Publications de Sebastian Urban Stich

Masked Training of Neural Networks with Partial Gradients

Martin Jaggi, Amirkeivan Mohtashami, Sebastian Urban Stich

State-of-the-art training algorithms for deep learning models are based on stochastic gradient descent (SGD). Recently, many variations have been explored: perturbing parameters for better accuracy (such as in Extra-gradient), limiting SGD updates to a sub ...

JMLR-JOURNAL MACHINE LEARNING RESEARCH2022

Stochastic distributed learning with gradient quantization and double-variance reduction

Sebastian Urban Stich, Konstantin Mishchenko

We consider distributed optimization over several devices, each sending incremental model updates to a central server. This setting is considered, for instance, in federated learning. Various schemes have been designed to compress the model updates in orde ...

TAYLOR & FRANCIS LTD2022

Quasi-Global Momentum: Accelerating Decentralized Deep Learning on Heterogeneous Data

Martin Jaggi, Tao Lin, Sebastian Urban Stich, Sai Praneeth Reddy Karimireddy

Decentralized training of deep learning models is a key element for enabling data privacy and on-device learning over networks. In realistic learning scenarios, the presence of heterogeneity across different clients' local datasets poses an optimization ch ...

JMLR-JOURNAL MACHINE LEARNING RESEARCH2021

LENA: Communication-Efficient Distributed Learning with Self-Triggered Gradient Uploads

Martin Jaggi, Sebastian Urban Stich, Hossein Shokri Ghadikolaei

In distributed optimization, parameter updates from the gradient computing node devices have to be aggregated in every iteration on the orchestrating server. When these updates are sent over an arbitrary commodity network, bandwidth and latency can be limi ...

MICROTOME PUBLISHING2021

Consensus Control for Decentralized Deep Learning

Martin Jaggi, Tao Lin, Sebastian Urban Stich, Anastasiia Koloskova, Lingjing Kong

Decentralized training of deep learning models enables on-device learning over networks, as well as efficient scaling to large compute clusters. Experiments in earlier works reveal that, even in a data-center setup, decentralized training often suffers fro ...

JMLR-JOURNAL MACHINE LEARNING RESEARCH2021

Semantic Perturbations with Normalizing Flows for Improved Generalization

Martin Jaggi, Tatjana Chavdarova, Sebastian Urban Stich

Data augmentation is a widely adopted technique for avoiding overfitting when training deep neural networks. However, this approach requires domain-specific knowledge and is often limited to a fixed set of hard-coded transformations. Recently, several work ...

IEEE2021

Advances and Open Problems in Federated Learning

Martin Jaggi, Sebastian Urban Stich, Yang Liu, Lie He, Ayfer Özgür Aydin, Florian Tramèr, Qiang Yang, Ananda Theertha Suresh, Badih Ghazi

Federated learning (FL) is a machine learning setting where many clients (e.g., mobile devices or whole organizations) collaboratively train a model under the orchestration of a central server (e.g., service provider), while keeping the training data decen ...

NOW PUBLISHERS INC2021

Critical Parameters for Scalable Distributed Learning with Large Batches and Asynchronous Updates

Martin Jaggi, Amirkeivan Mohtashami, Sebastian Urban Stich

It has been experimentally observed that the efficiency of distributed training with stochastic gradient (SGD) depends decisively on the batch size and—in asynchronous implementations—on the gradient staleness. Especially, it has been observed that the spe ...

MICROTOME PUBLISHING2021

The Error-Feedback Framework: Better Rates for SGD with Delayed Gradients and Compressed Updates

Sai Praneeth Reddy Karimireddy, Sebastian Urban Stich

We analyze (stochastic) gradient descent (SGD) with delayed updates on smooth quasi-convex and non-convex functions and derive concise, non-asymptotic, convergence rates. We show that the rate of convergence in all cases consists of two terms: (i) a stocha ...

Microtome Publishing2020

SCAFFOLD Stochastic Controlled Averaging for Federated Learning

Sai Praneeth Reddy Karimireddy, Sebastian Urban Stich, Ananda Theertha Suresh

Federated Averaging (FEDAVG) has emerged as the algorithm of choice for federated learning due to its simplicity and low communication cost. However, in spite of recent research efforts, its performance is not fully understood. We obtain tight convergence ...

JMLR-JOURNAL MACHINE LEARNING RESEARCH2020