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

Add support for sqrt on CUDA (#7953)

Browse source 
* cuda sqrt support

* enable cuda in pca

* fix comments in pca

* add test

* add sqrt to ggml_backend_cuda_supports_op

* fix test

* new line

* Use F32 sqrtf instead of F64 sqrt

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>

---------

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
This commit is contained in:
Calvin Laurenson 2024-06-16 15:23:04 -07:00 • committed by GitHub
parent 19b7a836f6
commit 43b35e38ba
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 71 additions and 8 deletions
Download patch file Download diff file Expand all files Collapse all files

16
examples/cvector-generator/pca.hpp
View file

@ -64,15 +64,15 @@ struct pca_model {
struct ggml_tensor * dev_eigenvector;
pca_model(struct ggml_tensor * t_input) {
// TODO: enable GPU support when support for GGML_OP_SQRT is added
// #ifdef GGML_USE_CUDA
// fprintf(stderr, "%s: using CUDA backend\n", __func__);
// backend = ggml_backend_cuda_init(0); // init device 0
// 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
if (!backend) {
fprintf(stderr, "%s: ggml_backend_cuda_init() failed\n", __func__);
}
#endif
// TODO: enable Metal support when support for GGML_OP_SQRT is added
// #ifdef GGML_USE_METAL
// fprintf(stderr, "%s: using Metal backend\n", __func__);
// backend = ggml_backend_metal_init();