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.
Course Outline
Introduction to CANN and Ascend AI Processors
- Understanding CANN: Its definition and function within Huawei’s AI compute ecosystem.
- An overview of Ascend processor architectures, including models like the 310 and 910.
- A summary of supported AI frameworks and the associated toolchain.
Model Conversion and Compilation
- Utilizing the ATC tool to convert models from TensorFlow, PyTorch, and ONNX.
- The process of creating and verifying OM model files.
- Strategies for managing unsupported operators and addressing typical conversion challenges.
Deployment with MindSpore and Other Frameworks
- Deploying models using MindSpore Lite.
- Integrating OM models via Python APIs or C++ SDKs.
- Utilizing the Ascend Model Manager for efficient workflow management.
Performance Optimization and Profiling
- Insights into AI Core, memory management, and tiling optimizations.
- Profiling model execution using dedicated CANN tools.
- Best practices for enhancing inference speed and resource efficiency.
Error Handling and Debugging
- Identifying and resolving common deployment errors.
- Interpreting logs and utilizing error diagnosis utilities.
- Conducting unit testing and functional validation for deployed models.
Edge and Cloud Deployment Scenarios
- Deploying solutions to Ascend 310 for edge computing applications.
- Integrating models with cloud-based APIs and microservices.
- Examining real-world case studies in computer vision and NLP.
Summary and Future Directions
Requirements
- Proficiency with Python-based deep learning frameworks, such as TensorFlow or PyTorch.
- A solid understanding of neural network architectures and standard model training workflows.
- Fundamental knowledge of the Linux Command Line Interface (CLI) and basic scripting.
Target Audience
- AI engineers focused on model deployment strategies.
- Machine learning practitioners aiming to leverage hardware acceleration.
- Deep learning developers constructing inference solutions.
14 Hours