Login / Signup

Impact of RoCE Congestion Control Policies on Distributed Training of DNNs.

Tarannum KhanSaeed RashidiSrinivas SridharanPallavi ShurpaliAditya AkellaTushar Krishna
Published in: HOTI (2022)
Keyphrases
  • control policies
  • multi agent
  • cooperative
  • finite horizon
  • optimal policy
  • decision making
  • control system
  • supervised learning
  • multistage
  • real time
  • neural network
  • action space
  • stochastic optimization problems