kernel
flash-attn3 / flash-attn /instantiations /flash_bwd_hdim192_bf16_softcap_sm90.cu
danieldk's picture
danieldk HF Staff
Convert FA3 to Kernel Hub format
eb8ddce
raw
history blame contribute delete
531 Bytes
// Copyright (c) 2024, Jay Shah, Ganesh Bikshandi, Ying Zhang, Vijay Thakkar, Pradeep Ramani, Tri Dao.
// Splitting the different template instantiations to different files to speed up compilation.
// This file is auto-generated. See "generate_kernels.py"
#include "flash_bwd_launch_template.h"
#ifndef FLASHATTENTION_DISABLE_HDIM192
template<>
void run_mha_bwd_<90, cutlass::bfloat16_t, 192, true>(Flash_bwd_params &params, cudaStream_t stream) {
run_mha_bwd_hdim192<90, cutlass::bfloat16_t, true>(params, stream);
}
#endif