Apache MXNet Vancouver Meetup
Details
This event is the occasion for the MXNet community to meet up, learn and network. Apache MXNet (incubating) is Amazon Web Services (AWS)'s preferred deep learning framework, it is designed for both efficiency and flexibility. It also allows you to mix symbolic and imperative programming to maximize efficiency and productivity.
If you are a MXNet user or curious about how MXNet differs from other libraries, this event is for you. It will be the occasion to meet some of the MXNet contributors and ask them your questions directly.
The evening will feature several technical talks where speakers will present on how they used Apache MXNet and deep learning to solve real-world problems.
If you are actively using MXNet and are interested in contributing a talk, contact Thomas Delteil at tdelteil@amazon.com directly with an abstract and a short bio.
Agenda:
6pm-6:30pm
Mingle, pizza + drinks will be provided
6:30pm-6:35pm
Welcome address
6:40pm-7:10pm
Supervised learning for Natural Language Understanding - Oliver Pringle, Data Scientist @ Finn AI
7:15pm-8:00pm
Amazon SageMaker Automatic Model Tuning with Apache MXNet - Thom Lane, Applied Scientist @ Amazon AI
8:00pm-8:30pm
Mingle
- Supervised learning for Natural Language Understanding
Abstract: Finn AI is an award-winning, conversational banking technology provider, specializing in AI-powered solutions built for personal banking and finance. In this session, Oliver Pringle, Data Scientist at Finn AI will provide an overview of the supervised learning models integral to Finn AI's natural language understanding engine and built using Apache MXNet.
Bio: Data Scientist at Finn AI by day, and extreme sports enthusiast by night, Oliver Pringle is passionate about all things deep learning. He holds a Master in Data Science and Mechanical Engineering and is driven to apply his technical skills to bring noticeable business impacts to emerging, innovative technology.
2) Amazon SageMaker Automatic Model Tuning with Apache MXNet
Abstract: Using AWS SageMaker (and the new Automatic Model Tuning feature), I will be presenting a mixture of theory and application of Bayesian Hyperparameter Optimization (with Gaussian Processes) to MXNet models. And hopefully getting the fastest CIFAR-10 (single GPU) training time on record!
Bio: Thom is a Machine Learning Scientist from Amazon Web Services who's been sharing his expertise in MXNet with internal and external customers over the last 8 months. He studied MORSE at University of Warwick (not the text encoding!) and crossed the pond from the UK (and then some) 18 months ago.
Hoping to see you all Tuesday to make this event a success :)
All the best,
Thomas Delteil
Applied Scientist @ Amazon AI
[IMPORTANT] To attend, RSVP by 9am on Monday August 27th so your name can make it to the list of attendees.