Horovod, Uber's open source distributed deep learning system, enables NVIDIA to scale model training from one to eight GPUs for their self-driving sensing and perception technologies.
Uber built Michelangelo, our machine learning platform, in 2015. Three years later, we reflect our journey to scaling ML at Uber and lessons learned along the way.
Uber open source projects leads give updates on seven of our projects, all of which will be showcased at the upcoming Uber Open Summit 2018.
Keynote speakers include Jim Zemlin, executive director of the Linux Foundation, and Zoubin Ghahramani, chief scientist at Uber AI Labs.
As we approach the New Year, Uber Open Source revisits some of Uber Engineering's most popular projects from 2017.
Uber Engineering introduces Horovod, an open source framework that makes it faster and easier to train deep learning models with TensorFlow.