Can Sagemaker-trained models be deployed to non-Sagemaker endpoints?

Question

I am looking for a way to train model which will be deployed for inferencing, but Google AI Platform or an on-premises VM, not to a Sagemaker endpoint. Are Sagemaker-trained generic enough to be used for that? If so, could you point me to documentation?

Answer

Hello JoshuaFox,   
  
Yes, you can deploy a SageMaker trained model outside of SageMaker. How to do that exactly and the documentation depends on the framework and framework version.   
  
For example lets say you want to deploy a TensorFlow 2.X model outside of SageMaker, you would use something https://www.tensorflow.org/tfx/tutorials/serving/rest_simple.  
  
HTH,   
  
Chaitanya

Can Sagemaker-trained models be deployed to non-Sagemaker endpoints?

Contenuto pertinente