Dynamically Adjusting Transformer Batch Size by Monitoring Gradient Direction Change.

Hongfei Xu Josef van Genabith Deyi Xiong Qiuhui Liu

Published in: CoRR (2020)

Keyphrases

gradient direction
batch size
single item
batch mode
gradient magnitude
poisson process
batch processing
lot sizing
zero crossing
image transformations
tensor field
fixed cost
image matching
service level
grey level