vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

llama/kompute: Add multi-GPU support

Browse source 
Signed-off-by: Feng Jiang <jiangfeng@kylinos.cn>
This commit is contained in:
Feng Jiang 2024-08-21 16:52:11 +08:00
parent 56c5f988eb
commit 424e3a52fe
1 changed files with 35 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

39
src/llama.cpp
View file

@ -2856,6 +2856,8 @@ static size_t llama_get_device_count(const llama_model & model) {
count = ggml_backend_sycl_get_device_count(); count = ggml_backend_sycl_get_device_count();
#elif defined(GGML_USE_VULKAN) #elif defined(GGML_USE_VULKAN)
count = ggml_backend_vk_get_device_count(); count = ggml_backend_vk_get_device_count();
#elif defined(GGML_USE_KOMPUTE)
count = ggml_backend_kompute_get_device_count();
#elif defined(GGML_USE_CANN) #elif defined(GGML_USE_CANN)
return ggml_backend_cann_get_device_count(); return ggml_backend_cann_get_device_count();
#endif #endif
@ -2952,6 +2954,11 @@ static size_t llama_get_device_memory(const llama_model & model, int device) {
size_t free; size_t free;
ggml_backend_vk_get_device_memory(device, &free, &total); ggml_backend_vk_get_device_memory(device, &free, &total);
return free; return free;
#elif defined(GGML_USE_KOMPUTE)
size_t total;
size_t free;
ggml_backend_kompute_get_device_memory(device, &free, &total);
return free;
#elif defined(GGML_USE_CANN) #elif defined(GGML_USE_CANN)
size_t total; size_t total;
size_t free; size_t free;
@ -16899,6 +16906,8 @@ size_t llama_max_devices(void) {
return GGML_SYCL_MAX_DEVICES; return GGML_SYCL_MAX_DEVICES;
#elif defined(GGML_USE_VULKAN) #elif defined(GGML_USE_VULKAN)
return GGML_VK_MAX_DEVICES; return GGML_VK_MAX_DEVICES;
#elif defined(GGML_USE_KOMPUTE)
return GGML_KOMPUTE_MAX_DEVICES;
#elif defined(GGML_USE_CANN) #elif defined(GGML_USE_CANN)
return GGML_CANN_MAX_DEVICES; return GGML_CANN_MAX_DEVICES;
#else #else
@ -17234,13 +17243,35 @@ struct llama_context * llama_new_context_with_model(
} }
#elif defined(GGML_USE_KOMPUTE) #elif defined(GGML_USE_KOMPUTE)
if (model->n_gpu_layers > 0) { if (model->n_gpu_layers > 0) {
auto * backend = ggml_backend_kompute_init(model->main_gpu); if (model->split_mode == LLAMA_SPLIT_MODE_NONE) {
if (backend == nullptr) { auto * backend = ggml_backend_kompute_init(model->main_gpu);
LLAMA_LOG_ERROR("%s: failed to initialize Kompute backend\n", __func__); if (!backend) {
LLAMA_LOG_ERROR("%s: failed to initialize Kompute backend\n", __func__);
llama_free(ctx);
return nullptr;
}
ctx->backends.push_back(backend);
} else if (model->split_mode == LLAMA_SPLIT_MODE_LAYER) {
size_t count = 0;
auto * devices =ggml_vk_available_devices(0, &count);
for (size_t i = 0; i < count; i++) {
LLAMA_LOG_INFO("Kompute: Found device #%d, %s, %s, max-alloc %ld, heap-size %lu\n",
devices[i].index, devices[i].vendor, devices[i].name,
devices[i].maxAlloc, devices[i].heapSize);
auto * backend = ggml_backend_kompute_init(devices[i].index);
if (!backend) {
LLAMA_LOG_ERROR("%s: failed to initialize Kompute backend\n", __func__);
llama_free(ctx);
return nullptr;
}
ctx->backends.push_back(backend);
}
free(devices);
} else {
LLAMA_LOG_ERROR("%s: Failed to init Kompute backend: split mode %d not supported\n", __func__, model->split_mode);
llama_free(ctx); llama_free(ctx);
return nullptr; return nullptr;
} }
ctx->backends.push_back(backend);
} }
#elif defined(GGML_USE_CANN) #elif defined(GGML_USE_CANN)
// with split_mode LLAMA_SPLIT_MODE_NONE or LLAMA_SPLIT_MODE_ROW, only the main GPU backend is used // with split_mode LLAMA_SPLIT_MODE_NONE or LLAMA_SPLIT_MODE_ROW, only the main GPU backend is used