Reputation: 429
How can I automatically update a SageMaker endpoint when the MLflow model aliased as `@Champion` changes?
I'm using a managed MLflow server on AWS SageMaker AI (https://www.youtube.com/watch?v=3xkz_5HOP6k&ab_channel=AWSEvents) to track experiments and model versions. Our data science team promotes production models by tagging the best version with a Champion alias. These models are deployed to SageMaker endpoints, which are then accessed via API Gateway.
I want to implement an automated pipeline that updates the SageMaker endpoint whenever a new model is assigned the Champion alias in MLflow. I'm currently considering a Lambda function that polls for alias changes, but I'm looking for a more efficient or managed solution. Has anyone implemented a dynamic endpoint update mechanism or can suggest an alternative approach?
Upvotes: -1
Views: 20
Answers (0)
Related Questions
- How do I print the model summary in PyTorch?
- Unable to properly register model and create Sagemaker Endpoint using Sagemaker Pipelines
- How to run the mlflow web-based user interface from Amazon SageMaker?
- How to access sagemaker model registry metrics from within the endpoint
- How to deploy the hugging face model via sagemaker pipeline
- AWS SageMaker TensorFlow Serving - Endpoint failure - CloudWatch log ref: "NET_LOG: Entering the event loop ..."
- How to update Sagemaker Endpoint with the newly Trained Model?
- Register SageMaker model in MLflow
- SageMaker endpoint for pretrained model
- AWS SageMaker: Use S3 pickled models instead of hosting on sagemaker