suggestion is to add default tag to the run, github repo , to identify which run was sent from which repo.

even preferable is to also add the github_SHA, so that it can be uniquely identified with code changes.

Move labels from the Dockerfile to the action.yml to speed up the docker build.

• Support yaml pipelines

Add support for runconfig

if user gives wrong "runconfig_python_file", or the path is wrong , the code throws below error

this error is not explicit why issue has come and looks like the code inherently threw an error. while actually there was in exception during module import execution

• Add default pipeline name, when pipeline should be registered

Feature brainstorming for GH integration

• Pass model back to artifact in actions
• Make a release flag so that the model will be made into a release automatically.
• Create a docker image of the model

Traceback (most recent call last):
File "/usr/local/lib/python3.8/site-packages/dotnetcore2/runtime.py", line 271, in attempt_get_deps
blob_deps_to_file()
File "/usr/local/lib/python3.8/site-packages/dotnetcore2/runtime.py", line 263, in blob_deps_to_file
blob = request.urlopen(deps_url, context=ssl_context)
File "/usr/local/lib/python3.8/urllib/request.py", line 222, in urlopen
return opener.open(url, data, timeout)
File "/usr/local/lib/python3.8/urllib/request.py", line 531, in open
response = meth(req, response)
File "/usr/local/lib/python3.8/urllib/request.py", line 640, in http_response
response = self.parent.error(
File "/usr/local/lib/python3.8/urllib/request.py", line 569, in error
return self._call_chain(*args)
File "/usr/local/lib/python3.8/urllib/request.py", line 502, in _call_chain
result = func(*args)
File "/usr/local/lib/python3.8/urllib/request.py", line 649, in http_error_default
raise HTTPError(req.full_url, code, msg, hdrs, fp)
urllib.error.HTTPError: HTTP Error 404: Not Found
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/code/main.py", line 240, in
main()
File "/code/main.py", line 187, in main
run.wait_for_completion(show_output=True)
File "/usr/local/lib/python3.8/site-packages/azureml/pipeline/core/run.py", line 294, in wait_for_completion
step_run.wait_for_completion(timeout_seconds=timeout_seconds - time_elapsed,
File "/usr/local/lib/python3.8/site-packages/azureml/pipeline/core/run.py", line 736, in wait_for_completion
return self._stream_run_output(timeout_seconds=timeout_seconds,
File "/usr/local/lib/python3.8/site-packages/azureml/pipeline/core/run.py", line 827, in _stream_run_output
print(final_details)
File "/usr/local/lib/python3.8/site-packages/azureml/data/_loggerfactory.py", line 129, in wrapper
return func(*args, **kwargs)
File "/usr/local/lib/python3.8/site-packages/azureml/data/abstract_dataset.py", line 766, in repr
steps = self._dataflow._get_steps()
File "/usr/local/lib/python3.8/site-packages/azureml/data/_loggerfactory.py", line 129, in wrapper
return func(*args, **kwargs)
File "/usr/local/lib/python3.8/site-packages/azureml/data/abstract_dataset.py", line 218, in _dataflow
dataprep().api._datastore_helper._set_auth_type(self._registration.workspace)
File "/usr/local/lib/python3.8/site-packages/azureml/dataprep/api/_datastore_helper.py", line 185, in _set_auth_type
get_engine_api().set_aml_auth(SetAmlAuthMessageArgument(auth_type, json.dumps(auth_value)))
File "/usr/local/lib/python3.8/site-packages/azureml/dataprep/api/engineapi/api.py", line 19, in get_engine_api
_engine_api = EngineAPI()
File "/usr/local/lib/python3.8/site-packages/azureml/dataprep/api/engineapi/api.py", line 110, in init
self._message_channel = launch_engine()
File "/usr/local/lib/python3.8/site-packages/azureml/dataprep/api/engineapi/engine.py", line 333, in launch_engine
dependencies_path = runtime.ensure_dependencies()
File "/usr/local/lib/python3.8/site-packages/dotnetcore2/runtime.py", line 285, in ensure_dependencies
if not attempt_get_deps():
File "/usr/local/lib/python3.8/site-packages/dotnetcore2/runtime.py", line 279, in attempt_get_deps
raise NotImplementedError(err_msg + '\n' + _unsupported_help_msg)
NotImplementedError: Linux distribution debian 11. does not have automatic support.
.NET Core 2.1 can still be used via dotnetcore2 if the required dependencies are installed.
Visit https://aka.ms/dotnet-install-linux for Linux distro specific .NET Core install instructions.
Follow your distro specific instructions to install dotnet-runtime-* and replace * with 2.1.

• allow multiple run submissions
• allow multiple pipeline subscriptions

I am using github workflows and aml-run action, I am setting a custom variables using run.log("my_age", 10) and want to read value of this variable on .yml file for next step. How can I achieve it? I see {{outputs.run_metrics}} gives me this custom value but I am unable to pick it.

If someone submits a pipeline to Azure Machine Learning, the convert_to_markdown function can receive nested dicts. The function cannot handle this and fails. This needs to be fixed.

Sample dict that needs to be parsed:
{'HD_9cbf07dc-db35-4700-adc3-92253a378992': {'best_child_by_primary_metric': {'metric_name': ['mse',
'mse',
'mse'],
'timestamp': ['2020-03-30 07:58:40.110108+00:00',
'2020-03-30 07:59:42.417631+00:00',
'2020-03-30 07:59:42.417631+00:00'],
'run_id': ['HD_9cbf07dc-db35-4700-adc3-92253a378992_2',
'HD_9cbf07dc-db35-4700-adc3-92253a378992_1',
'HD_9cbf07dc-db35-4700-adc3-92253a378992_1'],
'metric_value': [0.4896792910300023,
0.33284572706891424,
0.33284572706891424],
'final': [False, False, True]}},
'HD_9cbf07dc-db35-4700-adc3-92253a378992_0': {'AUC': 0.6562945178135493,
'RMSE': 0.05120842417634075,
'pAUC': 0.2812047761829226,
'mse': 0.9367719704449606,
'TimeSeries comparison': 'aml://artifactId/ExperimentRun/dcid.HD_9cbf07dc-db35-4700-adc3-92253a378992_0/TimeSeries comparison_1585555145.png'},
'HD_9cbf07dc-db35-4700-adc3-92253a378992_1': {'AUC': 0.3573103979675637,
'pAUC': 0.5616569664904285,
'RMSE': 0.6161010679565609,
'mse': 0.33284572706891424,
'TimeSeries comparison': 'aml://artifactId/ExperimentRun/dcid.HD_9cbf07dc-db35-4700-adc3-92253a378992_1/TimeSeries comparison_1585555146.png'},
'HD_9cbf07dc-db35-4700-adc3-92253a378992_2': {'AUC': 0.1200601984180878,
'pAUC': 0.5805687333099157,
'RMSE': 0.09691275977364577,
'mse': 0.4896792910300023,
'TimeSeries comparison': 'aml://artifactId/ExperimentRun/dcid.HD_9cbf07dc-db35-4700-adc3-92253a378992_2/TimeSeries comparison_1585555076.png'}}

I have two different pipelines in AzureML that I would like to trigger if any files related to them is changed in my repository. If there's only been changes to the files associated with one of the pipelines I only want to trigger that one and now the other. I've implemented this logic in a bash script that runs as a step in my github workflow and saves which of the pipelines to run to an environment variable.

My plan was then to make the aml-run action trigger a python script in azure-ml that will read in this environment variable and trigger the right pipelines. My problem is that if I set the environment variable in the Azure/aml-run@v1 step then it's not propagated to the run in azure-ml and I cannot retrieve it within the python script.

Is there any way I could get this variable into the python script I have defined in run_config.yml? I can see that there are arguments for setting both environment variables as well as arguments in the run_config.yml but would it be possible to change this based on what I send in to the aml-run action?

I'm using e1efa55 with azureml-sdk 1.21.0 and HyperDriveConfig, and getting

Traceback (most recent call last):
  File "/code/main.py", line 238, in <module>
    main()
  File "/code/main.py", line 189, in main
    run_metrics = run.get_metrics(recursive=True)
TypeError: get_metrics() got an unexpected keyword argument 'recursive'

Remove recursive=True marcindulak#1 as a workaround allows the run to finish.

If you are not using any kind of metric logging, using the run context from azureml sdk, this action crashes, because of an empty metrics list.

Following is my training script:

• (should be *.py) test_train.txt
• You can see that there is only the assignment of run context in line 20
• no other use of run.log(...) or others
• adding run.log(...) only one time, the pipeline completes the step successfully

• Mention that AutoML works too

If the user uses mlflow for tracking the markdown conversion fails.
This is something that needs to be addressed asap to reenable the feature for our advanced template.
@awmatheson and @ashishonce Can you please look into it?
I can help you setting this up.
This is related to the issue mentioned here Azure/aml-workspace#14

• Rename folders if required, so that the import works fine

The problem is also present in other aml-* github actions.

Hi,

I am trying to utilize this GitHub Action for a simple CI pipeline.
Unfortunately it offers no helpful traceback when it fails. There seems to be an issue when I tray to extend the runconfig_python_file.
I am currently using the runconfig_python_file from the test folder an simply adding a few lines of code. One thing I tried to do, was reading a yaml file to make the script more configurable.

Unfortunately then the Action fails already when I just want to import the yaml library without any helpful error information.

Script:

import os
import yaml

from azureml.core import (
    Workspace,
    Experiment,
    ComputeTarget,
    Environment,
    ScriptRunConfig,
)

def main(workspace):
    ws = workspace
    print("Workspace Name: \t{}".format(ws.name))
    print("Resource Group: \t{}".format(ws.resource_group))
    print("Location: \t\t{}".format(ws.location))
    print("Subscription ID: \t{}".format(ws.subscription_id))

    compute_name = 'my-compute'
    compute_target = ws.compute_targets[compute_name]
    print("Compute Cluster Name: \t{}".format(compute_name))

    print("Loading Environment")
    my_env= Environment.from_conda_specification(
        name="my_env",
        file_path="code/environment.yml"
    )

    tasks_source_dir =  '/tasks' 

    print("Loading script parameters")
    script_args = [
        "--kernel", "linear",
        "--penalty", 1.0
    ]
    
    print("Creating run config")
    run_config = ScriptRunConfig(
        source_directory=tasks_source_dir,
        script="T01_Test_Task.py",
        arguments=script_args,
        run_config="",
        compute_target=compute_target,
        environment=my_env
    )

    return run_config

GitHub Action YAML:

# Actions train a model on Azure Machine Learning
name: Continous Integration
on:
  push:
    branches:
      - dev
    # paths:
    #   - 'code/*'
jobs:
  train:
    runs-on: ubuntu-latest
    steps:
    # Checks-out your repository under $GITHUB_WORKSPACE, so your job can access it
    - name: Check Out Repository
      id: checkout_repository
      uses: actions/checkout@v2
      
    - name: Python Set Up
      uses: actions/setup-python@v4
      with:
        python-version: '3.8'
        # cache: 'pip'
    - run: pip install -r requirements.txt  # this requirements.txt contains pyyaml
    
    # Connect or Create the Azure Machine Learning Workspace
    - name: Connect/Create Azure Machine Learning Workspace
      id: aml_workspace
      uses: Azure/aml-workspace@v1
      with:
          azure_credentials: ${{ secrets.AZURE_CREDENTIALS }}
    
    # Connect or Create a Compute Target in Azure Machine Learning
    - name: Connect/Create Azure Machine Learning Compute Target
      id: aml_compute_training
      uses: Azure/aml-compute@v1
      with:
          azure_credentials: ${{ secrets.AZURE_CREDENTIALS }}
    
    # Submit a training run to the Azure Machine Learning
    - name: Submit training run
      id: aml_run
      uses: Azure/aml-run@v1
      with:
          azure_credentials: ${{ secrets.AZURE_CREDENTIALS }}
          parameters_file: "run.json"

Traceback (most recent call last):
  File "/code/main.py", line 240, in <module>
    main()
	Message: Failed to load RunConfiguration from path=code/train/run_config.yml name=None
	InnerException None
	ErrorResponse 
***
    "error": ***
        "code": "UserError",
        "message": "Failed to load RunConfiguration from path=code/train/run_config.yml name=None"
    ***
***
None
Error: Error when loading runconfig yaml definition your repository (Path: /code/train/run_config.yml).
Error: Error when loading pipeline yaml definition your repository (Path: /code/train/pipeline.yml).
Error: Error when loading python script or function in your repository which defines the experiment config (Script path: '/code/main.py', Function: 'main()').
Error: You have to provide either a yaml definition for your run, a yaml definition of your pipeline or a python script, which returns a runconfig (Pipeline, ScriptRunConfig, AutoMlConfig, Estimator, etc.). Please read the documentation for more details.
  File "/code/main.py", line [15](https://github.com/bayer-int/ch_daa_phinmo/runs/7499494632?check_suite_focus=true#step:10:16)3, in main
    raise AMLExperimentConfigurationException("You have to provide a yaml definition for your run, a yaml definition of your pipeline or a python script, which returns a runconfig. Please read the documentation for more details.")
utils.AMLExperimentConfigurationException: You have to provide a yaml definition for your run, a yaml definition of your pipeline or a python script, which returns a runconfig. Please read the documentation for more details.

problem is with integrationworkflow.yml , azureCredential Input has to be azure_credential as action input has an "_"

• Pass model back to artifact in actions

Add more detailed logging and info to the user what was submitted and what failed.