ModelPackage doesn't support SageMaker Pipeline parameter

[duk-amz]

Describe the bug
A clear and concise description of what the bug is.
Using ModelPackage in a SageMaker Pipeline for inference. I expect to create a pipeline, use a model_version (string) as a parameter, and use ModelStep to dynamically create SageMaker Model and make batch inference. But if find this leads to an error

model_version is ParameterString

AttributeError                            Traceback (most recent call last)
<ipython-input-36-62c24131e39c> in <module>
      7 step_create_model = ModelStep(
      8     name="Create_Model",
----> 9     step_args=model.create(instance_type="ml.m5.large"),
     10     depends_on = [step_lambda]
     11 )
/opt/conda/lib/python3.7/site-packages/sagemaker/workflow/pipeline_context.py in wrapper(*args, **kwargs)
    263             if run_func.__name__ in ["register", "create"]:
    264                 self_instance.sagemaker_session.init_model_step_arguments(self_instance)
--> 265                 run_func(*args, **kwargs)
    266                 context = self_instance.sagemaker_session.context
    267                 self_instance.sagemaker_session.context = None
/opt/conda/lib/python3.7/site-packages/sagemaker/model.py in create(self, instance_type, accelerator_type, serverless_inference_config, tags)
    471             accelerator_type=accelerator_type,
    472             tags=tags,
--> 473             serverless_inference_config=serverless_inference_config,
    474         )
    475 
/opt/conda/lib/python3.7/site-packages/sagemaker/model.py in _create_sagemaker_model(self, *args, **kwargs)
   1604         container_def = {"ModelPackageName": model_package_name}
   1605 
-> 1606         self._ensure_base_name_if_needed(model_package_name.split("/")[-1])
   1607         self._set_model_name_if_needed()
   1608 

AttributeError: 'Properties' object has no attribute 'split'

To reproduce
A clear, step-by-step set of instructions to reproduce the bug.

model_package_arn = "arn:xxxxxxxx:1" # Should be a ParameterString
model = ModelPackage(
    role = role,
    model_package_arn = model_package_arn,
    sagemaker_session = pipeline_session,
)
step_create_model = ModelStep(
    name="Create_Model",
    step_args=model.create(instance_type="ml.m5.large"),
)

Expected behavior
A clear and concise description of what you expected to happen.

The ModelPackage model can be used to create sagemaker model successfully.

Screenshots or logs
If applicable, add screenshots or logs to help explain your problem.

System information
A description of your system. Please provide:

• SageMaker Python SDK version:
• Framework name (eg. PyTorch) or algorithm (eg. KMeans):
• Framework version:
• Python version:
• CPU or GPU:
• Custom Docker image (Y/N):

Additional context
Add any other context about the problem here.

[qidewenwhen]

Hi @duk-amz thanks for reaching out!
Sorry that ModelStep was not initially designed to integrate with the ModelPackage as we didn't see such use case previously.
However, as you pointed it out here, we need to re-evaluate this feature. Adding this request to our backlog.

[ShwetaSingh801]

Hi @duk-amz,

Thanks for using SageMaker and taking the time to suggest ways to improve SageMaker Python SDK. We have added your feature request it to our backlog of feature requests and may consider putting it into future SDK versions. I will go ahead and close the issue now, please let me know if you have any more feedback. Let me know if you have any other questions.

Best,
Shweta