Theodore Vasiloudis

Currently inference tasks in our SageMaker pipeline generator always get their model input path by a variable [model_output_path](https://github.com/thvasilo/graphstorm/blob/2928c40553477fd5ab330a70bda859a28443a605/sagemaker/pipeline/create_sm_pipeline.py#L574) that gets set during the creation of a training task. We should...

Currently GConstruct does not offer ways to handle missing values for any of its transformations. We'll need to add that to bring it to parity with GSProcessing

Feature request: provide size hint for progress with generator input

1

Hi team, AFAIK when using a generator as input to parallel/delayed, there's no progress being printed with versbose=1 because there's no way to know the length of the generator in...

HyperparameterTuner does not preserve `disable_output_compression` setting of Estimator

2

**Describe the bug** When setting `disable_output_compression=True` in a PyTorch SageMaker Estimator that is used with HyperparameterTuner, the resulting HPO training jobs still compress the model output. This parameter is correctly...

*Issue #, if available:* *Description of changes:* * Add support for creating an HPO step in SM pipeline By submitting this pull request, I confirm that you can use, modify,...

[breaking] force logit output of focal loss to conform to (N, 2) shape for binary classification

1

*Issue #, if available:* https://github.com/awslabs/graphstorm/issues/1242 *Description of changes:* By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your...

Currently GConstruct saves a single transformation in the output gcon config even when the input has multiple columns. We should be saving one transformation per column and apply them individually...

GSConfig is a wrapper object for GraphStorm but it includes code that enforces log level at the application level: https://github.com/awslabs/graphstorm/blob/f0c38bbc2809dbbe2a956a93a31ba216fff13473/python/graphstorm/config/argument.py#L190-L198 The caller program should get to establish the logging level,...