Rediger

Del via

CLI (v2) Spark component YAML schema

  • Artikkel

APPLIES TO: Azure CLI ml extension v2 (current)

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.
type const Required. The type of component. spark
name string Required. Name of the component. Must start with lowercase letter. Allowed characters are lowercase letters, numbers, and underscore(_). Maximum length is 255 characters.
version string Version of the component. If omitted, Azure Machine Learning autogenerates a version.
display_name string Display name of the component in the studio UI. Can be nonunique within the workspace.
description string Description of the component.
tags object Dictionary of tags for the component.
code string Required. The location of the folder that contains source code and scripts for the component.
entry object Required. The entry point for the component. It could define a file.
entry.file string The location of the folder that contains source code and scripts for the component.
py_files object A list of .zip, .egg, or .py files, to be placed in the PYTHONPATH, for successful execution of the job with this component.
jars object A list of .jar files to include on the Spark driver, and the executor CLASSPATH, for successful execution of the job with this component.
files object A list of files that should be copied to the working directory of each executor, for successful execution of the job with this component.
archives object A list of archives that should be extracted into the working directory of each executor, for successful execution of the job with this component.
conf object The Spark driver and executor properties. See Attributes of the conf key
environment string or object The environment to use for the component. This value can be either a reference to an existing versioned environment in the workspace or an inline environment specification.

To reference an existing environment, use the azureml:<environment_name>:<environment_version> syntax or azureml:<environment_name>@latest (to reference the latest version of an environment).

To define an environment inline, follow the Environment schema. Exclude the name and version properties, because inline environments don't support them.
args string The command line arguments that should be passed to the component entry point Python script. These arguments may contain the paths of input data and the location to write the output, for example "--input_data ${{inputs.<input_name>}} --output_path ${{outputs.<output_name>}}"
inputs object Dictionary of component inputs. The key is a name for the input within the context of the component and the value is the input value.

Inputs can be referenced in the args using the ${{ inputs.<input_name> }} expression.
inputs.<input_name> number, integer, boolean, string or object One of a literal value (of type number, integer, boolean, or string) or an object containing a component input data specification.
outputs object Dictionary of output configurations of the component. The key is a name for the output within the context of the component and the value is the output configuration.

Outputs can be referenced in the args using the ${{ outputs.<output_name> }} expression.
outputs.<output_name> object The Spark component output. Output for a Spark component can be written to either a file or a folder location by providing an object containing the component output specification.

Attributes of the conf key

Key Type Description Default value
spark.driver.cores integer The number of cores for the Spark driver.
spark.driver.memory string Allocated memory for the Spark driver, in gigabytes (GB), for example, 2g.
spark.executor.cores integer The number of cores for the Spark executor.
spark.executor.memory string Allocated memory for the Spark executor, in gigabytes (GB), for example 2g.
spark.dynamicAllocation.enabled boolean Whether or not executors should be dynamically allocated as a True or False value. If this property is set True, define spark.dynamicAllocation.minExecutors and spark.dynamicAllocation.maxExecutors. If this property is set to False, define spark.executor.instances. False
spark.dynamicAllocation.minExecutors integer The minimum number of Spark executors instances, for dynamic allocation.
spark.dynamicAllocation.maxExecutors integer The maximum number of Spark executors instances, for dynamic allocation.
spark.executor.instances integer The number of Spark executor instances.

Component inputs

Key Type Description Allowed values Default value
type string The type of component input. Specify uri_file for input data that points to a single file source, or uri_folder for input data that points to a folder source. Learn more about data access. uri_file, uri_folder
mode string Mode of how the data should be delivered to the compute target. The direct mode passes in the URL of the storage location as the component input. You have full responsibility to handle storage access credentials. direct

Component outputs

Key Type Description Allowed values Default value
type string The type of component output. uri_file, uri_folder
mode string The mode of delivery of the output file(s) to the destination storage resource. direct

Remarks

The az ml component commands can be used for managing Azure Machine Learning Spark component.

Examples

Examples are available in the examples GitHub repository. Several are shown next.

YAML: A sample Spark component

# spark-job-component.yaml
$schema: https://azuremlschemas.azureedge.net/latest/sparkComponent.schema.json
name: titanic_spark_component
type: spark
version: 1
display_name: Titanic-Spark-Component
description: Spark component for Titanic data

code: ./src
entry:
  file: titanic.py

inputs:
  titanic_data:
    type: uri_file
    mode: direct

outputs:
  wrangled_data:
    type: uri_folder
    mode: direct

args: >-
  --titanic_data ${{inputs.titanic_data}}
  --wrangled_data ${{outputs.wrangled_data}}

conf:
  spark.driver.cores: 1
  spark.driver.memory: 2g
  spark.executor.cores: 2
  spark.executor.memory: 2g
  spark.dynamicAllocation.enabled: True
  spark.dynamicAllocation.minExecutors: 1
  spark.dynamicAllocation.maxExecutors: 4

YAML: A sample pipeline job with a Spark component

# attached-spark-pipeline-user-identity.yaml
$schema: https://azuremlschemas.azureedge.net/latest/pipelineJob.schema.json
type: pipeline
display_name: Titanic-Spark-CLI-Pipeline-2
description: Spark component for Titanic data in Pipeline

jobs:
  spark_job:
    type: spark
    component: ./spark-job-component.yml
    inputs:
      titanic_data: 
        type: uri_file
        path: azureml://datastores/workspaceblobstore/paths/data/titanic.csv
        mode: direct

    outputs:
      wrangled_data:
        type: uri_folder
        path: azureml://datastores/workspaceblobstore/paths/data/wrangled/
        mode: direct

    identity:
      type: user_identity

    compute: <ATTACHED_SPARK_POOL_NAME>

Next steps