vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-16 03:49:42 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

rename define to match upstream

Browse source
This commit is contained in:
Concedo 2025-05-23 17:10:12 +08:00
parent 22ef97d7d3
commit 499283c63a
6 changed files with 13 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

6
otherarch/sdcpp/ggml_extend.hpp
View file

@ -28,7 +28,7 @@
#include "model.h"
#ifdef SD_USE_CUBLAS
#ifdef SD_USE_CUDA
#include "ggml-cuda.h"
#endif
@ -782,7 +782,7 @@ __STATIC_INLINE__ struct ggml_tensor* ggml_nn_attention(struct ggml_context* ctx
struct ggml_tensor* k,
struct ggml_tensor* v,
bool mask = false) {
#if defined(SD_USE_FLASH_ATTENTION) && !defined(SD_USE_CUBLAS) && !defined(SD_USE_METAL) && !defined(SD_USE_VULKAN) && !defined(SD_USE_SYCL)
#if defined(SD_USE_FLASH_ATTENTION) && !defined(SD_USE_CUDA) && !defined(SD_USE_METAL) && !defined(SD_USE_VULKAN) && !defined(SD_USE_SYCL)
struct ggml_tensor* kqv = ggml_flash_attn(ctx, q, k, v, false); // [N * n_head, n_token, d_head]
#else
float d_head = (float)q->ne[0];
@ -938,7 +938,7 @@ __STATIC_INLINE__ struct ggml_tensor* ggml_nn_group_norm(struct ggml_context* ct
}
__STATIC_INLINE__ void ggml_backend_tensor_get_and_sync(ggml_backend_t backend, const struct ggml_tensor* tensor, void* data, size_t offset, size_t size) {
#if defined(SD_USE_CUBLAS) || defined(SD_USE_SYCL)
#if defined(SD_USE_CUDA) || defined(SD_USE_SYCL)
if (!ggml_backend_is_cpu(backend)) {
ggml_backend_tensor_get_async(backend, tensor, data, offset, size);
ggml_backend_synchronize(backend);