Overview
Discover PyTorch 1.0, a groundbreaking open-source deep learning framework that bridges the gap between research and production in AI development. Explore how this innovative tool from Meta accelerates the transition from cutting-edge research to large-scale production, combining both research and production needs into a single, powerful framework. Learn about the PyTorch ecosystem, unframework approach, and machine learning execution flow. Gain insights into Facebook's AI innovation process, including model development, computation graph toolkits, and distributed training techniques. Understand how PyTorch 1.0 addresses mobile fragmentation and supports the advancement of AI in various domains, including open translation at Facebook. Delve into the open-source tools and hardware that contribute to the framework's effectiveness, such as Twin Lakes, Tioga Pass, and Bryce Canyon.
Syllabus
Intro
Fair Is Advancing Al in the Open
Translation at Facebook
Open Source Tools
Open Source Hardware
Twin Lakes
Tioga Pass
Bryce Canyon
PyTorch Ecosystem
Unframework
Machine Learning Execution Flow
Putting It Together
Research to Production at Facebook
Model development
Computation Graph Toolkits
Declarative Toolkits
Imperative Toolkits
Distributed Training Inside Each Iteration
Mobile Fragmentation
Taught by
Meta Developers