PYNQ-Torch: a framework to develop PyTorch accelerators on the PYNQ platform
The following repository provides information on setting up the environment as well as examples for using the proposed framework. It is organised as follows:
- The PyTorch Installation folder provides steps on how to install PyTorch v1.2.0 on the PYNQ OS v2.4. It also provides an example on how to use the tools effectively to embed an accelerator within an application, which includes Vivado project files.
- The SampleRNN folder showcases the accelerators being used within a state-of-the-art RNN application. Again, includes Vivado project files.
If you use any part of this work please cite:
Manohar Vohra, Stefano Fasciani, PYNQ-Torch: a framework to develop PyTorch accelerators on the PYNQ platform, in Proceedings of 2019 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT 2019), 2019
@INPROCEEDINGS {,
author = "Manohar Vohra, Stefano Fasciani",
title = "PYNQ-Torch: a framework to develop PyTorch accelerators on the PYNQ platform",
booktitle = "Proceedings of 2019 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT 2019)",
year = "2019",
organization = "IEEE"
}