the optimal compute configuration. Distributed training allows you to train on multiple nodes to speed up training time.
Deploying large-scale models for real-time inferencing is challenging because of the model's size. Learn what you can do and which frameworks you can use to optimize your model's performance during model scoring.