FutureWarning: torch.cuda.amp.custom_bwd(args...)
is deprecated. Please use torch.amp.custom_bwd(args..., device_type='cuda')
instead.
#20370
Labels
Bug description
I'm getting the error
specifically the issue seems to be caused by
fairscale
:What version are you seeing the problem on?
v2.4
How to reproduce the bug
python code is almost standard distributed training code with
FSDPStrategy
train strategy and it was working before:Error messages and logs
Undefined number of following logging
Environment
More info
No response
The text was updated successfully, but these errors were encountered: