+49 (0) 30 5201 4923
info@orb-experts.com
This course is designed for engineers aiming to escalate Machine Learning workloads on Azure. Conducted by seasoned professionals, the training will take you through the intricacies of Kubeflow, Kubernetes, and TensorFlow.
What You Will Achieve:
• Grasp the nuances of Kubeflow and how it interfaces with Kubernetes.
• Acquire the skills to set up and manage Azure Kubernetes Service (AKS).
• Learn to create robust Kubernetes pipelines for ML model automation.
• Gain experience in multi-GPU and parallel machine training with TensorFlow.
• Extend ML capabilities with Azure's managed services.
• Familiarity with machine learning and cloud computing concepts.
• Basic understanding of containers (Docker) and orchestration (Kubernetes).
• Working knowledge of command-line interfaces.
• Python programming experience is beneficial but not mandatory.
*We customize the course outline and content to your specific needs and relevant use cases.
1. Introduction:
• Comparison: Kubeflow on Azure, On-premise, and other public clouds
• Architectural Overview of Kubeflow
2. Setting the Stage:
• Activating an Azure Account
• Launching GPU-Enabled Virtual Machines
• User Roles and Permissions: A Primer
3. Building the Environment:
• Preparing the Build Environment
• Introduction to TensorFlow Models and Datasets
• Packaging Code: Dockerization
4. Deployment Infrastructure:
• Introduction to Azure Kubernetes Service (AKS)
• Kubernetes Cluster Initialization with AKS
5. Data Management:
• Data Staging: Training and Validation
• Configuring Kubeflow Pipelines
6. Training and Monitoring:
• Launching a Training Job
• Real-Time Monitoring of Training Jobs
7. Post-Deployment:
• Cleaning Up Resources
• Troubleshooting Tips
8. Summary and Conclusion:
• Key Takeaways
• Next Steps in Leveraging Azure for ML Workloads
Hands-on learning with expert instructors at your location for organizations.
Master new skills guided by experienced instructors from anywhere.
