v0.3.6
Merge branch 'main' into sync/npu
[example] make gpt example directory more clear (#2353)
[example]add gpt2 benchmark example script. (#5295)
[devops] fix extention building (#5427)
[hotfix] fix typo of openmoe model source (#5403)
[bug] fix get_default_parser in examples (#4764)
[npu] change device to accelerator api (#5239)