ExternalVendorCode/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/whisper.cpp.git synced 2025-06-23 00:48:59 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

SYCL : Move to compile time oneMKL interface backend selection for NVIDIA backend (llama/10584)

Browse Source 
* [SYCL] Move to Compile Time backend selection on oneMKL Interface for NVIDIA backend

Move to compile time selection to backend to avoid latency at run time.
Add it to all mkl gemm calls and only for NVIDIA backend.

Signed-off-by: nscipione <nicolo.scipione@codeplay.com>

* Formatting

* Address PR comments to increase readibility

---------

Signed-off-by: nscipione <nicolo.scipione@codeplay.com>
This commit is contained in:
Nicolò Scipione
2024-12-04 02:29:20 +01:00
committed by Georgi Gerganov
parent 6f1ba9d82d
commit 966433fdf2
4 changed files with 50 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
ggml/src/ggml-sycl/ggml-sycl.cpp
View File

@ -2573,12 +2573,17 @@ inline void ggml_sycl_op_mul_mat_sycl(
const float alpha = 1.0f;
const float beta = 0.0f;
#if !GGML_SYCL_DNNL
# ifdef GGML_SYCL_NVIDIA
SYCL_CHECK(CHECK_TRY_ERROR(oneapi::mkl::blas::column_major::gemm(
*stream, oneapi::mkl::transpose::trans,
oneapi::mkl::transpose::nontrans, row_diff, src1_ncols, ne10,
dpct::get_value(&alpha, *stream), src0_ddf_i, ne00,
src1_ddf1_i, ne10, dpct::get_value(&beta, *stream),
oneapi::mkl::backend_selector<oneapi::mkl::backend::cublas>{ *stream }, oneapi::mkl::transpose::trans,
oneapi::mkl::transpose::nontrans, row_diff, src1_ncols, ne10, dpct::get_value(&alpha, *stream), src0_ddf_i,
ne00, src1_ddf1_i, ne10, dpct::get_value(&beta, *stream), dst_dd_i, ldc)));
# else
SYCL_CHECK(CHECK_TRY_ERROR(oneapi::mkl::blas::column_major::gemm(
*stream, oneapi::mkl::transpose::trans, oneapi::mkl::transpose::nontrans, row_diff, src1_ncols, ne10,
dpct::get_value(&alpha, *stream), src0_ddf_i, ne00, src1_ddf1_i, ne10, dpct::get_value(&beta, *stream),
dst_dd_i, ldc)));
# endif
#else
auto dnnl_stream = ctx.stream_dnnl(stream);
DnnlGemmWrapper::row_gemm(dnnl_stream, false, true, src1_ncols, row_diff, ne10, src1_ddf1_i, DnnlGemmWrapper::to_dt<float>(),