| // Copyright (c) 2024, Tri Dao. | |
| // Splitting the different head dimensions to different files to speed up compilation. | |
| // This file is auto-generated. See "generate_kernels.py" | |
| namespace FLASH_NAMESPACE { | |
| template<> | |
| void run_mha_fwd_<cutlass::bfloat16_t, 256, true>(Flash_fwd_params ¶ms, cudaStream_t stream) { | |
| run_mha_fwd_hdim256<cutlass::bfloat16_t, true>(params, stream); | |
| } | |
| } // namespace FLASH_NAMESPACE |