Create Model Step fails when when we pass the URL as a Pipeline Parameter for model_data

[rohangpatil]

Describe the bug
A clear and concise description of what the bug is.

To reproduce
The below code works

model = XGBoostModel(
      model_data="s3://<...>/model.tar.gz",
      role=role,
      sagemaker_session=sagemaker_session,
      entry_point="inference_entrypoint.py",
      framework_version="1.5-1",
    )

This doesnt work.

model_url = ParameterString(name="ModelURL")
model = XGBoostModel(
      model_data=model_url,
      role=role,
      sagemaker_session=sagemaker_session,
      entry_point="inference_entrypoint.py",
      framework_version="1.5-1",
    )

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

Screenshots or logs
This is not what is expected.

This is expected

System information
A description of your system. Please provide:

SageMaker Python SDK version: 2.87.0
Framework name (eg. PyTorch) or algorithm (eg. KMeans): XGBoost
Framework version: 1.5-1
Python version: 3.7.10
CPU or GPU: CPU
Custom Docker image (Y/N): N

Additional context
Add any other context about the problem here.

[jessieweiyi]

This issue is similar to #3078. Related to ParameterString.

[jerrypeng7773]

We have a PR under review to fix this, as a workaround, you can give the parameter a default value like this

model_url = ParameterString(name="ModelURL", default="my-model")

and then override this parameter to the actual value while starting the pipeline execution.

Sorry for the inconvenience.

[rohangpatil]

We have a PR under review to fix this, as a workaround, you can give the parameter a default value like this

model_url = ParameterString(name="ModelURL", default="my-model")

and then override this parameter to the actual value while starting the pipeline execution.

Sorry for the inconvenience.

Thanks! Some how this workaround works.

[qidewenwhen]

@rohangpatil The fix has been released. Could you check if it works for you?
Closing this issue and feel free to reopen if you have any questions.