从 Azure 机器学习工作区获取所有现有管道

Question

I use azure machine learning services (aml) to run a ml-model. When I go to the GUI of AML I can see all the exisiting piplines, but I can't see how they are scheduled. I need to get all puplished piplines and the belonging meta data.

How I can get information about an existing pipline with the python sdk?

Answer 1

Thanks @Lukas Lötters, Iam converting this to answer so that may help other community members:

So, From azureml.pipeline.core.schedule import ScheduleRecurrence , Schedule.list(ws) to get information about an existing pipeline with the python sdk.

Azure documentation about Schedule Recurrence class and Schedule.list .

Answer 2

once published a pipeline and was able to see from notebook.

Every publish will create a REST endpoint as shown here .

published_pipeline1 = pipeline_run1.publish_pipeline(
     name="Published_Titanic_Pipeline_Notebook",
     description="Titanic_Pipeline_Notebook Published Pipeline Description",
     version="1.0")

Metadata from Azure ML: You can enable Web service schema (OpenAPI specification) in the service and consume it by querying the swagger URL of the service.