RATQ is a fixed-length quantizer based on random Hadamard rotation and adaptive uniform quantization that nearly attains the information-theoretic lower bound on gradient precision for convex stochastic optimization.
cpSGD: Communication- efficient and differentially-private distributed SGD,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
RATQ: A Universal Fixed-Length Quantizer for Stochastic Optimization
RATQ is a fixed-length quantizer based on random Hadamard rotation and adaptive uniform quantization that nearly attains the information-theoretic lower bound on gradient precision for convex stochastic optimization.