aws / aws/sagemaker-python-sdk
ModelBuilder.register does not generate a repack step with a pipeline session in V3 but does in V2
- Dominant language
- Python
- Stars
- 2.3k
- Forks
- 1.3k
- Avg merge
- 1d 22h
- Merged PRs (30d)
- 35
Description
**PySDK Version**
- [ ] PySDK V2 (2.x)
- [ X] PySDK V3 (3.x)
**Describe the bug**
ModelBuilder.register does not generate a repack step with a pipeline session in V3 but does in V2
**To reproduce**
A clear, step-by-step set of instructions to reproduce the bug.
The provided code need to be **complete** and **runnable**, if additional data is needed, please include them in the issue.
```
model_builder = ModelBuilder(
image_uri=image_uri,
s3_model_data_url=data_url,
source_code=SourceCode(
entry_script="infer.py",
source_dir=s3_script_archive,
),
role_arn=role,
sagemaker_session=pipeline_session,
env_vars={"SAGEMAKER_DEFAULT_INVOCATIONS_ACCEPT": "application/json"},
)
...
step_register = ModelStep(
name="RegisterModel",
step_args=model_builder.register(
content_types=["application/jsonlines"],
response_types=["application/json"],
inference_instances=[inference_instance_type],
transform_instances=[inference_instance_type],
model_package_group_name=model_package_group_name,
approval_status="Approved",
)
)
```
**Expected behavior**
A clear and concise description of what you expected to happen.
In V2, the model was repacked with the code present. In V3, the model is not, leaving the model unable to be repacked
**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**: 2
- **Framework name (eg. PyTorch) or algorithm (eg. KMeans)**: ScikitLearn
- **Framework version**: 1.4.2
- **Python version**: 3.11
- **CPU or GPU**: CPU
- **Custom Docker image (Y/N)**: N
**Additional context**
Add any other context about the problem here.
Contributor guide
Research direction
Start at the ModelBuilder.register entry point used with pipeline_session and compare its V3 behavior with the V2 behavior described in the issue. Reproduce the ModelStep registration using the shown SourceCode and model data, then verify that a repack step is generated and the registered model includes the code.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- aws, python, scikit-learn
- Domain
- cloud, machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 1-2 days
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 48/100