SageMaker endpoint deployment fails despite available ml.m5.large quota

### Product Version

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

### Issue Description

I am trying to deploy a JumpStart model using an ml.m5.large endpoint.

Service Quotas shows an applied quota of 4 with utilization 0.

SageMaker also shows 0 of 4 quota used, but the deployment fails with the error: limit: 0, in use: 0, requested: 1.

Both SageMaker and Service Quotas are using the same AWS Region.

Steps to reproduce:

Open SageMaker JumpStart.

1. Select a model and choose Deploy.
2. Select ml.m5.large.
3. Set instance count to 1.
4. Click Deploy.
5. Deployment fails with quota limit 0, even though Service Quotas shows 4 available.

### Expected Behavior

_No response_

### Observed Behavior

_No response_

### Product Category

JumpStart

### Feedback Category

Service Quotas

### Other Details

_No response_

Start by reproducing the deployment in SageMaker Studio JumpStart with an ml.m5.large instance, then compare the quota shown in SageMaker with the same-region value in Service Quotas. Done means the deployment succeeds with the available quota or the service clearly reports and resolves the quota discrepancy.

Written by the indexing model from the issue text.