candle/candle-flash-attn/kernels/flash_fwd_hdim160_fp16_sm80.cu

11 lines
392 B
Plaintext

// Copyright (c) 2023, Tri Dao.
// Splitting the different head dimensions to different files to speed up compilation.
// This file is auto-generated. See "generate_kernels.py"
#include "flash_fwd_launch_template.h"
template<>
void run_mha_fwd_<cutlass::half_t, 160, false>(Flash_fwd_params &params, cudaStream_t stream) {
run_mha_fwd_hdim160<cutlass::half_t, false>(params, stream);
}