mct-nightly 1.4.0.8062022.post354

A Model Compression Toolkit for neural networks

Model Compression Toolkit (MCT)

Model Compression Toolkit (MCT) is an open-source project for neural network model optimization under efficient, constrained hardware. This project provides researchers, developers, and engineers tools for optimizing and deploying state-of-the-art neural networks on efficient hardware. Specifically, this project aims to apply quantization and pruning schemes to compress neural networks.

Currently, this project supports hardware-friendly post-training quantization (HPTQ) with Tensorflow 2 and Pytorch [1].

The MCT project is developed by researchers and engineers working at Sony Semiconductors Israel.

For more information, please visit our project website.

Table of Contents

• Getting Started
• Supported features
• Results
• Contributions
• License

Getting Started

This section provides a quick starting guide. We begin with installation via source code or pip server. Then, we provide a short usage example.

Installation

See the MCT install guide for the pip package, and build from the source.

From Source

git clone https://github.com/sony/model_optimization.git
python setup.py install

From PyPi - latest stable release

pip install model-compression-toolkit

A nightly package is also available (unstable):

pip install mct-nightly

To run MCT, one of the supported frameworks, Tenosflow/Pytorch, needs to be installed.

For using with Tensorflow please install the packages: tensorflow, tensorflow-model-optimization

For using with Pytorch (experimental) please install the packages: torch

MCT is tested with:

• Tensorflow version 2.7
• Pytorch version 1.10.0

Usage Example

For an example of how to use the post-training quantization, using Keras, please use this link.

For an example using Pytorch (experimental), please use this link.

For more examples please see the tutorials' directory.

Supported Features

Quantization:

• Post Training Quantization for Keras models.
• Post Training Quantization for Pytorch models (experimental).
• Gradient-based post-training (Experimental, Keras only).
• Mixed-precision post-training quantization (Experimental).

Tensorboard Visualization (Experimental):

• CS Analyzer: compare a model compressed with the original model to analyze large accuracy drops.
• Activation statistics and errors

Results

Keras

As part of the MCT library, we have a set of example networks on image classification. These networks can be used as examples when using the package.

• Image Classification Example with MobileNet V1 on ImageNet dataset

Network Name	Float Accuracy	8Bit Accuracy	Comments
MobileNetV1 [2]	70.558	70.418

For more results please see [1]

Pytorch

We quantized classification networks from the torchvision library. In the following table we present the ImageNet validation results for these models:

Network Name	Float Accuracy	8Bit Accuracy
MobileNet V2 [3]	71.886	71.444
ResNet-18 [3]	69.86	69.63
SqueezeNet 1.1 [3]	58.128	57.678

Contributions

MCT aims at keeping a more up-to-date fork and welcomes contributions from anyone.

*You will find more information about contributions in the Contribution guide.

License

Apache License 2.0.

References

[1] Habi, H.V., Peretz, R., Cohen, E., Dikstein, L., Dror, O., Diamant, I., Jennings, R.H. and Netzer, A., 2021. HPTQ: Hardware-Friendly Post Training Quantization. arXiv preprint.

[2] MobilNet from Keras applications.

[3] TORCHVISION.MODELS