search

megvii-model / DetNAS

MIT License
291 stars 50 forks source link

only one GPU, I met the error that "Default process group is not initialized" #35

Closed nanzh-19 closed 2 years ago

nanzh-19 commented 3 years ago

I only have one GPU, and when I trained the supernet on coco, there was an error that "AssertionError: Default process group is not initialized" I think that i should not use "SyncBN" to distributed calculating. how can I set it up so I can train with only one GPU?

yukang2017 commented 2 years ago

Sorry for the late reply. I think you can replace SyncBN with GN when using only one GPU.