แก้ไข

แชร์ผ่าน

CLI (v2) deployment template YAML schema

APPLIES TO: Azure CLI ml extension v2 (current)

The source JSON schema can be found at https://azuremlschemas.azureedge.net/latest/deploymentTemplate.schema.json.

Note

The YAML syntax detailed in this document is based on the JSON schema for the latest version of the ML CLI v2 extension. This syntax is guaranteed only to work with the latest version of the ML CLI v2 extension. You can find the schemas for older extension versions at https://azuremlschemasprod.azureedge.net/.

YAML syntax

Key Type Description Allowed values Default value
$schema string The YAML schema. If you use the Azure Machine Learning VS Code extension to author the YAML file, including $schema at the top of your file enables you to invoke schema and resource completions.
name string Required. Name of the deployment template.
version string or integer Version of the deployment template.
description string Description of the deployment template.
tags object Dictionary of tags for the deployment template.
type string Type of the deployment template.
deployment_template_type string The deployment template type.
environment string or object The environment to use for the deployment template. This value can be either a reference to an existing versioned environment in a registry, or an inline environment specification.

To reference an existing environment in a registry, use the azureml://registries/<registry-name>/environments/<environment-name> syntax.

To define an environment inline, follow the Environment schema.

As a best practice for production scenarios, you should create the environment separately and reference it here.
environment_variables object Dictionary of environment variable key-value pairs to set for the deployment. You can access these environment variables from your scoring scripts.
instance_count integer The number of instances to use for the deployment. Specify the value based on the workload you expect.
default_instance_type string The default instance type to use when deploying with this template.
allowed_instance_types string The allowed instance type that can be used when deploying with this template.
model_mount_path string The path to mount the model in the container.
scoring_path string The path for the scoring endpoint.
scoring_port integer The port for the scoring endpoint.
liveness_probe object Liveness probe settings for monitoring the health of the container regularly. See ProbeSettings for the set of configurable properties.
readiness_probe object Readiness probe settings for validating if the container is ready to serve traffic. See ProbeSettings for the set of configurable properties.
request_settings object Request settings for the deployment. See RequestSettings for the set of configurable properties.

ProbeSettings

Key Type Description Default value
failure_threshold integer When a probe fails, the system tries failure_threshold times before giving up. Giving up in the case of a liveness probe means the container is restarted. In the case of a readiness probe, the container is marked Unready. Minimum value is 1. 30
initial_delay integer The number of seconds after the container has started before the probe is initiated. Minimum value is 1. 10
method string The HTTP method to use for the probe.
path string The path for the probe.
period integer How often (in seconds) to perform the probe. 10
port integer The port to probe.
scheme string The scheme to use for the probe (for example, HTTP or HTTPS).
success_threshold integer The minimum consecutive successes for the probe to be considered successful after having failed. Minimum value is 1. 1
timeout integer The number of seconds after which the probe times out. Minimum value is 1. 2

RequestSettings

Key Type Description Default value
request_timeout_ms integer The request timeout in milliseconds. 5000
max_concurrent_requests_per_instance integer The maximum number of concurrent requests per instance allowed for the deployment. 1

Remarks

Deployment templates provide a reusable configuration for deploying models. They define the environment, infrastructure settings, and probe configurations that can be applied when creating deployments.

Examples

Examples are shown below.

YAML: basic

$schema: https://azuremlschemas.azureedge.net/latest/deploymentTemplate.schema.json
name: my-deployment-template
version: 1
description: Basic deployment template example
environment: azureml:my-environment:1
instance_count: 1
default_instance_type: Standard_DS3_v2

YAML: with environment variables and probes

$schema: https://azuremlschemas.azureedge.net/latest/deploymentTemplate.schema.json
name: my-deployment-template
version: 1
description: Deployment template with environment variables and health probes
environment: azureml://registries/azureml/environments/minimal-ubuntu20.04-py38-cpu-inference:latest
environment_variables:
  MODEL_PATH: /var/azureml-app/model
  SCORING_TIMEOUT: "60"
instance_count: 3
default_instance_type: Standard_DS3_v2
scoring_path: /score
scoring_port: 8080
liveness_probe:
  initial_delay: 30
  period: 10
  timeout: 2
  success_threshold: 1
  failure_threshold: 3
readiness_probe:
  initial_delay: 10
  period: 5
  timeout: 2
  success_threshold: 1
  failure_threshold: 3
request_settings:
  request_timeout_ms: 10000
  max_concurrent_requests_per_instance: 2

YAML: with allowed instance type

$schema: https://azuremlschemas.azureedge.net/latest/deploymentTemplate.schema.json
name: my-deployment-template-restricted
version: 1
description: Deployment template with instance type restrictions
environment: azureml:my-environment:1
instance_count: 1
default_instance_type: Standard_DS3_v2
allowed_instance_types: Standard_DS3_v2

YAML: with inline environment

$schema: https://azuremlschemas.azureedge.net/latest/deploymentTemplate.schema.json
name: my-deployment-template-inline-env
version: 1
description: Deployment template with inline environment definition
environment:
  name: inline-environment
  image: mcr.microsoft.com/azureml/minimal-ubuntu20.04-py38-cpu-inference:latest
  inference_config:
    liveness_route:
      path: /health
      port: 5001
    readiness_route:
      path: /ready
      port: 5001
    scoring_route:
      path: /score
      port: 5001
instance_count: 1
default_instance_type: Standard_DS3_v2

YAML: with model mount path

$schema: https://azuremlschemas.azureedge.net/latest/deploymentTemplate.schema.json
name: my-deployment-template-custom-mount
version: 1
description: Deployment template with custom model mount path
environment: azureml:my-environment:1
instance_count: 1
default_instance_type: Standard_DS3_v2
model_mount_path: /var/azureml-app/models

Next steps

  • Last updated on