ROCm / composable_kernel

Composable Kernel: Performance Portable Programming Model for Machine Learning Tensor Operators
https://rocm.docs.amd.com/projects/composable_kernel/en/latest/
Other
308 stars 125 forks source link

Backward weight convolution kernel error when enable profiling #1184

Closed iq136boy closed 1 month ago

iq136boy commented 8 months ago

We found the backward weight convolution kernels will lead to errors when enable profiling for the ck invoker run() functions, which made the ck-based solver failed in MIOpen.
We have the following observation: the ck-profiling is enabled, the ck wrw kernel introduces the errors that cause the precision issue. The ck-profiling is disabled, the result is correct but the time will be 0 in this case. This lead to a situation: we can either get the correct profiling information OR get the correct result from ck, but not both of them. I have proposed a workaround: PR2770 in MIOpen to address this issue. It would be nice if CK can do some fix to let us get both the profiling info and result simultaneously.

ppanchad-amd commented 2 months ago

@iq136boy Is this fixed in the latest ROCm 6.2? If so, please close the ticket. Thanks!

huanrwan-amd commented 1 month ago

Close the ticket as @iq136boy will use their own profiling tool.