Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour

Priya Goyal, Piotr Dollár, Ross Girshick, +6

arXiv (Cornell University)

Sgd Minibatch SizeConvolutional Neural NetworkMachine VisionMachine LearningData Science +11

Machine Learning at Facebook: Understanding Inference at the Edge

Carole-Jean Wu, David Brooks, Kevin Chen, +22

Artificial IntelligenceEngineeringMachine LearningLink PredictionData Science +14

Software-hardware co-design for fast and scalable training of deep learning recommendation models

Dheevatsa Mudigere, Yuchen Hao, Jianyu Huang, +50

Artificial IntelligenceEngineeringMachine LearningComputer ArchitectureScalable Training +18

GPT-4o System Card

OpenAI, A. M. Hurst, Adam Lerer, +364

arXiv (Cornell University)

Artificial IntelligenceEngineeringMachine LearningSpeech RecognitionNatural Language Processing +16

Deep Learning Inference in Facebook Data Centers: Characterization, Performance Optimizations and Hardware Implications

Jongsoo Park, Maxim Naumov, Protonu Basu, +25

arXiv (Cornell University)

Graph Neural NetworkEngineeringMachine LearningData ScienceMachine Learning Model +13

High performance ultra-low-precision convolutions on mobile devices

Andrew Tulloch, Yangqing Jia

arXiv (Cornell University)

Convolutional Neural NetworkMobile Deep LearningMachine LearningEngineeringMobile Devices +14

High-performance, Distributed Training of Large-scale Deep Learning Recommendation Models.

Dheevatsa Mudigere, Yuchen Hao, Jianyu Huang, +48

arXiv (Cornell University)

Cluster ComputingEngineeringMachine LearningAdvanced ComputingDistributed Training +16

Mixed-Precision Embedding Using a Cache

Jie Yang, Jianyu Huang, Jongsoo Park, +2

arXiv (Cornell University)

EngineeringMachine LearningComputer ArchitectureMixed-precision EmbeddingTable Compression +15

Accurate Low-Rank Approximations Via a Few Iterations of Alternating Least Squares

Arthur Szlam, Andrew Tulloch, Mark Tygert

SIAM Journal on Matrix Analysis and Applications

Numerical AnalysisEngineeringMachine LearningMatrix EntriesAlternating Least Squares +15

On Periodic Functions as Regularizers for Quantization of Neural Networks

Maxim Naumov, Utku Diril, Jongsoo Park, +3

arXiv (Cornell University)

Convolutional Neural NetworkEngineeringMachine LearningAutoencodersFunctional Analysis +18