Course Outline
Preparing Machine Learning Models for Production Deployment
- Packaging models using Docker
- Exporting models from TensorFlow and PyTorch
- Managing versioning and storage requirements
Serving Models on Kubernetes
- An introduction to inference servers
- Deploying TensorFlow Serving and TorchServe
- Configuring model endpoints
Optimizing Inference Performance
- Implementing effective batching strategies
- Managing concurrent request processing
- Tuning for optimal latency and throughput
Autoscaling Machine Learning Workloads
- Utilizing the Horizontal Pod Autoscaler (HPA)
- Leveraging the Vertical Pod Autoscaler (VPA)
- Implementing Kubernetes Event-Driven Autoscaling (KEDA)
Provisioning GPUs and Managing Resources
- Setting up GPU-enabled nodes
- An overview of the NVIDIA device plugin
- Defining resource requests and limits for ML workloads
Strategies for Model Rollout and Release
- Executing blue/green deployments
- Applying canary rollout patterns
- Conducting A/B testing for model evaluation
Monitoring and Observability for Production ML
- Tracking key metrics for inference workloads
- Establishing best practices for logging and tracing
- Creating dashboards and setting up alerting mechanisms
Security and Reliability Considerations
- Securing model endpoints
- Implementing network policies and access control
- Safeguarding high availability
Conclusion and Future Directions
Requirements
- A solid grasp of containerized application workflows
- Practical experience with Python-based machine learning models
- Basic familiarity with Kubernetes core concepts
Target Audience
- ML engineers
- DevOps engineers
- Platform engineering teams
Testimonials (4)
basic understanding of container/kubernetes and how they interact features of the openshift plattform
Eric Scholze - NOW IT GmbH
Course - Introduction to Containers, Kubernetes & OpenShift
About the microservices and how to maintenance kubernetes
Yufri Isnaini Rochmat Maulana - Bank Indonesia
Course - Advanced Platform Engineering: Scaling with Microservices and Kubernetes
How trainer deliver knowledge so effectively
Vu Thoai Le - Reply Polska sp. z o. o.
Course - Certified Kubernetes Administrator (CKA) - exam preparation
The knowledge and exchanges with Augustin