@misc{indiciaec45100cc30b3,
  title = {Parm: Efficient Training of Large Sparsely-Activated Models with Dedicated Schedules},
  author = {Xinglin Pan and Wenxiang Lin and Shaohuai Shi and Xiaowen Chu and Weinong Sun and Bo Li},
  year = {2024},
  url = {https://arxiv.org/abs/2407.00599},
  note = {Source identifier: 2407.00599}
}