Past Meetup

Building and Scaling your First Machine Learning model. Lessons from AlphaZero.


Needs a location


Following the success of their AlphaGo program that achieved superhuman performance in the game of Go, DeepMind has generalized their reinforcement learning approach into a single AlphaZero algorithm that can achieve "almost alien-like" performance across many challenging domains. Starting from random play and without any domain knowledge, within 24 hours, AlphaZero taught itself to play chess well enough to defeat a world-champion program in games of chess and shogi (Japanese chess).

For our October meetup, we have invited the Oracle Cloud team to give us a tour of the AlphaZero algorithm and share the architecture, protocols and other technologies that made it possible to implement DeepMind's revolutionary algorithm. In this interactive session, the team will also answer your questions and share best practices for building/scaling your first neural network/ML model.

For our more advanced practitioners, this discussion will include a deep dive of popular model formats including Tensorflow, Caffe2, ONNX, and TensorRT, and how to deploy these models to production using various serving technologies like GraphPipe.

Space is limited so please reserve your seat ASAP.