aws / aws/amazon-sagemaker-feedback

Support for OCI image manifest containers for inference endpoints

Open
#180 2 comments 5 reactions 0 assignees View on GitHub
feature-request
Dominant language
No language data
Stars
10
Forks
3
PR merge metrics
No merged PRs in 30d

Description

### Product Version

- [ ] Amazon SageMaker Studio Classic
- [ ] Amazon SageMaker Studio
- [x] It is not related to SageMaker Studio

### Product Category

Inference and Endpoints

### Description

When trying to use a recent image for serving inference requests from a Sagemaker AI endpoint, we received this error:

```software.amazon.awssdk.services.sagemaker.model.SageMakerException: Unsupported manifest media type application/vnd.oci.image.manifest.v1+json for image /:1.0.0. Ensure that valid manifest media type is used for specified image. (Service: SageMaker, Status Code: 400, Request ID: *****)```

It would be great to be able to use a vendor neutral and open standard for our container builds. The OCI format is (partially) supported by ECR, so we're able to store our images in AWS, but unable to make use of them for Sagemaker AI Inference endpoint.

### Other Details

Same issue reported here: https://repost.aws/questions/QUQgFRXiCYTjCtyxH_e2x9yg/oci-format-unsupported-when-creating-a-sagemaker-model

Contributor guide

Open the contributing guide

Research direction

Review the reported SageMakerException and the linked AWS re:Post report to understand the OCI manifest validation failure. Confirm the endpoint behavior and define done as accepting OCI image manifest media types for inference endpoint images.

Written by the indexing model from the issue text.

Assessment

Tech stack
aws
Domain
cloud, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.