tessl/pypi-azure-ai-ml

Microsoft Azure Machine Learning Client Library for Python providing comprehensive SDK for ML workflows including job execution, pipeline components, model deployment, and AutoML capabilities

—

Pending

Quality

Pending

Does it follow best practices?

Impact

Pending

No eval scenarios have been run

Overview

Eval results

Files

Client and Authentication

Name: tessl/pypi-azure-ai-ml
Author: tessl

The MLClient class is the main entry point for all Azure Machine Learning operations. It provides a unified interface to interact with Azure ML services and manages authentication, workspace connections, and operation groups.

Capabilities

MLClient Initialization

Creates the main client for Azure Machine Learning operations with credential and workspace configuration.

class MLClient:
    def __init__(
        self,
        credential,
        subscription_id: str = None,
        resource_group_name: str = None,
        workspace_name: str = None,
        registry_name: str = None,
        **kwargs
    ):
        """
        Initialize MLClient for Azure ML operations.
        
        Parameters:
        - credential: Azure credential (DefaultAzureCredential, ClientSecretCredential, etc.)
        - subscription_id: Azure subscription ID
        - resource_group_name: Resource group containing the workspace
        - workspace_name: Azure ML workspace name
        - registry_name: Registry name (for registry operations)
        - cloud: Cloud environment ("AzureCloud", "AzureChinaCloud", etc.)
        - **kwargs: Additional configuration options
        """

Usage Example

from azure.ai.ml import MLClient
from azure.identity import DefaultAzureCredential, ClientSecretCredential

# Using default credential (recommended for development)
ml_client = MLClient(
    credential=DefaultAzureCredential(),
    subscription_id="your-subscription-id",
    resource_group_name="your-resource-group",
    workspace_name="your-workspace"
)

# Using service principal authentication
credential = ClientSecretCredential(
    tenant_id="your-tenant-id",
    client_id="your-client-id",
    client_secret="your-client-secret"
)

ml_client = MLClient(
    credential=credential,
    subscription_id="your-subscription-id",
    resource_group_name="your-resource-group",
    workspace_name="your-workspace"
)

# For registry operations
registry_client = MLClient(
    credential=DefaultAzureCredential(),
    subscription_id="your-subscription-id",
    resource_group_name="your-resource-group",
    registry_name="your-registry"
)

Operation Groups

MLClient organizes functionality into operation groups for different resource types.

class MLClient:
    @property
    def jobs(self) -> JobOperations: ...
    
    @property
    def models(self) -> ModelOperations: ...
    
    @property
    def data(self) -> DataOperations: ...
    
    @property
    def environments(self) -> EnvironmentOperations: ...
    
    @property
    def components(self) -> ComponentOperations: ...
    
    @property
    def compute(self) -> ComputeOperations: ...
    
    @property
    def datastores(self) -> DatastoreOperations: ...
    
    @property
    def workspaces(self) -> WorkspaceOperations: ...
    
    @property
    def online_endpoints(self) -> OnlineEndpointOperations: ...
    
    @property
    def batch_endpoints(self) -> BatchEndpointOperations: ...
    
    @property
    def online_deployments(self) -> OnlineDeploymentOperations: ...
    
    @property
    def batch_deployments(self) -> BatchDeploymentOperations: ...
    
    @property
    def schedules(self) -> ScheduleOperations: ...
    
    @property
    def connections(self) -> WorkspaceConnectionsOperations: ...
    
    @property
    def registries(self) -> RegistryOperations: ...
    
    @property
    def feature_sets(self) -> FeatureSetOperations: ...
    
    @property
    def feature_stores(self) -> FeatureStoreOperations: ...
    
    @property
    def feature_store_entities(self) -> FeatureStoreEntityOperations: ...
    
    @property
    def indexes(self) -> IndexOperations: ...
    
    @property
    def serverless_endpoints(self) -> ServerlessEndpointOperations: ...
    
    @property
    def marketplace_subscriptions(self) -> MarketplaceSubscriptionOperations: ...

Usage Example

from azure.ai.ml import MLClient
from azure.identity import DefaultAzureCredential

ml_client = MLClient(
    credential=DefaultAzureCredential(),
    subscription_id="your-subscription-id",
    resource_group_name="your-resource-group",
    workspace_name="your-workspace"
)

# Access different operation groups
jobs = ml_client.jobs.list()
models = ml_client.models.list()
compute_targets = ml_client.compute.list()
environments = ml_client.environments.list()

Load Functions

Utility functions for loading Azure ML entities from YAML files.

def load_job(source: str) -> Job:
    """Load job from YAML file."""

def load_model(source: str) -> Model:
    """Load model from YAML file."""

def load_data(source: str) -> Data:
    """Load data asset from YAML file."""

def load_environment(source: str) -> Environment:
    """Load environment from YAML file."""

def load_component(source: str) -> Component:
    """Load component from YAML file."""

def load_compute(source: str) -> Compute:
    """Load compute target from YAML file."""

def load_datastore(source: str) -> Datastore:
    """Load datastore from YAML file."""

def load_workspace(source: str) -> Workspace:
    """Load workspace from YAML file."""

def load_online_endpoint(source: str) -> OnlineEndpoint:
    """Load online endpoint from YAML file."""

def load_batch_endpoint(source: str) -> BatchEndpoint:
    """Load batch endpoint from YAML file."""

def load_online_deployment(source: str) -> OnlineDeployment:
    """Load online deployment from YAML file."""

def load_batch_deployment(source: str) -> BatchDeployment:
    """Load batch deployment from YAML file."""

def load_connection(source: str) -> WorkspaceConnection:
    """Load workspace connection from YAML file."""

def load_registry(source: str) -> Registry:
    """Load registry from YAML file."""

def load_feature_set(source: str) -> FeatureSet:
    """Load feature set from YAML file."""

def load_feature_store(source: str) -> FeatureStore:
    """Load feature store from YAML file."""

def load_feature_store_entity(source: str) -> FeatureStoreEntity:
    """Load feature store entity from YAML file."""

def load_index(source: str) -> Index:
    """Load index from YAML file."""

def load_serverless_endpoint(source: str) -> ServerlessEndpoint:
    """Load serverless endpoint from YAML file."""

def load_marketplace_subscription(source: str) -> MarketplaceSubscription:
    """Load marketplace subscription from YAML file."""

def load_capability_host(source: str) -> CapabilityHost:
    """Load capability host from YAML file."""

def load_model_package(source: str) -> ModelPackage:
    """Load model package from YAML file."""

def load_workspace_connection(source: str) -> WorkspaceConnection:
    """Load workspace connection from YAML file."""

Usage Example

from azure.ai.ml import load_job, load_model, load_environment

# Load entities from YAML files
job = load_job("./job.yml")
model = load_model("./model.yml")
environment = load_environment("./environment.yml")

# Submit the loaded job
submitted_job = ml_client.jobs.create_or_update(job)

Types

class ValidationException(Exception):
    """Exception raised for validation errors."""
    
class ErrorCategory:
    USER_ERROR = "UserError"
    SYSTEM_ERROR = "SystemError"
    UNKNOWN = "Unknown"

class ErrorTarget:
    """Error target constants for exception handling."""

class ValidationErrorType:
    INVALID_VALUE = "INVALID VALUE"
    UNKNOWN_FIELD = "UNKNOWN FIELD"
    MISSING_FIELD = "MISSING FIELD"
    FILE_OR_FOLDER_NOT_FOUND = "FILE OR FOLDER NOT FOUND"
    CANNOT_SERIALIZE = "CANNOT DUMP"
    CANNOT_PARSE = "CANNOT PARSE"
    RESOURCE_NOT_FOUND = "RESOURCE NOT FOUND"
    GENERIC = "GENERIC"

Install with Tessl CLI