This course provides an inside look at TensorFlow Lite, offering a high-level overview of deploying machine learning models on mobile and IoT devices. The course covers topics such as TensorFlow Lite interpreter, model conversion, inference, performance benchmarking, and optimization. The teaching method includes technical deep dives by the TensorFlow team members. The course is intended for individuals interested in deploying machine learning models on mobile and IoT devices.
Overview
Syllabus
Intro
Outline
Why on device?
TensorFlow vs TensorFlow Lite
Why not TFMobile?
What is TensorFlow Lite?
Interpreter
Acceleration (Delegates)
Model conversion (Python)
Inference (Java)
Selective Registration (Bazel)
Selective Registration (C++)
Performance
Benchmarking (Android)
Inference w/ NNAPI
Inference w/ GPU passthrough
Fast execution
Model conversion w/ Post-Training Quant (Hybrid)
Optimization
Model conversion w/ Post-Training Quant (Full)
Documentation
Model repository
TensorFlow Lite Roadmap
Questions?
Taught by
TensorFlow
