HGQ 0.1.3

High Granularity Quantizarion for hls4ml

High Granularity Quantization for Ultra-Fast Inference on FPGAs

HGQ is a method for quantization aware training of neural works to be deployed on FPGAs, which allows for per-weight and per-activation bitwidth optimization.

Depending on the specific application, HGQ could achieve up to 10x resource reduction compared to the traditional AutoQkeras approach, while maintaining the same accuracy. For some more challenging tasks, where the model is already under-fitted, HGQ could still improve the performance under the same on-board resource consumption. For more details, please refer to our paper (link coming not too soon).

This repository implements HGQ for tensorflow.keras models. It is independent of the QKeras project.

Notice: this repository is still under development, and the API might change in the future.

This package is still under development. Any API might change without notice at any time

Installation

pip install HGQ, and you are good to go. Note that HGQ requires python3.10 and tensorflow>=2.11.

Usage Guide

Please refer to the usage guide for more details. This repo contains some use cases for HGQ.

FAQ

Please refer to the FAQ for more details.

Citation

The paper is not ready. Please check back later.

Important Usage Notice

When using HGQ in your project, remember to...

• Make sure inp_q is in the input quantizer layer's name.
• Add ResetMinMax callback during training.
• Call compute_bops before converting to hls4ml model.