-
Notifications
You must be signed in to change notification settings - Fork 1.2k
Description
PySDK Version
- PySDK V2 (2.x)
- PySDK V3 (3.x)
Describe the bug
The feature_processor subpackage is missing from the published sagemaker-mlops wheel on PyPI. The source code exists in the repository under sagemaker-mlops/src/sagemaker/mlops/feature_store/feature_processor/, but the built wheel does not include this directory. This makes it impossible to import feature_processor after installing sagemaker-mlops[feature-processor].
To reproduce
Note, all of these steps were tested on a JupyterLab space in SageMaker AI Studio.
- Install the sagemaker packages:
pip install sagemaker==3.5.0
pip install sagemaker-mlops==1.5.0
pip install sagemaker-mlops[feature-processor]- Restart the Python kernel.
- Attempt to import the feature processor:
from sagemaker.mlops.feature_store.feature_processor import (
feature_processor,
FeatureGroupDataSource,
)- Observe the error:
ModuleNotFoundError: No module named 'sagemaker.mlops.feature_store.feature_processor' - Confirm the directory is missing from the installed package:
ls /opt/conda/lib/python3.12/site-packages/sagemaker/mlops/feature_store/Output shows no feature_processor directory:
__init__.py dataset_builder.py ingestion_manager_pandas.py
__pycache__ feature_definition.py inputs.py
athena_query.py feature_utils.py
Expected behavior
The feature_processor subpackage should be included in the published sagemaker-mlops wheel, and the import should work after installing sagemaker-mlops[feature-processor].
Screenshots or logs
╭──────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ in <module>:1 │
│ │
│ ❱ 1 from sagemaker.mlops.feature_store.feature_processor import ( │
│ 2 │ feature_processor, │
│ 3 │ FeatureGroupDataSource, │
│ 4 │ CSVDataSource, │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
ModuleNotFoundError: No module named 'sagemaker.mlops.feature_store.feature_processor'
System information
- SageMaker Python SDK version: 3.5.0 (sagemaker-mlops 1.5.0)
- Framework name (eg. PyTorch) or algorithm (eg. KMeans): N/A
- Framework version: N/A
- Python version: 3.12
- CPU or GPU: CPU
- Custom Docker image (Y/N): N
Additional context
N/A