Making PyTorch AI Models Portable Using ONNX

Keith Pijanowski

Rate me:

5.00/5 (2 votes)

8 Sep 2020CPOL5 min read

6.3K

In this article I provide a brief overview of PyTorch for those looking for a deep learning framework for building and training neural networks.

Here I then show how to convert PyTorch models to the ONNX format using the conversion tool which is already a part of PyTorch itself. I also show the best practice of adding metadata to the exported model.

Views

Daily Counts

Downloads

Weekly Counts

This article is part of the series 'CodeProject’s Next Top Model ◁ Prev View All Next ▷

License

This article, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

Written By

Keith Pijanowski

Technical Lead Bank of New York Mellon

United States

Keith is a sojourner in the software industry. He has over 30 years of experience building and bringing applications to market. He has worked for startups and large enterprises in roles ranging from tech lead to business development manager. He is currently a senior engineer on BNY Mellon's Distribution Analytics team where he is building data pipelines from on-premise data sources to the cloud.

Making PyTorch AI Models Portable Using ONNX

Views

Downloads

License

Comments and Discussions