Merci d'avoir envoyé votre demande ! Un membre de notre équipe vous contactera sous peu.
Merci d'avoir envoyé votre réservation ! Un membre de notre équipe vous contactera sous peu.
Plan du cours
Introduction to Huawei CloudMatrix
- CloudMatrix ecosystem and deployment flow
- Supported models, formats, and deployment modes
- Typical use cases and supported chipsets
Preparing Models for Deployment
- Model export from training tools (MindSpore, TensorFlow, PyTorch)
- Using ATC (Ascend Tensor Compiler) for format conversion
- Static vs dynamic shape models
Deploying to CloudMatrix
- Service creation and model registration
- Deploying inference services via UI or CLI
- Routing, authentication, and access control
Serving Inference Requests
- Batch vs real-time inference flows
- Data preprocessing and postprocessing pipelines
- Calling CloudMatrix services from external apps
Monitoring and Performance Tuning
- Deployment logs and request tracking
- Resource scaling and load balancing
- Latency tuning and throughput optimization
Integration with Enterprise Tools
- Connecting CloudMatrix with OBS and ModelArts
- Using workflows and model versioning
- CI/CD for model deployment and rollback
End-to-End Inference Pipeline
- Deploying a complete image classification pipeline
- Benchmarking and validating accuracy
- Simulating failover and system alerts
Summary and Next Steps
Pré requis
- An understanding of AI model training workflows
- Experience with Python-based ML frameworks
- Basic familiarity with cloud deployment concepts
Audience
- AI ops teams
- Machine learning engineers
- Cloud deployment specialists working with Huawei infrastructure
21 Heures
