Model Compression Toolkit User Guide¶
Overview¶
Model Compression Toolkit (MCT) is an open source project for neural networks optimization that enables users to compress and quantize models. This project enables researchers, developers and engineers an easily way to optimized and quantized state-of-the-art neural network.
Currently, MCT supports hardware-friendly post training quantization (HPTQ) with Tensorflow 2 [1].
MCT project is developed by researchers and engineers working in Sony Semiconductors Israel.
Install¶
See the MCT install guide for the pip package, and build from source.
From Source:
git clone https://github.com/sony/model_optimization.git
python setup.py install
From PyPi:
pip install model-compression-toolkit
A nightly version is also available:
pip install mct-nightly
Supported Features¶
Quantization:
Hardware-friendly Post Training Quantization [1]
Gradient base post training using knowledge distillation (Experimental)
Visualization:
Quickstart¶
Take a look of how you can start using MCT in just a few minutes
API Documentation¶
Please visit the MCT API documentation here
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.