Run DDP from a flast/api endpoint #10477
Unanswered
FrancescoSaverioZuppichini
asked this question in
DDP / multi-GPU / multi-node
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Hi Guys,
I have the following training routine
It works like a charm but as soon as I try to wrap it into an endpoint (Flask or FastAPI) the ddp is not able to start the processes, no errors it just hangs there
same thing with Flask
I believe this problem is linked to how distributed training works but I don't know how to solve it
Thanks
Francesco
Beta Was this translation helpful? Give feedback.
All reactions