How to save algorithm results from scoring script?

greatcoconut 1

How can I save the results of the algorithm to a file in the scoring script?
I am using AzureML v1 and python sdk.

YutongTie-MSFT 53,766 Reputation points

2023-01-10T00:27:12.833+00:00

Hello @greatcoconut

Do you mean your output of your jobs?

The result will automatically go to the place you designed. When writing changes, we recommend writing files to storage via an Azure Machine Learning dataset with an OutputFileDatasetConfig object. See how to create an OutputFileDatasetConfig. If you have set as above, you need to check your result in your outputFile.

Otherwise, write files to the ./outputs and/or ./logs folder.

Reference doc - https://learn.microsoft.com/en-us/azure/machine-learning/how-to-save-write-experiment-files#where-to-write-files

I personally recommend the first one way, please see where and how you should write your output - https://learn.microsoft.com/en-us/azure/machine-learning/v1/how-to-train-with-datasets#where-to-write-training-output

I hope this helps!

Regards,
Yutong
greatcoconut 1 Reputation point

2023-01-10T12:04:37.597+00:00

Thanks for the answer. I understand how data passing between pipeline components works because I use them elsewhere. My question concerns saving the results directly in the scoring script, e.g. after making a prediction. I know that I can send a request for scoring using REST, receive a response and then save it, but I am asking about saving on the side of the scoring script.

Your answer