11 July 2019
In collaboration with Machine intelligence Garage we are offering this hands on workshop which gives you a quick overview of several deep learning frameworks. With each framework, you’ll learn about the framework’s benefits, supported platforms, installation considerations, and supported back ends.
Deep learning isn’t a single approach but rather a class of algorithms and topologies that you can apply to a broad spectrum of problems. While deep learning is certainly not new, it is experiencing explosive growth because of the intersection of deeply layered neural networks and the use of GPUs to accelerate their execution.
Big data has also fed this growth. Because deep learning relies on supervised learning algorithms, the more data, the better to build these deep learning structures.
18.30 - 19.00 : Registration, Food, Drinks and Networking
19.00 - 21.00 : Introduction, followed by a walkthrough of the Use Case
Why GPUs Matter For Deep Learning ~ 15 mins
What even is "Deep Learning"? Seriously, behind the scenes, how does this stuff work? We'll see a really quick worked example for recognising hand written digits, discussing the theory, and the infrastructure limitations that make GPUs THE platform for Deep Learning.
Working with BIG data - Going beyond a GPU ~ 15 mins
As we continually search for bigger, more complicated, problems the demands we place on the GPU continue to grow rapidly. So where next - how can we continue to build models and algorithms that scale as our data does or enable us to tackle even more intricate problems? In this session we'll talk about the Open Source Large Model Support. IBM's solution to the problem that enables TensorFlow, Caffe and PyTorch developers to overcome the GPU memory limitation and keep growing their deep learning workloads.
Style Transfer - Replicating Art with AI (LAB) ~ 60 mins
In our hands on lab we'll discuss how Style Transfer can be used to take the "style" or visual effect from one image and apply it to another, mimicking the style of an artist. We'll explore the theory behind the process, before getting hands on and running our own Jupyter notebook in Watson Studio to perform Style Transfer on our own images using the Watson Machine Learning service.
Chris Parsons - Machine Learning Developer Advocate at IBM
Mo Haghighi - Head of Developer Advocacy at IBM(Europe)