Illia Silin 6a4cfe125b Remove the workaround for bf16 attention tests. (#586) ...

* remove workanround in bf16 attention test

* clean up another workaround

[ROCm/composable_kernel commit: 06f1fc864c]

2023-02-14 18:06:24 -06:00

..

block

[Navi3x-LWPCK-545] Block-wise GEMM + Real GEMM_WMMA_FP16 (#541)

2023-01-16 20:06:01 -06:00

device

Gemm+layernorm instance, ckProfiler, client example (#568)

2023-02-09 15:02:55 -06:00

element

Add GemmAddSoftmaxGemm support for MSFT ORT (instances and client API) (#576)

2023-02-08 14:34:45 -06:00

grid

Remove the workaround for bf16 attention tests. (#586)

2023-02-14 18:06:24 -06:00

thread

Batchnorm-forward implemented using welford method to calculate variance (#403)

2022-10-27 18:52:54 -06:00

warp

[Navi3x-LWPCK-545] Block-wise GEMM + Real GEMM_WMMA_FP16 (#541)

2023-01-16 20:06:01 -06:00