Merge branch 'develop' of https://github.com/ROCm/composable_kernel into wip-async-tr-fa

This commit is contained in:
aska-0096
2025-08-12 09:04:41 +00:00
12 changed files with 1055 additions and 169 deletions

View File

@@ -525,10 +525,10 @@ bool run(const ck_tile::ArgParser& arg_parser)
flop += nhead * (static_cast<std::size_t>(2) * mask.get_unmaskarea() * hdim_q +
static_cast<std::size_t>(2) * mask.get_unmaskarea() * hdim_v);
num_byte += nhead * (sizeof(QDataType) * real_seqlen_q * hdim_q +
sizeof(KDataType) * real_seqlen_k * hdim_q +
sizeof(VDataType) * hdim_v * real_seqlen_k +
sizeof(ODataType) * real_seqlen_q * hdim_v);
num_byte += nhead * (sizeof(QDataType) * real_seqlen_q * hdim_q +
sizeof(ODataType) * real_seqlen_q * hdim_v);
num_byte += nhead_k * (sizeof(KDataType) * real_seqlen_k * hdim_q +
sizeof(VDataType) * hdim_v * real_seqlen_k);
}
}