QONNX: Representing Arbitrary-Precision Quantized Neural Networks
Abstract
We present extensions to the Open Neural Network Exchange (ONNX) intermediate representation format to represent arbitrary-precision quantized neural networks. We first introduce support for low precision quantization in existing ONNX-based quantization formats by leveraging integer clipping, resulting in two new backward-compatible variants: the quantized operator format with clipping and quantize-clip-dequantize (QCDQ) format. We then introduce a novel higher-level ONNX format called quantized ONNX (QONNX) that introduces three new operators -- Quant, BipolarQuant, and Trunc -- in order to represent uniform quantization. By keeping the QONNX IR high-level and flexible, we enable targeting a wider variety of platforms. We also present utilities for working with QONNX, as well as examples of its usage in the FINN and hls4ml toolchains. Finally, we introduce the QONNX model zoo to share low-precision quantized neural networks.
Cite
@article{arxiv.2206.07527,
title = {QONNX: Representing Arbitrary-Precision Quantized Neural Networks},
author = {Alessandro Pappalardo and Yaman Umuroglu and Michaela Blott and Jovan Mitrevski and Ben Hawks and Nhan Tran and Vladimir Loncar and Sioni Summers and Hendrik Borras and Jules Muhizi and Matthew Trahms and Shih-Chieh Hsu and Scott Hauck and Javier Duarte},
journal= {arXiv preprint arXiv:2206.07527},
year = {2022}
}
Comments
9 pages, 5 figures, Contribution to 4th Workshop on Accelerated Machine Learning (AccML) at HiPEAC 2022 Conference