Explanation:
The question asks which class to use to prepare information for submitting a training job in Azure ML SDK v2. The community discussion clearly indicates that the 'command' class is the correct answer, with multiple comments providing supporting evidence. Comment #945837 shows actual code using 'command' to define a training job with code, command, environment, and compute parameters before submitting via MLClient. Comment #964912 explains that commands are used to execute scripts and submit them as jobs. Comment #1142911 clarifies that while MLClient is used to submit the job, the 'command' class is specifically for preparing/configure the run. The other options are incorrect: MLClient (A) is for workspace connection and job submission, not preparation; BuildContext (B) is not a standard Azure ML SDK v2 class for job preparation; EndpointConnection (C) is related to endpoints, not training jobs.
Ultimate access to all questions.
No comments yet.
You are using the Azure Machine Learning SDK v2 for Python and notebooks to train a model. You have created a compute target, an environment, and a training script using Python code.
Which class should you use to submit a training job?
A
MLClient
B
BuildContext
C
EndpointConnection
D
command