ScriptRunConfig Class

Reference

Represents configuration information for submitting a training run in Azure Machine Learning.

A ScriptRunConfig packages together the configuration information needed to submit a run in Azure ML, including the script, compute target, environment, and any distributed job-specific configs.

Once a script run is configured and submitted with the submit, a ScriptRun is returned.

Class ScriptRunConfig constructor.

Inheritance: azureml._logging.chained_identity.ChainedIdentity

ScriptRunConfig

Constructor

ScriptRunConfig(source_directory, script=None, arguments=None, run_config=None, _telemetry_values=None, compute_target=None, environment=None, distributed_job_config=None, resume_from=None, max_run_duration_seconds=2592000, command=None, docker_runtime_config=None)

Parameters

Name	Description
source_directory Required	str A local directory containing code files needed for a run.
script Required	str The file path relative to the source_directory of the script to be run.
arguments Required	list or str Optional command line arguments to pass to the training script. Arguments are passed in pairs, for example, ['–arg1', arg1_val, '–arg2', arg2_val].
run_config Required	RunConfiguration Optional run configuration to use.
_telemetry_values Required	dict Internal use only.
compute_target Required	AbstractComputeTarget or str The compute target where training will happen. This can either be a ComputeTarget object, the name of an existing ComputeTarget, or the string "local". If no compute target is specified, your local machine will be used.
environment Required	Environment The environment to use for the run. If no environment is specified, azureml.core.runconfig.DEFAULT_CPU_IMAGE will be used as the Docker image for the run.
distributed_job_config Required	TensorflowConfiguration, MpiConfiguration or PyTorchConfiguration For jobs that require additional distributed job-specific configurations.
resume_from Required	DataPath The DataPath containing the checkpoint or model files from which to resume the experiment.
max_run_duration_seconds Required	The maximum time allowed for the run. The system will attempt to automatically cancel the run if it took longer than this value. :type max_run_duration_seconds: int
command Required	list[str] or str The command to be submitted for the run. The command property can also be used instead of script/arguments. Both command and script/argument properties cannot be used together to submit a run. To submit a script file using the command property - ['python', 'train.py', '–arg1', arg1_val] To run an actual command - ['ls']
docker_runtime_config Required	DockerConfiguration For jobs that require Docker runtime-specific configurations.
source_directory Required	str A local directory containing code files needed for a run.
script Required	str The file path relative to the source_directory of the script to be run.
arguments Required	list[str] Optional command line arguments to pass to the training script. Arguments are passed in pairs, for example, ['–arg1', arg1_val, '–arg2', arg2_val].
run_config Required	RunConfiguration Optional run configuration to use.
_telemetry_values Required	dict Internal use only.
compute_target Required	AbstractComputeTarget or str The compute target where training will happen. This can either be a ComputeTarget object, the name of an existing ComputeTarget, or the string "local". If no compute target is specified, your local machine will be used.
environment Required	Environment The environment to use for the run. If no environment is specified, azureml.core.runconfig.DEFAULT_CPU_IMAGE will be used as the Docker image for the run.
distributed_job_config Required	TensorflowConfiguration or MpiConfiguration or PyTorchConfiguration For jobs that require additional distributed job-specific configurations.
resume_from Required	DataPath The DataPath containing the checkpoint or model files from which to resume the experiment.
max_run_duration_seconds Required	int The maximum time allowed for the run. The system will attempt to automatically cancel the run if it took longer than this value.
command Required	list[str] or str The command to be submitted for the run. The command property can also be used instead of script/arguments. Both command and script/argument properties cannot be used together to submit a run. To submit a script file using the command property - ['python', 'train.py', '–arg1', arg1_val] To run an actual command - ['ls']
docker_runtime_config Required	DockerConfiguration For jobs that require Docker runtime-specific configurations.

Remarks

The Azure Machine Learning SDK provides you with a series of interconnected classes, that are designed to help you train and compare machine learning models that are related by the shared problem that they are solving.

An Experiment acts as a logical container for these training runs. A ScriptRunConfig object is used to configure the information necessary for submitting a training run as part of an Experiment. When a run is submitted using a ScriptRunConfig object, the submit method returns an object of type ScriptRun. Then returned ScriptRun object gives you programmatic access to information about the training run. ScriptRun is a child class of Run.

The key concept to remember is that there are different configuration objects that are used to submit an experiment, based on what kind of run you want to trigger. The type of the configuration object then informs what child class of Run you get back from the submit method. When you pass a ScriptRunConfig object in a call to Experiment's submit method, you get back a ScriptRun object. Examples of other run objects returned include AutoMLRun (returned for an AutoML run) and PipelineRun (returned for a Pipeline run).

The following sample shows how to submit a training script on your local machine.


   from azureml.core import ScriptRunConfig, Experiment

   # create or load an experiment
   experiment = Experiment(workspace, 'MyExperiment')
   # create or retrieve a compute target
   cluster = workspace.compute_targets['MyCluster']
   # create or retrieve an environment
   env = Environment.get(ws, name='MyEnvironment')
   # configure and submit your training run
   config = ScriptRunConfig(source_directory='.',
                            script='train.py',
                            arguments=['--arg1', arg1_val, '--arg2', arg2_val],
                            compute_target=cluster,
                            environment=env)
   script_run = experiment.submit(config)

The following example shows how to submit a training script on your cluster using the command property instead of script and arguments.


   from azureml.core import ScriptRunConfig, Experiment
   # create or load an experiment
   experiment = Experiment(workspace, 'MyExperiment')
   # create or retrieve a compute target
   cluster = workspace.compute_targets['MyCluster']
   # create or retrieve an environment
   env = Environment.get(ws, name='MyEnvironment')
   # configure and submit your training run
   config = ScriptRunConfig(source_directory='.',
                            command=['python', 'train.py', '--arg1', arg1_val],
                            compute_target=cluster,
                            environment=env)
   script_run = experiment.submit(config)

The following sample shows how to run a command on your cluster.


   from azureml.core import ScriptRunConfig, Experiment
   # create or load an experiment
   experiment = Experiment(workspace, 'MyExperiment')
   # create or retrieve a compute target
   cluster = workspace.compute_targets['MyCluster']
   # create or retrieve an environment
   env = Environment.get(ws, name='MyEnvironment')
   # configure and submit your training run
   config = ScriptRunConfig(source_directory='.',
                            command=['ls', '-l'],
                            compute_target=cluster,
                            environment=env)
   script_run = experiment.submit(config)

For more examples showing how to work with ScriptRunConfig, see:

the article Select and use a compute target to train your model
these training notebooks

Attributes

MAX_DURATION_SECONDS_DEFAULT

MAX_DURATION_SECONDS_DEFAULT = 2592000

Share via