Login / Signup
Parm: Efficient Training of Large Sparsely-Activated Models with Dedicated Schedules.
Xinglin Pan
Wenxiang Lin
Shaohuai Shi
Xiaowen Chu
Weinong Sun
Bo Li
Published in:
INFOCOM (2024)
Keyphrases
</>
accurate models
bayesian framework
model selection
linear model
parameter estimation
scheduling problem
probabilistic model
prior knowledge
website
data sets
online learning
real time
statistical model
statistical models
computationally expensive
multiscale
training phase