Details
Title
Stich, Sebastian Urban
Sciper ID
278401
Affiliated labs
MLO
Publications
Accelerated Stochastic Matrix Inversion: General Theory and Speeding up BFGS Rules for Faster Second-Order Optimization
Critical Parameters for Scalable Distributed Learning with Large Batches and Asynchronous Updates
Don't Use Large Mini-Batches, Use Local SGD
Dynamic Model Pruning with Feedback
Ensemble Distillation for Robust Model Fusion in Federated Learning
Error Feedback Fixes SignSGD and other Gradient Compression Schemes
Extrapolation for Large-batch Training in Deep Learning
Local SGD Converges Fast and Communicates Little
On Matching Pursuit and Coordinate Descent
Safe Adaptive Importance Sampling
See complete list of publications (18)
Critical Parameters for Scalable Distributed Learning with Large Batches and Asynchronous Updates
Don't Use Large Mini-Batches, Use Local SGD
Dynamic Model Pruning with Feedback
Ensemble Distillation for Robust Model Fusion in Federated Learning
Error Feedback Fixes SignSGD and other Gradient Compression Schemes
Extrapolation for Large-batch Training in Deep Learning
Local SGD Converges Fast and Communicates Little
On Matching Pursuit and Coordinate Descent
Safe Adaptive Importance Sampling
See complete list of publications (18)
Use for
Stich, Sebastian
Link to search
All resources
Record appears in
Authorities > People