Oral: Larq Compute Engine: Design, Benchmark and Deploy State-of-the-Art Binarized Neural Networks
Apr 4, 2021
Speakers
Tom Bannink
Sprecher:in · 0 Follower:innen
Arash Bakhtiari
Sprecher:in · 0 Follower:innen
Adam Hillier
Sprecher:in · 0 Follower:innen
Lukas Geiger
Sprecher:in · 0 Follower:innen
Tim De Bruin
Sprecher:in · 0 Follower:innen
Leon Overweel
Sprecher:in · 0 Follower:innen
Jelmer Neeven
Sprecher:in · 0 Follower:innen
Koen Helwegen
Sprecher:in · 0 Follower:innen
About
We introduce Larq Compute Engine, the world's fastest Binarized Neural Network (BNN) inference engine, and use this framework to investigate several important questions about the efficiency of BNNs and to design a new state-of-the-art BNN architecture. LCE provides highly optimized implementations of binary operations and accelerates binary convolutions by 8.5 - 18.5x compared to their full-precision counterparts on Pixel 1 phones. LCE's integration with Larq and a sophisticated MLIR-based converter allow users to move smoothly from training to deployment. By extending TensorFlow and TensorFlow Lite, LCE supports models which combine binary and full-precision layers, and can be easily integrated into existing applications. Using LCE, we analyze the performance of existing BNN computer vision architectures and develop QuickNet, a simple, easy-to-reproduce BNN that outperforms existing binary networks in terms of latency and accuracy on ImageNet. Furthermore, we investigate the impact of full-precision shortcuts and the relationship between number of MACs and model latency. We are convinced that empirical performance should drive BNN architecture design and hope this work will facilitate others to design, benchmark and deploy binary models.We introduce Larq Compute Engine, the world's fastest Binarized Neural Network (BNN) inference engine, and use this framework to investigate several important questions about the efficiency of BNNs and to design a new state-of-the-art BNN architecture. LCE provides highly optimized implementations of binary operations and accelerates binary convolutions by 8.5 - 18.5x compared to their full-precision counterparts on Pixel 1 phones. LCE's integration with Larq and a sophisticated MLIR-based conve…
Organizer
MLSys 2021
Konto · 159 Follower:innen
Categories
KI und Datenwissenschaft
Kategorie · 10,8k Präsentationen
About MLSys 2021
The Conference on Machine Learning and Systems targets research at the intersection of machine learning and systems. The conference aims to elicit new connections amongst these fields, including identifying best practices and design principles for learning systems, as well as developing novel learning methods and theory tailored to practical machine learning workflows.
Like the format? Trust SlidesLive to capture your next event!
Professional recording and live streaming, delivered globally.
Sharing
Recommended Videos
Presentations on similar topic, category or speaker
Directions for Deep Learning Hardware
Später ansehen
Ewigspeicher-Fortschrittswert: 0 = 0.0%
To Bridge Neural Network Design and Real-World Performance: A Behaviour Study for Neural Networks
Später ansehen
Xiaohu Tang, …
Ewigspeicher-Fortschrittswert: 0 = 0.0%
CODE: Compiler-based Neuron-aware Ensemble training
Später ansehen
Ewigspeicher-Fortschrittswert: 0 = 0.0%
Directions for Deep Learning Hardware
Später ansehen
Ewigspeicher-Fortschrittswert: 0 = 0.0%
Oral: Pipelined Backpropagation at Scale: Training Large Models without Batches
Später ansehen
Atli Kosson, …
Ewigspeicher-Fortschrittswert: 0 = 0.0%
Oral: A Deep Learning Based Cost Model for Automatic Code Optimization
Später ansehen
Ewigspeicher-Fortschrittswert: 0 = 0.0%