v0.3.9
[misc] refactor launch API and tensor constructor (#5666)
[example] make gpt example directory more clear (#2353)
[chore] refactor profiler utils
Support 4d parallel + flash attention (#5789)
[hotfix] fix OpenMOE example import path (#5697)
[example]add gpt2 benchmark example script. (#5295)
[devops] remove post commit ci (#5566)