fmha_fwd_hdim32.cu 720 Bytes
Newer Older
1
2
// Copyright (c) 2022, Tri Dao.

Tri Dao's avatar
Tri Dao committed
3
// Splitting the different head dimensions to different files to speed up compilation.
4
5
6
7

#include "fmha_fwd_launch_template.h"

void run_fmha_fwd_hdim32(Launch_params<FMHA_fprop_params> &launch_params) {
8
    FP16_SWITCH(launch_params.params.is_bf16, ({
9
10
11
12
13
14
15
        if (launch_params.params.seqlen_k == 128) {
            using Kernel_traits = FMHA_kernel_traits<128, 32, 16, 1, 4, 0x08u, elem_type>;
            run_fmha_fwd_loop<Kernel_traits>(launch_params);
        } else if (launch_params.params.seqlen_k >= 256) {
            using Kernel_traits = FMHA_kernel_traits<256, 32, 16, 1, 4, 0x08u, elem_type>;
            run_fmha_fwd_loop<Kernel_traits>(launch_params);
        }
16
    }));
17
}