Cross account image access support for Sagemaker AI Inference Serverless

### 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 deploy a Sagemaker AI serverless inference endpoint, we received the following error:
```
ClientError: An error occurred (ValidationException) when calling the CreateEndpointConfig operation:

The container image URI is not a prebuilt SageMaker algorithm or deep learning image, or the bring-your-own container image URI does not reside in an Amazon ECR repository within the same AWS account creating the endpoint.
```

Which indicates that the container used for serverless must reside in the same account. Currently, our organization uses multiple accounts and we have our ECR repositories in a single account which can be accessed from a number of our other accounts. This allows us to deploy the same container across the organization without having to store multiple copies of the image in multiple accounts.

### Other Details

Issue also occasionally pops up in stack overflow, for example: https://stackoverflow.com/questions/76227877/sagemaker-serverless-inference-error-bring-your-own-container-image-uri-does-no

No source files, tests, or entry points are named. Start by reviewing the CreateEndpointConfig error and the linked Stack Overflow report to understand the serverless image restriction; done would mean defining and implementing support for deploying an endpoint from an ECR image in another AWS account.

Written by the indexing model from the issue text.