CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-15 11:17:31 +00:00
Code Issues Packages Projects Releases Wiki Activity

SYCL: Initial set_rows kernel implementation (#14562)

Browse Source 
* SYCL: Initial set_rows kernel implementation

* Revert max_threads to 256

* Refactor set_rows and address review comments

* Deduplicate conversion function

* Remove guard before kernel launch and refactor

* Fix and add back SFINAE
This commit is contained in:
Akarshan Biswas
2025-07-10 13:59:38 +05:30
committed by GitHub
parent 435a6d10d6
commit 704bb7a71c
4 changed files with 145 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
ggml/src/ggml-sycl/ggml-sycl.cpp
View File

@@ -41,6 +41,7 @@
#include "ggml-sycl/element_wise.hpp"
#include "ggml-sycl/presets.hpp"
#include "ggml-sycl/gemm.hpp"
#include "ggml-sycl/set_rows.hpp"
#include "ggml-sycl/sycl_hw.hpp"
#include "ggml-sycl/getrows.hpp"
#include "ggml.h"
@@ -3605,6 +3606,9 @@ static bool ggml_sycl_compute_forward(ggml_backend_sycl_context & ctx, struct gg
case GGML_OP_GET_ROWS:
ggml_sycl_get_rows(ctx, dst);
break;
case GGML_OP_SET_ROWS:
ggml_sycl_op_set_rows(ctx, dst);
break;
case GGML_OP_DUP:
ggml_sycl_dup(ctx, dst);
break;
@@ -4299,7 +4303,7 @@ static bool ggml_backend_sycl_device_supports_op(ggml_backend_dev_t dev, const g
{
// TODO: add support
// ref: https://github.com/ggml-org/llama.cpp/pull/14274
return false;
return (op->type == GGML_TYPE_F32 || (op->type == GGML_TYPE_F16 && op->src[0]->type == GGML_TYPE_F32 && op->src[1]->type == GGML_TYPE_I64));
} break;
case GGML_OP_CPY:
{