Introduction to Gluon NLP & CV and ONNX

Join us for the second MXNet Copenhagen Meetup to learn about Gluon NLP and the ONNX format. This is a great chance to meet one of the top AWS ML & AI speakers in a more "hyggelig" setting.

Speakers:

Julien Simon
As the Global AI & Machine Learning Evangelist, Julien focuses on helping developers and enterprises bring their ideas to life. He frequently speaks at conferences and he's also actively blogging at https://medium.com/@julsimon.

Cosmin Catalin Sanda
MXNet enthusiast mixing Data Science and Data Engineering at AudienceProject ApS. Blogging at https://cosminsanda.com.

Agenda:

17:00 Check In
17:45 Introduction to Gluon NLP (with Julien)
19:15 MXNet to ONNX to ML.NET (with Cosmin)

Presentations:

Introduction to Gluon NLP & Gluon CV
The Gluon library in Apache MXNet provides a clear, concise, and simple API for deep learning. It makes it easy to prototype, build, and train deep learning models without sacrificing training speed. Gluon also includes GluonCV and GluonNLP, two toolkits implementing state of the art deep learning algorithms for computer vision and natural language processing. In this code-level session, we’ll first introduce you to Gluon, comparing it to the symbolic MXNet API. Then, we’ll take a look at GluonCV and GluonNLP. Of course, we’ll run demos using Python and Jupyter notebooks.

MXNet to ONNX to ML.NET
ML.NET is a new machine learning library from Microsoft. It is appealing because it targets .NET experts. However, it is missing the power of a production ready deep learning library such as MXNet. To overcome this, ML.NET allows for inference using ONNX models. ONNX is an interchangeable format for deep learning models. In this short presentation you will see how to go from modeling a neural network in MXNet to doing inference in ML.NET with the help of ONNX.