This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(70.4, 96%, 16%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EBJ%3C/text%3E%3C/svg%3E)
Hi,
I am using the PythonSDK with azureml-core==1.46.0 to submit an experiment with Experiment(workspace, experiment_name).submit(pipeline, pipeline_parameters). In this Run I am training and registering a model. I would like to link the Run/Job in this registered Model. The AzureML-UI has already a column prepared for this:
I use Model.register(workspace, model_path, model_name, ...). Is there a way to link this Job/Run in the register-step?
Thanks
Jan
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(22.400000000000002, 11%, 12%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ERM%3C/text%3E%3C/svg%3E)
@Berger Jan (ITKG/EPM-ISY1.1) Thanks for the question, you can link the Run/Job in the registered Model. Run.register_model() will allow you to see the Experiment name and the Run ID in the Azure Machine Learning Studio.
Here is the sample for managing the Runs.
Thanks for the quick answer!
I tried your proposal but get an error saying it can not find the model. To add more context: I store several files in a folder "outputs". I tried to upload the folder to the run with Run.get_context().upload_folder("outputs", target) but get an error as well saying it is not a valid folder.
Do you have any more tipps for this?
Can you please add more details about the use case and steps that you performed. Here are steps how to dynamically create data sets from PipelineParameters".
Suppose the parameter values are datastore_name and relative_path, user could call Dataset.File.from_files to get the data from path first, then dataset._ensure_save to save it to workspace. about a pipeline includes two steps, one to create data, second to show the data path, you may have to replace the relative path to a real one to run the pipeline successfully.
Thanks for the reply. I fixed the problem with uploading the folder and can register the model now from the Run. The Run Id of the registered model now points to the PythonStep where I called this code (register step). Even though I would have prefered a direct link to the Pipeline Run Id, it still is linked as a parent and therefore works for me.