-
Notifications
You must be signed in to change notification settings - Fork 1.2k
Closed
Closed
Copy link
Labels
Description
Describe the bug
Tags are't inherited from the sagemaker_config
to hyperparameter tuning requests like they are for other job types.
To reproduce
Can't provide a reprex.
Expected behavior
Tags should be inherited the same way they are for other job types.
Screenshots or logs
Here is an example to show the tags being added for a training job but not for a hyperparameter tuning job:
System information
A description of your system. Please provide:
- SageMaker Python SDK version: 2.224.1
- Framework name (eg. PyTorch) or algorithm (eg. KMeans): Scikit-learn
- Framework version: 0.23-1
- Python version: 3.10.14
- CPU or GPU: CPU
- Custom Docker image (Y/N): N
Additional context
I've traced the issue back from when we call .fit
from a sagemaker.tuner.HyperparameterTuner
(with an estimator
supplied):
- call
self._fit_with_estimator
https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/tuner.py#L1039 - call
_TuningJob.start_new
https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/tuner.py#L1046-L1050 - call
sagemaker_session.create_tuning_job
https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/tuner.py#L2149 - call
sagemaker_session.create_hyper_parameter_tuning_job
https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/session.py#L3238 - call
self.get_tuning_request
https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/session.py#L3225-L3233 - missing the call to add tags from config https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/session.py#L3289-L3291