AutoMoE: Neural Architecture Search for Efficient Sparsely Activated Transformers.
Ganesh JawaharSubhabrata MukherjeeXiaodong LiuYoung Jin KimMuhammad Abdul-MageedLaks V. S. LakshmananAhmed Hassan AwadallahSébastien BubeckJianfeng GaoPublished in: CoRR (2022)