fpgaconvnet-hls 0.1.2

Optimiser for mapping convolutional neural network models to FPGA platforms.

fpgaConvNet HLS

This repository is part of the fpgaConvNet framework, designed to solve the complex mapping problem of Convolutional Neural Networks (CNN) onto Field Programmable Gate Array (FPGA) devices. The HLS repository contains the hardware implementation of CNN building blocks, and performs the mapping automation of a CNN model description to hardware.

Setup

The following programs are required:

1. Vivado HLS (2019.1, 2019.2)
2. python (>=3.8)

Once these programs are installed, you can setup the project from pypi:

python -m pip install fpgaconvnet-hls

Usage

You can see example usage in the tests/networks folder as well as in the fpgaconvnet-tutorial repository. Below is a quick example of how a configuration can be loaded and used to generate and test hardware.

from fpgaconvnet.hls.generate.network import GenerateNetwork

# create instance of the network
gen_net = GenerateNetwork("model-name", "model-config.json", "model.onnx")

# generate hardware and create HLS project for partition 0
gen_net.create_partition_project(0)

# run HLS synthesis for partition 0
gen_net.generate_partition_hardware(0)

---

Please feel free to ask questions or post any issues!