How INT8 Quantization Made My Neural Network 60% Smaller: A TinyML Model Compression Experiment
中文摘要
使用INT8量化将TinyML心电图心律失常检测模型体积缩小60%的实验探索。
English Summary
An exploration of using INT8 quantization to reduce a TinyML ECG arrhythmia detection model's size by 60%.
原文节选
Exploring what happened when I used INT8 quantization to compress an ECG arrhythmia detection model for TinyML. Continue reading on Medium »