Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2025-01-26 03:12:23 +01:00
Code Issues Packages Projects Releases Wiki Activity

rpc-server : add support for the SYCL backend (#10934)

Browse Source
This commit is contained in:
Radoslav Gerganov 2024-12-23 10:39:30 +02:00 committed by GitHub
parent b92a14a841
commit 86bf31cfe6
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 12 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
examples/rpc/rpc-server.cpp
View File

@ -12,6 +12,10 @@
#include "ggml-vulkan.h" #include "ggml-vulkan.h"
#endif #endif
#ifdef GGML_USE_SYCL
#include "ggml-sycl.h"
#endif
#include "ggml-rpc.h" #include "ggml-rpc.h"
#ifdef _WIN32 #ifdef _WIN32
# include <windows.h> # include <windows.h>
@ -91,6 +95,12 @@ static ggml_backend_t create_backend() {
if (!backend) { if (!backend) {
fprintf(stderr, "%s: ggml_backend_vulkan_init() failed\n", __func__); fprintf(stderr, "%s: ggml_backend_vulkan_init() failed\n", __func__);
} }
#elif GGML_USE_SYCL
fprintf(stderr, "%s: using SYCL backend\n", __func__);
backend = ggml_backend_sycl_init(0); // init device 0
if (!backend) {
fprintf(stderr, "%s: ggml_backend_sycl_init() failed\n", __func__);
}
#endif #endif
// if there aren't GPU Backends fallback to CPU backend // if there aren't GPU Backends fallback to CPU backend
@ -106,6 +116,8 @@ static void get_backend_memory(size_t * free_mem, size_t * total_mem) {
ggml_backend_cuda_get_device_memory(0, free_mem, total_mem); ggml_backend_cuda_get_device_memory(0, free_mem, total_mem);
#elif GGML_USE_VULKAN #elif GGML_USE_VULKAN
ggml_backend_vk_get_device_memory(0, free_mem, total_mem); ggml_backend_vk_get_device_memory(0, free_mem, total_mem);
#elif GGML_USE_SYCL
ggml_backend_sycl_get_device_memory(0, free_mem, total_mem);
#else #else
#ifdef _WIN32 #ifdef _WIN32
MEMORYSTATUSEX status; MEMORYSTATUSEX status;