AMD ROCm: torch.backends.cudnn.benchmark
should be set to False
by default on ROCm
#2552
Labels
bug
Something isn't working
Describe the bug
With AMD HIP,
torch.backends.cudnn.benchmark
defaults toTrue
. On CUDA, it defaults toFalse
.While this is an upstream PyTorch decision (or bug), having
benchmark
default toTrue
by default is probably a bad idea in our case, because we have wildly different input shapes. Since CNN kernel benchmarking is performed on a per-shape basis, it performs those slow benchmarks very often, so training times are worsened rather than improved.Probably should add some common per-backend quirks file for this kind of change in SB?
Expected behaviour
On AMD ROCm, it should be set to
False
, probably with a warning since we break the default.To Reproduce
No response
Environment Details
No response
Relevant Log Output
No response
Additional Context
No response
The text was updated successfully, but these errors were encountered: