Alert button
Picture for Michael Crawshaw

Michael Crawshaw

Alert button

EPISODE: Episodic Gradient Clipping with Periodic Resampled Corrections for Federated Learning with Heterogeneous Data

Feb 14, 2023
Michael Crawshaw, Yajie Bao, Mingrui Liu

Figure 1 for EPISODE: Episodic Gradient Clipping with Periodic Resampled Corrections for Federated Learning with Heterogeneous Data
Figure 2 for EPISODE: Episodic Gradient Clipping with Periodic Resampled Corrections for Federated Learning with Heterogeneous Data
Figure 3 for EPISODE: Episodic Gradient Clipping with Periodic Resampled Corrections for Federated Learning with Heterogeneous Data
Figure 4 for EPISODE: Episodic Gradient Clipping with Periodic Resampled Corrections for Federated Learning with Heterogeneous Data
Viaarxiv icon

Robustness to Unbounded Smoothness of Generalized SignSGD

Aug 23, 2022
Michael Crawshaw, Mingrui Liu, Francesco Orabona, Wei Zhang, Zhenxun Zhuang

Figure 1 for Robustness to Unbounded Smoothness of Generalized SignSGD
Figure 2 for Robustness to Unbounded Smoothness of Generalized SignSGD
Figure 3 for Robustness to Unbounded Smoothness of Generalized SignSGD
Figure 4 for Robustness to Unbounded Smoothness of Generalized SignSGD
Viaarxiv icon

Fast Composite Optimization and Statistical Recovery in Federated Learning

Jul 17, 2022
Yajie Bao, Michael Crawshaw, Shan Luo, Mingrui Liu

Figure 1 for Fast Composite Optimization and Statistical Recovery in Federated Learning
Figure 2 for Fast Composite Optimization and Statistical Recovery in Federated Learning
Figure 3 for Fast Composite Optimization and Statistical Recovery in Federated Learning
Figure 4 for Fast Composite Optimization and Statistical Recovery in Federated Learning
Viaarxiv icon

SLAW: Scaled Loss Approximate Weighting for Efficient Multi-Task Learning

Sep 16, 2021
Michael Crawshaw, Jana Košecká

Figure 1 for SLAW: Scaled Loss Approximate Weighting for Efficient Multi-Task Learning
Figure 2 for SLAW: Scaled Loss Approximate Weighting for Efficient Multi-Task Learning
Figure 3 for SLAW: Scaled Loss Approximate Weighting for Efficient Multi-Task Learning
Figure 4 for SLAW: Scaled Loss Approximate Weighting for Efficient Multi-Task Learning
Viaarxiv icon

Multi-Task Learning with Deep Neural Networks: A Survey

Sep 10, 2020
Michael Crawshaw

Figure 1 for Multi-Task Learning with Deep Neural Networks: A Survey
Figure 2 for Multi-Task Learning with Deep Neural Networks: A Survey
Figure 3 for Multi-Task Learning with Deep Neural Networks: A Survey
Figure 4 for Multi-Task Learning with Deep Neural Networks: A Survey
Viaarxiv icon