Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2025-01-01 00:39:00 +01:00
Code Issues Packages Projects Releases Wiki Activity

cuda : add comment

Browse Source
This commit is contained in:
Georgi Gerganov 2023-12-05 18:15:51 +02:00
parent dd86df82e6
commit 4adb1d69d9
No known key found for this signature in database
GPG Key ID: 449E073F9DC10735
1 changed files with 1 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
ggml-cuda.cu
View File

@ -7849,6 +7849,7 @@ static void ggml_cuda_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1
#endif // GGML_CUDA_FORCE_DMMV #endif // GGML_CUDA_FORCE_DMMV
if (use_mul_mat_vec_q) { if (use_mul_mat_vec_q) {
// NOTE: this kernel does not support ggml_nrows(src1) > 1
ggml_cuda_op_mul_mat(src0, src1, dst, ggml_cuda_op_mul_mat_vec_q, true); ggml_cuda_op_mul_mat(src0, src1, dst, ggml_cuda_op_mul_mat_vec_q, true);
} else { } else {
ggml_cuda_op_mul_mat(src0, src1, dst, ggml_cuda_op_dequantize_mul_mat_vec, false); ggml_cuda_op_mul_mat(src0, src1, dst, ggml_cuda_op_dequantize_mul_mat_vec, false);