114-quantization-simplified-mode

Simplified Post-Training Quantization of Image Classification Models with OpenVINO™

This tutorial demonstrates how to perform INT8 quantization with an image classification model, using the Simplified Mode in Post-Training Optimization Tool (part of OpenVINO). A ResNet20 model and Cifar10 dataset are used. The code in this tutorial is designed to be extendable to custom models and datasets.

Notebook Contents

The tutorial consists of the following steps:

• Downloading and preparing the ResNet20 model and the calibration dataset.
• Preparing the model for quantization.
• Compressing the model by using the simplified mode.
• Comparing performance of the original and quantized models.
• Demonstrating the results of the optimized model.