Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Duration 14 hours
Course Outline
Introduction to CANN and Ascend AI Processing Units
- Definition of CANN and its function within the Huawei AI computing ecosystem
- Overview of Ascend processor architectures, including Series 310 and 910
- Summary of supported AI frameworks and the associated toolchain for government applications
Model Transformation and Compilation Procedures
- Utilizing the ATC utility for model conversion from TensorFlow, PyTorch, and ONNX standards
- Generation and verification of OM model files
- Management of unsupported operators and mitigation of common conversion challenges
Deployment via MindSpore and Third-Party Frameworks
- Implementation of models using MindSpore Lite
- Integration of OM models via Python interfaces or C++ SDKs
- Operational use of the Ascend Model Manager for government infrastructure
Performance Optimization and Profiling Protocols
- Analysis of AI Core, memory, and tiling optimization techniques
- Profiling of model execution using CANN diagnostic tools
- Best practices for enhancing inference latency and resource efficiency in public sector deployments
Error Management and Debugging Procedures
- Identification and resolution of common deployment anomalies
- Interpretation of logs and application of error diagnostic utilities
- Unit testing and functional validation of deployed models to ensure accountability
Edge and Cloud Deployment Architectures
- Deployment to Ascend 310 for edge-based applications
- Integration with cloud-based APIs and microservices for government use n
- Practical case studies in computer vision and natural language processing
Summary and Recommended Follow-up Actions
Requirements
- Proficiency with Python-based deep learning frameworks, such as TensorFlow or PyTorch
- Working knowledge of neural network architectures and model training workflows
- Fundamental familiarity with Linux command-line interfaces and scripting
Target Audience
- AI engineers involved in model deployment initiatives
- Machine learning practitioners focused on hardware acceleration strategies
- Deep learning developers constructing inference solutions for public sector applications
