On the Training Instability of Shuffling SGD with Batch Normalization.

David Xing Wu Chulhee Yun Suvrit Sra

Published in: ICML (2023)

Keyphrases

stochastic gradient descent
batch mode
training process
support vector
preprocessing
training set
test set
online algorithms
information retrieval
information systems
pairwise
training algorithm
batch size