Discover how to optimize real-time inference with just 7 lines of code in this informative MLOps Community meetup talk. Explore the OpenVINO toolkit, its installation methods, and key components like the model optimizer and runtime. Learn about supported devices, performance hints, auto-batching, and handling different input shapes. Gain insights into the Open Model Zoo, quantization techniques, and the broader Intel AI software portfolio. Watch live demonstrations of object detection and NLP model optimization, and understand how to leverage OpenVINO for faster inference across various deployment scenarios. Delve into practical tips, ecosystem adoption, and future developments in this comprehensive overview of accelerating AI workloads.
The 7 Lines of Code You Need to Run Faster Real-time Inference
Overview
Syllabus
[] Introduction to Adrian Boguszewski
[] Introduction to OpenVino Toolkit
[] The 7 Lines of Code You Need to Run Faster Real-time Inference
[] Demo
[] The challenge
[] "But running on CPU is slow..."
[] The solution
[] OpenVINO
[] Installation methods
[] Model optimizer
[] Neural Network any format
[] Intermediate Representation IR
[] Post-training Optimization Tool POT
[] OpenVINO Runtime
[] Supported Devices
[] Auto device
[] Performance hints
[] Auto Batching
[] Input data with different shapes
[] Static shape
[] Dynamic shape
[] Open Model Zoo
[] Performance
[] Demo
[] Run live object detection
[] Quantize NLP Model with post
[] OpenVINO Notebooks
[] Intel Developer Cloud for the Edge
[] OpenVINO Ecosystem Adoption
[] Intel AI Software Portfolio
[] Main Takeaways
[] Win $50 in Swag!
[] Adrian's start with OpenVINO
[] Deploying OpenVINO outside the local system
[] Developing tricks with OpenVINO
[] OpenVINO Hosted Service
[] Wrap up
Taught by
MLOps.community
