[Feature Request] Serverless Inference with VPC Config


I would like to use a Sagemaker Model with a custom VPC Configuration, which is currently not possible with Serverless Inference. Is this feature planned? More generally: Is there a roadmap somewhere for Serverless Inference?

asked 2 years ago895 views
3 Answers

Any updates on this?

Btw, you should add a warning to the documentation of Model.deploy() here: https://sagemaker.readthedocs.io/en/v2.169.0/api/inference/model.html as I've been getting a ValidationException and I've been trying to debug it for hours, without a clue for why it's failing. I am also using a VPC config. Honestly AWS, fix your damn documentation.

answered a year ago
  • Hi! We also lost several hours because of incomplete and misleading documentation. What we ended up doing is to create a VPC network for our Redis Cluster which Sagemaker was not part of and handled the caching in the API calling the Sagemaker endpoint. Maybe you can use a similar approach.


SageMaker Serverless Inference is currently in preview and VPC support is not available but as the feature you are asking for is an important one and is on the roadmap( unfortunately I cannot share the exact details of the timelines here)

answered 2 years ago

How long will this be on preview. I hope when it comes out, it will have support for VPC

answered a year ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions