profiler/include/profiler/profile_layernorm_impl.hpp · aba0880dd5badf8411b503e24e4c134fc6125fbe · gaoqiong / composable_kernel

rocking authored Aug 12, 2023

* Add normalization splitK to layernorm and groupnorm instances

* Fix bug of GetKPerThread()

* Refine naming

* clang format

03b8119e

profile_layernorm_impl.hpp 9.3 KB