sagemaker-runtime

class moto.sagemakerruntime.models.SageMakerRuntimeBackend(region_name: str, account_id: str)

Implementation of SageMakerRuntime APIs.

Implemented features for this service

• [X] invoke_endpoint

  This call will return static data by default.

  You can use a dedicated API to override this, by configuring a queue of expected results.

  A request to get_query_results will take the first result from that queue. Subsequent requests using the same details will return the same result. Other requests using a different QueryExecutionId will take the next result from the queue, or return static data if the queue is empty.

  Configuring this queue by making an HTTP request to /moto-api/static/sagemaker/endpoint-results. An example invocation looks like this:

  expected_results = {
      "account_id": "123456789012",  # This is the default - can be omitted
      "region": "us-east-1",  # This is the default - can be omitted
      "results": [
          {
               "Body": "first body",
               "ContentType": "text/xml",
               "InvokedProductionVariant": "prod",
               "CustomAttributes": "my_attr",
           },
          # other results as required
      ],
  }
  requests.post(
      "http://motoapi.amazonaws.com/moto-api/static/sagemaker/endpoint-results",
      json=expected_results,
  )
  
  client = boto3.client("sagemaker", region_name="us-east-1")
  details = client.invoke_endpoint(EndpointName="asdf", Body="qwer")
  

• [X] invoke_endpoint_async

• [ ] invoke_endpoint_with_response_stream