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 Processors
- Defining CANN and exploring its strategic role within Huawei’s AI compute stack.
- Examining the architecture of Ascend processors (including 310, 910, and others).
- Reviewing the supported AI frameworks and an overview of the associated toolchain.
Model Conversion and Compilation
- Utilizing the ATC tool for model conversion from TensorFlow, PyTorch, and ONNX.
- Generating and validating OM model files for deployment.
- Addressing unsupported operators and resolving common conversion challenges.
Deploying with MindSpore and Other Frameworks
- Implementing model deployment strategies using MindSpore Lite.
- Integrating OM models via Python APIs or C++ SDKs.
- Collaborating effectively with the Ascend Model Manager.
Performance Optimization and Profiling
- Understanding optimizations related to AI Core, memory management, and tiling.
- Profiling model execution processes using CANN diagnostic tools.
- Applying best practices to enhance inference speed and resource efficiency.
Error Handling and Debugging
- Identifying common deployment errors and implementing effective resolutions.
- Interpreting logs and utilizing error diagnosis tools for root cause analysis.
- Conducting unit testing and functional validation for deployed models.
Edge and Cloud Deployment Scenarios
- Deploying to Ascend 310 for targeted edge applications.
- Integrating solutions with cloud-based APIs and microservice architectures.
- Analyzing real-world case studies in computer vision and Natural Language Processing (NLP).
Summary and Next Steps
Requirements
- Proficiency in Python-based deep learning frameworks, including TensorFlow or PyTorch.
- A solid grasp of neural network architectures and standard model training workflows.
- Fundamental knowledge of the Linux Command Line Interface (CLI) and scripting basics.
Target Audience
- AI engineers focused on the model deployment lifecycle.
- Machine learning practitioners aiming to leverage hardware acceleration capabilities.
- Deep learning developers building robust inference solutions.
