Error Compensated Distributed SGD can be Accelerated
–Neural Information Processing Systems
In this work, we show for the first time that error compensated gradient compression methods can be accelerated. In particular, we propose and study the error compensated loopless Katyusha method, and establish an accelerated linear convergence rate under standard assumptions.
Neural Information Processing Systems
Aug-19-2025, 03:05:10 GMT
- Country:
- North America > United States
- Virginia (0.04)
- Europe > Russia
- Central Federal District > Moscow Oblast > Moscow (0.04)
- Asia
- China > Hong Kong (0.04)
- Russia (0.04)
- Middle East
- Jordan (0.04)
- Saudi Arabia > Mecca Province
- Thuwal (0.04)
- North America > United States
- Genre:
- Research Report > New Finding (0.46)
- Technology: