Y
Hacker News
new
|
ask
|
show
|
jobs
Revisiting Small Batch Training for Deep NNs; batch_size less than 64 is best
(
arxiv.org
)
1 points
by
evc123
2982 days ago