MLSUS-02: Consider AI services and pre-trained models
Consider whether the workload needs to be developed as a custom model. Many workloads can use managed AI services accessible through an API. Using these services means that you won’t need to provision your own resources to collect, store, and process training data and to prepare, train, tune, and deploy an ML model.
If adopting a fully managed AI service is not appropriate, evaluate if you can use pre-existing datasets, algorithms, or models. You can also fine-tune an existing model starting from a pre-trained model. Using pre-trained models from third parties can reduce the resources needed for data preparation and model training.
Implementation plan
-
Use pre-trained AWS AI services - AWS AI services
integrate with applications through APIs to address common use cases such as personalized recommendations, image recognition, language analysis and translation, modernizing contact centers, improving safety and security, and increasing customer engagement. -
Use pre-trained models from AWS Marketplace - AWS Marketplace
offers over 1,400 ML-related assets that you can subscribe to. -
Use pre-trained models from SageMaker AI JumpStart - SageMaker AI JumpStart provides pre-trained, open-source models for a wide range of problem types to help you get started with machine learning. You can incrementally train and tune these models before deployment.
Documents
Blogs
Videos
