mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-23 22:34:36 +00:00
fixed multiple definition issue of bfp16/fp32 conversion function when building ckProfiler (#51)
* fixed bfloat16 issues
* refactor type_convert
Co-authored-by: Chao Liu <chao.liu2@amd.com>
[ROCm/composable_kernel commit: 0a66c54e95]
This commit is contained in:
@@ -106,12 +106,12 @@ void profile_conv(int do_verification,
|
||||
{
|
||||
case 0: break;
|
||||
case 1:
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_2{-5, 5});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_2{-5, 5});
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_2<InDataType>{-5, 5});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_2<WeiDataType>{-5, 5});
|
||||
break;
|
||||
default:
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_3<float>{0.0, 1.0});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_3<float>{-0.5, 0.5});
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_3<InDataType>{0.0, 1.0});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_3<WeiDataType>{-0.5, 0.5});
|
||||
}
|
||||
|
||||
if(do_verification)
|
||||
|
||||
@@ -122,12 +122,12 @@ void profile_gemm(int do_verification,
|
||||
{
|
||||
case 0: break;
|
||||
case 1:
|
||||
a_m_k.GenerateTensorValue(GeneratorTensor_2{-5, 5});
|
||||
b_k_n.GenerateTensorValue(GeneratorTensor_2{-5, 5});
|
||||
a_m_k.GenerateTensorValue(GeneratorTensor_2<ADataType>{-5, 5});
|
||||
b_k_n.GenerateTensorValue(GeneratorTensor_2<BDataType>{-5, 5});
|
||||
break;
|
||||
default:
|
||||
a_m_k.GenerateTensorValue(GeneratorTensor_3<float>{0.0, 1.0});
|
||||
b_k_n.GenerateTensorValue(GeneratorTensor_3<float>{-0.5, 0.5});
|
||||
a_m_k.GenerateTensorValue(GeneratorTensor_3<ADataType>{0.0, 1.0});
|
||||
b_k_n.GenerateTensorValue(GeneratorTensor_3<BDataType>{-0.5, 0.5});
|
||||
}
|
||||
|
||||
if(do_verification)
|
||||
|
||||
Reference in New Issue
Block a user