Demjan Grubic

QSGD: Communication-Optimal Stochastic Gradient Descent, with Applications to Training Neural Networks

Dan Alistarh, Jerry Li, Ryota Tomioka, 2016, 1610.02132.

QSGD: Communication-Optimal Stochastic Gradient Descent, with Applications to Training Neural Networks

M. Vojnovic, Dan Alistarh, Demjan Grubic, 2016, 1610.02132.

QSGD: Communication-Optimal Stochastic Gradient Descent, with Applications to Training Neural Networks

M. Vojnovic, Dan Alistarh, Demjan Grubic, 2016, 1610.02132.

QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding

Dan Alistarh, Demjan Grubic, Jerry Li, 2016, NIPS.

QSGD: Communication-Optimal Stochastic Gradient Descent, with Applications to Training Neural Networks

Dan Alistarh, Demjan Grubic, Jerry Li, 2016, 1610.02132.

Synchronous Multi-GPU Deep Learning with Low-Precision Communication: An Experimental Study

Dan Alistarh, Ce Zhang, Leo K. Tam, 2018 .