SageMaker Processing jobs have a less informative config than Training jobs

Hi all,

SageMaker integrates fairly well with wandb through SageMaker | Weights & Biases Documentation. However, whereas training jobs automatically add all arguments (i.e. “hyperparameters” in an Estimator Estimators — sagemaker 2.151.0 documentation) and some extra information (such as the sagemaker_training_job_name, sagemaker_region) to the wandb config, processing jobs do not. I could of course manually add these to my processing job wandb runs, but would rather have it handled through wandb seamless API.

Is there an intention to close this gap?

PS this post was flagged as ‘spam’ because it is considered an ‘advertisement’. Please let me know why. How can I advertise something when it is not even working? Furthermore, I am only providing relevant facts, I am running into a (minor) limitation, and there is no customer service.

Thanks a lot @laut I have proceeded with a feature request on this, as it would be very helpful indeed. We will reach out here once there are any updates.

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.