Using Local Computer as a Compute Target for Azure AutoML in ML Studio

Question

Is it possible to connect a local computer as a compute target to an Azure Machine Learning workspace and use it as a compute target for AutoML within Azure ML Studio, instead of using a compute cluster? I tried going through the documentation but could not find anything useful. could someone please guide me through the process or share code snippets/any relevant documentation please?

Thanks in advance!

Answer 1

Ramasamy Seenivasagan Greetings!

Is it possible to connect a local computer as a compute target to an Azure Machine Learning workspace and use it as a compute target for AutoML within Azure ML Studio, instead of using a compute cluster?

Yes, it is possible to connect a local computer as a compute target to an Azure Machine Learning workspace and use it for AutoML.

You can use any of the following resources for a training compute target for most jobs. Not all resources can be used for automated machine learning, machine learning pipelines, or designer.

You can select the compute target where your training script will run on. If no compute target is specified in the ScriptRunConfig, or if compute_target='local', Azure Machine Learning will execute your script locally. See Configure and submit training jobs.

Also, note that if your compute target is your local machine, you're responsible for ensuring that all the necessary packages are available in the Python environment where the script runs.

I would suggest you, check Manage compute resources for model training and deployment in studio and Set up a Python development environment for Azure Machine Learning for more details.

Do let me know if that helps or have any other queries.

If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further queries do let us know.