Dive into the Model-Compression-Toolkit (MCT) with our collection of tutorials, covering a wide range of compression techniques for Keras and Pytorch models. We provide both Python scripts and interactive Jupyter notebooks for an engaging and hands-on experience.
Post-Training Quantization (PTQ)
Tutorial | Included Features |
---|---|
MobileNetV2 | ✅ PTQ |
Mixed-Precision MobileNetV2 | ✅ PTQ ✅ Mixed-Precision |
Nanodet-Plus | ✅ PTQ |
YoloV8-nano | ✅ PTQ |
EfficientDetLite0 | ✅ PTQ ✅ sony-custom-layers integration |
Gradient-Based Post-Training Quantization (GPTQ)
Tutorial | Included Features |
---|---|
MobileNetV2 | ✅ GPTQ |
Quantization-Aware Training (QAT)
Tutorial | Included Features |
---|---|
QAT on MNIST | ✅ QAT |
Structured Pruning
Tutorial | Included Features |
---|---|
Fully-Connected Model Pruning | ✅ Pruning |
Export Quantized Models
Tutorial | Included Features |
---|---|
Exporter Usage | ✅ Export |
Debug Tools
Tutorial | Included Features |
---|---|
Network Editor Usage | ✅ Network Editor |
Quick-Start with Torchvision
Tutorial |
---|
Quick Start - Torchvision Pretrained Model |
Post-Training Quantization (PTQ)
Tutorial | Included Features |
---|---|
Training & Quantizing Model on MNIST | ✅ PTQ |
Mixed-Precision MobileNetV2 on Cifar100 | ✅ PTQ ✅ Mixed-Precision |
SSDLite MobileNetV3 Quantization | ✅ PTQ |
Quantization-Aware Training (QAT)
Tutorial | Included Features |
---|---|
QAT on MNIST | ✅ QAT |
Structured Pruning
Tutorial | Included Features |
---|---|
Fully-Connected Model Pruning | ✅ Pruning |
Data Generation
Tutorial | Included Features |
---|---|
Data-Free Quantization using Data Generation | ✅ PTQ ✅ Data-Free Quantization ✅ Data Generation |
Export Quantized Models
Tutorial | Included Features |
---|---|
Exporter Usage | ✅ Export |