FFT-based Gradient Sparsification for the Distributed Training of Deep Neural Networks

preview_player
Показать описание
Linnan Wang (Brown University) Wei Wu (Los Alamos National Laboratory) Junyu Zhang (University of Minnesota, Twin Cities) Hang Liu (Stevens Institute of Technology) George Bosilca (University of Tennessee) Maurice Herlihy, Rodrigo Fonseca (Brown University)
  1. HPDC ACM
Рекомендации по теме
FFT-based Gradient Sparsification 0:23:51

FFT-based Gradient Sparsification for the Distributed Training of Deep Neural Networks

Butterfly Transform: An 0:01:00

Butterfly Transform: An Efficient FFT Based Neural Architecture Design

MUG '23 Day 0:34:50

MUG '23 Day 3 - Stable-TopK: Gradient Sparsification with Temporal Patterns

Intro to Deep 0:07:08

Intro to Deep Learning Model Sparsification

How can we 0:40:58

How can we leverage FFT-based methods in convolutional networks?

[DeepBayes2019]: Day 6, 0:05:52

[DeepBayes2019]: Day 6, Practical session 2. Sparsification of deep neural networks

Graph streaming algorithms 0:19:46

Graph streaming algorithms Graph sparsification

Using Sparsification Recipes 0:54:50

Using Sparsification Recipes with PyTorch

[shp4783] Detecting Poisoning 0:02:57

[shp4783] Detecting Poisoning Attacks on Federated Learning Using Gradient-Weighted Class Activation

Dynamic Data Structures 0:44:21

Dynamic Data Structures in Machine Learning by Omri Weinstein

High-Performance Communication Strategies 1:00:31

High-Performance Communication Strategies in Parallel and Distributed Deep Learning

Signal Reconstruction Using 0:15:37

Signal Reconstruction Using Neural Networks

Ep 31. Analog 0:53:01

Ep 31. Analog Modulation and Over-the-Air Aggregation [Wireless Future Podcast]

Optimizing Power and 1:01:09

Optimizing Power and Performance For Machine Learning at the Edge - Model Deployment Overview | Arm

Nondeterministic Extensions of 0:59:21

Nondeterministic Extensions of the Strong Exponential Time Hypothesis and Consequences for Non-reduc

Performance Tuning of 1:01:08

Performance Tuning of MPI Collectives - Status Quo and Open Problems

PyTorch Release by 0:17:31

PyTorch Release by Joe Spisak | PyTorch Ecosystem Day 2021

Choco-SGD Communication Efficient 0:20:36

Choco-SGD Communication Efficient Decentralized Learning | AI & ML on the Edge | Anastasia Kolos...

'Fault-Tolerance for HPC: 0:01:59

'Fault-Tolerance for HPC: Theory and Practice' at SC22

ECE 804 - 1:10:02

ECE 804 - Fall 2012 - Lecture 004 with Dr. Yann LeCun

Sketchy Decisions: Convex 0:48:50

Sketchy Decisions: Convex Low-Rank Matrix Optimization with Optimal Storage

[REFAI Seminar 11/11/21] 1:12:57

[REFAI Seminar 11/11/21] Energy-Efficient AI ASIC Designs: CNN Accelerator and LSTM Accelerator

Randomization methods for 1:00:59

Randomization methods for big-data with Stephen Becker

EuroMPI 2022 Sess. 0:31:14

EuroMPI 2022 Sess. III - Correctness and Error Handling: 'Error-Handling in MPI Continuations&a...