Azure ML integration

This documents outlines the integrations between Olive and Azure Machine Learning. Discover how to use your Azure Machine Learning assets within Olive.

Azure Machine Learning client

To use Azure Machine Learning assets in your workspace, please include your workspace information as the azureml_client in configuration JSON file. Please find more details here

Olive Core

Using AzureML registered model

You can run Olive workflow with your AML workspace registered model. In the input model section, define the model config as:

"model_path": {
    "type": "azureml_model",
    "name": "<model_name>",
    "version": "<model_version>"
 }

Olive will automatically download the model and run the workflow in the specified target or host with this model as input model.

Using AzureML curated model

You can run Olive workflow with AML registered model. In the input model section, define the model config as:

"model_path": {
    "type": "azureml_registry_model",
    "name": "model_name",
    "registry_name": "registry_name",
    "version": 1
}

Olive will automatically download the model and run the workflow in the specified target or host with this model as input model.

Note: you don’t need the azureml_client section for AzureML curated model.

Using model stored in AzureML datastore

You can specify your model path from an AzureML datastore as:

"model_path": {
    "type": "azureml_datastore",
    "azureml_client": {
        "subscription_id": "my_subscription_id",
        "resource_group": "my_resource_group",
        "workspace_name": "my_workspace"
    },
    "datastore_name": "my_datastore",
    "relative_path": "model_dir/my_model.pt" // Relative path to the resource from the datastore root
}

Using a model from an AzureML job output

You can specify your model path from an AzureML job output as:

"model_path": {
    "type": "azureml_job_output",
    "azureml_client": {
        "subscription_id": "my_subscription_id",
        "resource_group": "my_resource_group",
        "workspace_name": "my_workspace"
    },
    "job_id": "my_job_id", // id of the job
    "output_name": "my_output_name", // name of the job output
    "relative_path": "model_dir/my_model.pt" // Relative path to the resource from the job output root
}

Using data stored in AzureML datastore

You can use data files or folders that are stored in your Azure ML datastore as:

"data_dir": {
    "type": "azureml_datastore",
    "azureml_client": {
        "subscription_id": "my_subscription_id",
        "resource_group": "my_resource_group",
        "workspace_name": "my_workspace"
    },
    "datastore_name": "my_datastore",
    "relative_path": "data_dir" // Relative path to the resource from the datastore root
}

Using Azure ML compute as host or target

You can specify your Azure ML Compute as an Olive System and use it as a host to run Pass, or a target to evaluate the model.

"systems": {
    "aml_system": {
        "type": "AzureML",
        "aml_compute": "cpu-cluster",
        "aml_docker_config": {
            "base_image": "mcr.microsoft.com/azureml/openmpi4.1.0-ubuntu20.04",
            "conda_file_path": "conda.yaml"
        }
    }
}

Then you can specify where to use it in the Engine config:

{
    "host": "aml_system",
    "target": "aml_system",
}

Packaging the output models to AzureML workspace

Azure ML Models

AzureMLModels packaging will register the output models to your Azure Machine Learning workspace. Please find more details here.

Azure ML Data

AzureMLData packaging will upload the output models to your Azure Machine Learning workspace as Data assets. Please find more details here.

Connect your own device to Azure ML as target or host by Azure Arc

If you have your own device, you have the option to link it to your Azure ML Workspace as a Compute via Azure Arc. This allows you to use it as a Target for evaluating your model.

Please follow this instruction to setup your local device: Self-hosted Kubernetes cluster

Azure ML Helper Scripts

Olive offers several scripts to assist you in managing your Azure ML assets.

For details on the available scripts, refer to: Azure ML scripts