MMLBaseEstimatorRunConfig Class

  • Reference

Abstract base class for all Estimator run configs.

DEPRECATED. Use the RunConfiguration class.

Initialize the MMLBaseEstimatorRunConfig.

Inheritance
RunConfiguration
MMLBaseEstimatorRunConfig

Constructor

MMLBaseEstimatorRunConfig(compute_target, vm_size=None, vm_priority=None, entry_script=None, script_params=None, node_count=None, process_count_per_node=None, distributed_backend=None, use_gpu=None, use_docker=None, custom_docker_base_image=None, custom_docker_image=None, image_registry_details=None, user_managed=False, conda_packages=None, pip_packages=None, environment_definition=None, inputs=None, source_directory_data_store=None, shm_size=None)

Parameters

compute_target
AbstractComputeTarget or str
Required

The compute target where training will happen. This can either be an object or the string "local".

vm_size
str
default value: None

The VM size of the compute target that will be created for the training.

Supported values: Any Azure VM size.

vm_priority
str
default value: None

The VM priority of the compute target that will be created for the training. If not specified, 'dedicated' is used.

Supported values: 'dedicated' and 'lowpriority'.

This takes effect only when the vm_size parameter is specified in the input.

entry_script
str
default value: None

The relative path to the file used to start training.

script_params
dict
default value: None

A dictionary containing parameters that will be passed as arguments to the entry_script.

node_count
int
default value: None

The number of nodes in the compute target used for training. Only the the AmlCompute target is supported for distributed training (node_count > 1).

process_count_per_node
int
default value: None

When using MPI as an execution backend, the number of processes per node.

distributed_backend
str
default value: None

The communication backend for distributed training.

Supported values: 'mpi' and 'ps'.

'mpi': MPI/Horovod 'ps': parameter server

This parameter is required when any of node_count, process_count_per_node, worker_count, or parameter_server_count > 1.

When node_count == 1 and process_count_per_node == 1, no backend will be used unless a backend is explicitly set. Only the azureml.core.compute.AmlCompute target is supported for distributed training.

use_gpu
bool
default value: None

Specifies whether the environment to run the experiment should support GPUs. If true, a GPU-based default Docker image will be used in the environment. If false, a CPU-based image will be used. Default Docker images (CPU or GPU) will be used only if the custom_docker_image parameter is not set. This setting is used only in Docker-enabled compute targets.

use_docker
bool
default value: None

Specifies whether the environment to run the experiment should be Docker-based.

custom_docker_base_image
str
default value: None

The name of the Docker image from which the image to use for training will be built.

DEPRECATED. Use the custom_docker_image parameter.

If not set, a default CPU-based image will be used as the base image.

custom_docker_image
str
default value: None

The name of the Docker image from which the image to use for training will be built. If not set, a default CPU-based image will be used as the base image.

image_registry_details
ContainerRegistry
default value: None

The details of the Docker image registry.

user_managed
bool
default value: False

Specifies whether Azure ML reuses an existing Python environment. If false, a Python environment is created based on the conda dependencies specification.

conda_packages
list
default value: None

List of strings representing conda packages to be added to the Python environment for the experiment.

pip_packages
list
default value: None

A list of strings representing pip packages to be added to the Python environment for the experiment.

environment_definition
Environment
default value: None

The environment definition for the experiment. It includes PythonSection, DockerSection, and environment variables. Any environment option not directly exposed through other parameters to the Estimator construction can be set using this parameter. If this parameter is specified, it will take precedence over other environment related parameters like use_gpu, custom_docker_image, conda_packages, or pip_packages and errors will be reported on these invalid combinations.

inputs
list
default value: None

A list of DataReference or DatasetConsumptionConfig objects to use as input.

source_directory_data_store
str
default value: None

The backing data store for the project share.

shm_size
str
default value: None

The size of the Docker container's shared memory block. For more information, see Docker run reference. If not set, the default azureml.core.environment._DEFAULT_SHM_SIZE is used.