Sam Witteveen and I started the TensorFlow and Deep Learning Singapore group on MeetUp in February 2017, and the twenty-eighth MeetUp - which was a TensorFlow World Extended event - was hosted by Google Singapore, and recapped the TensorFlow World event we attended in Silicon Valley at the end of October.
Sam’s talk was an overview of the latest research into Transformer-based models : “The latest in Transformers : T5 and more”.
Aurélien Géron showed us how we should be “Preprocessing large datasets with Apache Beam and TF Transform” - With lots of code examples, and a specific use-case with DataFlow on Google Cloud.
My talk was “Neural Structured Learning (NSL): Training with Structured Signals”. In the talk, I briefly discussed the kinds of problems that NSL can tackle, and quickly showed the blog/code for a supervised learning example. For a simpler-to-read example, I then went through my updated version of Google adversarial learning example notebook for robustifying MNIST digit recognition.
The slides for my talk are here :
If there are any questions about the presentation please ask below, or contact me using the details given on the slides themselves.
PS: And if you liked the content, please ‘star’ my Deep Learning Workshop repo :: Star
blog comments powered by Disqus