azure-machine-learning

# Azure Machine Learning Skill This skill provides expert guidance for Azure Machine Learning. Covers troubleshooting, best practices, decision making, architecture & design patterns, limits & quotas, security, configuration, integrations & coding patterns, and deployment. It combines local quick-reference content with remote documentation fetching capabilities. ## How to Use This Skill > **IMPORTANT for Agent**: Use the **Category Index** below to locate relevant sections. For categories with line ranges (e.g., `L35-L120`), use `read_file` with the specified lines. For categories with file links (e.g., `[security.md](security.md)`), use `read_file` on the linked reference file > **IMPORTANT for Agent**: If `metadata.generated_at` is more than 3 months old, suggest the user pull the latest version from the repository. If `mcp_microsoftdocs` tools are not available, suggest the user install it: [Installation Guide](https://github.com/MicrosoftDocs/mcp/blob/main/README.md) This skill requires **network access** to fetch documentation content: - **Preferred**: Use `mcp_microsoftdocs:microsoft_docs_fetch` with query string `from=learn-agent-skill`. Returns Markdown. - **Fallback**: Use `fetch_webpage` with query string `from=learn-agent-skill&accept=text/markdown`. Returns Markdown. ## Category Index | Category | Lines | Description | |----------|-------|-------------| | Troubleshooting | L37-L65 | Diagnosing and fixing Azure ML failures and errors across pipelines, endpoints, AutoML, networking, Kubernetes, environments, data access, prompt flow, and known platform issues. | | Best Practices | L66-L80 | Guidance on optimizing AutoML and training, handling imbalance/overfitting, preparing data, batch/inference performance, monitoring models, and reducing Azure ML compute and cost. | | Decision Making | L81-L108 | Guides for planning and making migration, upgrade, networking, DR, data, compute, deployment, and monitoring decisions across Azure ML v1/v2, Fabric, Prompt Flow, and Agent Framework. | | Architecture & Design Patterns | L109-L114 | Designing real-time inference architectures with online endpoints and building RAG solutions using Azure ML vector stores, including deployment, scaling, and integration patterns. | | Limits & Quotas | L115-L124 | Limits, quotas, and availability for Azure ML: regional/sovereign support, VM SKUs, workspace soft delete, and capacity planning for managed online endpoints. | | Security | L125-L174 | Securing Azure ML workspaces, endpoints, and data: encryption, identity/RBAC, network isolation/VNet, Key Vault secrets, policies, compliance, and secure access to other Azure/on-prem resources. | | Configuration | L175-L408 | Configuring Azure ML components, compute, networking, AutoML, YAML schemas, monitoring, and Prompt Flow so you can build, train, deploy, and manage ML workflows and infrastructure. | | Integrations & Coding Patterns | L409-L451 | Integrating Azure ML with data platforms, REST/MLflow APIs, Spark, Databricks/Synapse/Fabric, and building/debugging prompt flow/RAG tools and deployments. | | Deployment | L452-L481 | Deploying and operationalizing models and pipelines on Azure ML (online/batch endpoints, CI/CD, MLOps, prompt flow, RAG, HF/MLflow/ONNX), including rollout strategies and cross-workspace/registry use. | ### Troubleshooting | Topic | URL | |-------|-----| | Troubleshoot Azure ML designer component error codes | https://learn.microsoft.com/en-us/azure/machine-learning/component-reference/designer-error-codes?view=azureml-api-2 | | Resolve common Azure AutoML forecasting issues | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-automl-forecasting-faq?view=azureml-api-2 | | Debug Azure ML online endpoints locally with VS Code | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-debug-managed-online-endpoints-visual-studio-code?view=azureml-api-2 | | Diagnose and fix Azure ML pipeline failures in studio | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-debug-pipeline-failure?view=azureml-api-2 | | Troubleshoot Azure ML pipeline performance with profiling | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-debug-pipeline-performance?view=azureml-api-2 | | Diagnose and fix Azure ML pipeline reuse issues | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-debug-pipeline-reuse-issues?view=azureml-api-2 | | Troubleshoot Azure automated ML experiment failures | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-auto-ml?view=azureml-api-2 | | Troubleshoot Azure ML batch endpoints and jobs | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-batch-endpoints?view=azureml-api-2 | | Troubleshoot data access issues in Azure ML SDK v2 | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-data-access?view=azureml-api-2 | | Troubleshoot Azure ML data labeling project creation | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-data-labeling?view=azureml-api-2 | | Troubleshoot Azure ML environment image build failures | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-environments?view=azureml-api-2 | | Troubleshoot Azure ML Kubernetes compute workloads | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-kubernetes-compute?view=azureml-api-2 | | Troubleshoot Azure ML Kubernetes extension deployment | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-kubernetes-extension?view=azureml-api-2 | | Diagnose Azure ML managed virtual network issues | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-managed-network?view=azureml-api-2 | | Troubleshoot Azure ML online endpoint deployment and scoring errors | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-online-endpoints?view=azureml-api-2 | | Resolve 'descriptors cannot be created directly' in Azure ML | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-protobuf-descriptor-error?view=azureml-api-2 | | Troubleshoot private endpoint access to Azure ML workspaces | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-secure-connection-workspace?view=azureml-api-2 | | Fix 'Validation for schema failed' errors in Azure ML CLI v2 | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-validation-for-schema-failed-error?view=azureml-api-2 | | Diagnose and fix Azure ML workspace issues | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-workspace-diagnostic-api?view=azureml-api-2 | | Review Azure Machine Learning current known issues | https://learn.microsoft.com/en-us/azure/machine-learning/known-issues/azure-machine-learning-known-issues?view=azureml-api-2 | | Known issue: Invalid certificate during AKS deployment | https://learn.microsoft.com/en-us/azure/machine-learning/known-issues/inferencing-invalid-certificate?view=azureml-api-2 | | Known issue: Updating Azure ML Kubernetes compute fails | https://learn.microsoft.com/en-us/azure/machine-learning/known-issues/inferencing-updating-kubernetes-compute-appears-to-succeed?view=azureml-api-2 | | Troubleshoot common prompt flow issues in Azure ML | https://learn.microsoft.com/en-us/azure/machine-learning/prompt-flow/troubleshoot-guidance?view=azureml-api-2 | | Troubleshoot common prompt flow issues in Azure ML | https://learn.microsoft.com/en-us/azure/machine-learning/prompt-flow/troubleshoot-guidance?view=azureml-api-2 | | Troubleshoot Azure ML managed feature store errors | https://learn.microsoft.com/en-us/azure/machine-learning/troubleshooting-managed-feature-store?view=azureml-api-2 | ### Best Practices | Topic | URL | |-------|-----| | Mitigate overfitting and imbalance in Azure AutoML | https://learn.microsoft.com/en-us/azure/machine-learning/concept-manage-ml-pitfalls?view=azureml-api-2 | | Understand Azure ML model monitoring concepts and practices | https://learn.microsoft.com/en-us/azure/machine-learning/concept-model-monitoring?view=azureml-api-2 | | Optimize and manage Azure Machine Learning costs | https://learn.microsoft.com/en-us/azure/machine-learning/concept-plan-manage-cost?view=azureml-api-2 | | Design feature set transformations in Azure ML | https://learn.microsoft.com/en-us/azure/machine-learning/feature-set-specification-transformation-concepts?view=azureml-api-2 | | Author batch scoring scripts for AML batch deployments | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-batch-scoring-script?view=azureml-api-2 | | Tune Azure ML Kubernetes inference router performance | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-kubernetes-inference-routing-azureml-fe?view=azureml-api-2 | | Optimize Azure Machine Learning compute costs | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-manage-optimize-cost?view=azureml-api-2 | | Prepare image datasets for Azure AutoML vision | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-prepare-datasets-for-automl-images?view=azureml-api-2 | | Apply distributed GPU training patterns in Azure ML | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-train-distributed-gpu?view=azureml-api-2 | | Optimize AutoML for small object detection in images | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-use-automl-small-object-detect?view=azureml-api-2 | | Optimize checkpoint performance for large Azure ML models with Nebula | https://learn.microsoft.com/en-us/azure/machine-learning/reference-checkpoint-performance-for-large-models?view=azureml-api-2 | ### Decision Making | Topic | URL | |-------|-----| | Choose between managed and custom network isolation in Azure ML | https://learn.microsoft.com/en-us/azure/machine-learning/concept-network-isolation-configurations?view=azureml-api-2 | | Choose migration paths from Azure ML Data Import to Fabric | https://learn.microsoft.com/en-us/azure/machine-learning/data-import-migration-guide?view=azureml-api-2 | | Plan failover and disaster recovery for Azure ML | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-high-availability-machine-learning?view=azureml-api-2 | | Manage and migrate imported data assets in Azure ML | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-manage-imported-data-assets?view=azureml-api-2 | | Decide and plan migration from Azure ML v1 to v2 | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-migrate-from-v1?view=azureml-api-2 | | Move Azure ML workspaces between subscriptions | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-move-workspace?view=azureml-api-2 | | Plan Azure ML network isolation architecture | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-network-isolation-planning?view=azureml-api-2 | | Select and use vendor companies for Azure ML data labeling | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-outsource-data-labeling?view=azureml-api-2 | | Map Azure ML v1 datasets to v2 data assets | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-assets-data?view=azureml-api-2 | | Migrate Azure ML model management from SDK v1 to v2 | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-assets-model?view=azureml-api-2 | | Upgrade Azure ML script runs to v2 command jobs | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-command-job?view=azureml-api-2 | | Migrate Azure ML deployment endpoints from SDK v1 to v2 | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-deploy-endpoints?view=azureml-api-2 | | Upgrade Azure ML pipeline endpoints from v1 to v2 | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-deploy-pipelines?view=azureml-api-2 | | Upgrade Azure ML AutoML workflows from SDK v1 to v2 | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-execution-automl?view=azureml-api-2 | | Migrate Azure ML hyperparameter tuning to v2 sweep jobs | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-execution-hyperdrive?view=azureml-api-2 | | Migrate Azure ML parallel run step to SDK v2 parallel job | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-execution-parallel-run-step?view=azureml-api-2 | | Upgrade Azure ML pipelines from SDK v1 to v2 jobs | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-execution-pipeline?view=azureml-api-2 | | Migrate Azure ML local runs from SDK v1 to v2 | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-local-runs?view=azureml-api-2 | | Upgrade ACI web services to Azure ML managed online endpoints | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-managed-online-endpoints?view=azureml-api-2 | | Compare and migrate Azure ML compute management v1 to v2 | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-resource-compute?view=azureml-api-2 | | Migrate datastore management from AML v1 to v2 | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-resource-datastore?view=azureml-api-2 | | Decide how to upgrade Azure ML workspaces to SDK v2 | https://learn.microsoft.com/en-us/azure/machine-learning/migrate-to-v2-resource-workspace?view=azureml-api-2 | | Select and interpret Azure ML generative AI monitoring metrics | https://learn.microsoft.com/en-us/azure/machine-learning/prompt-flow/concept-model-monitoring-generative-ai-evaluation-metrics?view=azureml-api-2 | | Plan migration from Prompt Flow to Agent Framework | https://learn.microsoft.com/en-us/azure/machine-learning/prompt-flow/migrate-prompt-flow-to-agent-framework?view=azureml-api-2 | ### Architecture & Design Patterns | Topic | URL | |-------|-----| | Plan real-time inference with Azure ML online endpoints | https://learn.microsoft.com/en-us/azure/machine-learning/concept-endpoints-online?view=azureml-api-2 | | Use Azure ML vector stores for RAG architectures | https://learn.microsoft.com/en-us/azure/machine-learning/concept-vector-stores?view=azureml-api-2 | ### Limits & Quotas | Topic | URL | |-------|-----| | Check regional availability for standard model deployments | https://learn.microsoft.com/en-us/azure/machine-learning/concept-endpoint-serverless-availability?view=azureml-api-2 | | Understand soft delete retention for ML workspaces | https://learn.microsoft.com/en-us/azure/machine-learning/concept-soft-delete?view=azureml-api-2 | | Manage Azure ML resource quotas and limits | https://learn.microsoft.com/en-us/azure/machine-learning/how-to-manage-quotas?view=azureml-api-2 |