Skip to main content
Skip table of contents

Quantization Guided Training (QGT) Tutorial - PDF

We provide an example tutorial using Quantization Guided Training (QGT) to fine-tune and quantize a pretrained ResNet50 model (ImageNet 1000 classes). The purpose is to show how to use the QGT API using hyperparameters to achieve the best accuracy while reducing bit-precision to reduce model size.

Click here to view tutorial.

Click here to download tutorial.

JavaScript errors detected

Please note, these errors can depend on your browser setup.

If this problem persists, please contact our support.