The task used in install-terraform.yml file on line 5 is causing the ambiguous error below.
To correct that is needed to change TerraformInstaller@0 for ms-devlabs.custom-terraform-tasks.custom-terraform-installer-task.TerraformInstaller@0

Thank you!

Thanks for creating this repo. We are using it for Azure MLOps pipelining of deep learning jobs.

We keep getting this error that our jobs are cancelling after 60 minutes.

Could you guide which part of your pipeline we should modify so that it gets fixed?

I am using PyTorch DistributedDataParallel for the training.

I tried adding the timeout variables to pipeline.yaml from you but somehow it is not in the schema also I am not sure if that would fix the problem.

Here's an overview of using MLOps Templates from you for our code:

Please note while we get a BUILD FAILED email the job in the pipeline keeps running. I also had it once finish the training and save the model checkpoints but it seems it might be producing wrong results (perhaps because of same error).

Any help you could provide in this respect is really appreciated.
Here's the link provided in BUILD FAILED email: https://learn.microsoft.com/en-us/azure/devops/pipelines/process/phases?view=azure-devops&tabs=yaml#timeouts

Here's the link to pipeline job schema: https://azuremlschemas.azureedge.net/latest/pipelineJob.schema.json

Currently this workload wants to reference Service Principal, but the newish standard is to actually use OpenID Connect (OIDC) and managed identity. https://github.com/Azure/login#login-with-openid-connect-oidc-recommended

Currently the MLOps_v2 calls back for

secrets:
      creds:

Doesn't allow you to use OIDC

          client-id: ${{ secrets.AZURE_CLIENT_ID }}
          tenant-id: ${{ secrets.AZURE_TENANT_ID }}
          subscription-id: ${{ secrets.AZURE_SUBSCRIPTION_ID }}

Includes:

• Registering additional datasets other than "training"/"scoring".
• Passing datasets definitions as separate training parameter

Additional argument to specify type of environment definition that supports folder & Dockerfile, appart from conda.

There are important files that Microsoft projects should all have that are not present in this repository. A pull request has been opened to add the missing file(s). When the pr is merged this issue will be closed automatically.

Microsoft teams can learn more about this effort and share feedback within the open source guidance available internally.

Merge this pull request

Getting the following error in connect-to-workspace.yml:

AuthenticationException: AzureCliAuthentication requires azure-cli>=2.30.0 to be installed in the same python environment where azureml-sdk is installed.

This repository is currently missing a LICENSE.MD file outlining its license. A license helps users understand how to use your project in a compliant manner. You can find the standard MIT license text at the Microsoft repo templates LICENSE file: https://github.com/microsoft/repo-templates/blob/main/shared/LICENSE.
If you would like to learn more about open source licenses, please visit the document at https://aka.ms/license.

tf-ado-deploy-infra.yml from mlops-project-template asks for create-sp-variables.yml, but that file does not exist in templates/infra.