vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-11 01:24:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

move position id remap out of ggml to avoid int32 cuda operations

Browse source
This commit is contained in:
HimariO 2025-03-16 00:35:19 +08:00
parent e18f6a3238
commit c891300c1e
2 changed files with 108 additions and 43 deletions
Download patch file Download diff file Expand all files Collapse all files

48
examples/llava/qwen2vl-cli.cpp
View file

@ -370,14 +370,14 @@ static void debug_test_mrope_2d() {
// 1. Initialize backend
ggml_backend_t backend = NULL;
std::string backend_name = "";
#ifdef GGML_USE_CUDA
fprintf(stderr, "%s: using CUDA backend\n", __func__);
backend = ggml_backend_cuda_init(0); // init device 0
backend_name = "cuda";
if (!backend) {
fprintf(stderr, "%s: ggml_backend_cuda_init() failed\n", __func__);
}
#endif
// #ifdef GGML_USE_CUDA
// fprintf(stderr, "%s: using CUDA backend\n", __func__);
// backend = ggml_backend_cuda_init(0); // init device 0
// backend_name = "cuda";
// if (!backend) {
// fprintf(stderr, "%s: ggml_backend_cuda_init() failed\n", __func__);
// }
// #endif
// if there aren't GPU Backends fallback to CPU backend
if (!backend) {
backend = ggml_backend_cpu_init();
@ -490,14 +490,14 @@ static void debug_patch_layout() {
// 1. Initialize backend
ggml_backend_t backend = NULL;
std::string backend_name = "";
#ifdef GGML_USE_CUDA
fprintf(stderr, "%s: using CUDA backend\n", __func__);
backend = ggml_backend_cuda_init(0); // init device 0
backend_name = "cuda";
if (!backend) {
fprintf(stderr, "%s: ggml_backend_cuda_init() failed\n", __func__);
}
#endif
// #ifdef GGML_USE_CUDA
// fprintf(stderr, "%s: using CUDA backend\n", __func__);
// backend = ggml_backend_cuda_init(0); // init device 0
// backend_name = "cuda";
// if (!backend) {
// fprintf(stderr, "%s: ggml_backend_cuda_init() failed\n", __func__);
// }
// #endif
// if there aren't GPU Backends fallback to CPU backend
if (!backend) {
backend = ggml_backend_cpu_init();
@ -615,14 +615,14 @@ static void debug_test_get_rows() {
// 1. Initialize backend
ggml_backend_t backend = NULL;
std::string backend_name = "";
#ifdef GGML_USE_CUDA
fprintf(stderr, "%s: using CUDA backend\n", __func__);
backend = ggml_backend_cuda_init(0); // init device 0
backend_name = "cuda";
if (!backend) {
fprintf(stderr, "%s: ggml_backend_cuda_init() failed\n", __func__);
}
#endif
// #ifdef GGML_USE_CUDA
// fprintf(stderr, "%s: using CUDA backend\n", __func__);
// backend = ggml_backend_cuda_init(0); // init device 0
// backend_name = "cuda";
// if (!backend) {
// fprintf(stderr, "%s: ggml_backend_cuda_init() failed\n", __func__);
// }
// #endif
// if there aren't GPU Backends fallback to CPU backend
if (!backend) {
backend = ggml_backend_cpu_init();