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

ggml : fix OpenCL broadcast requirement for ggml_mul (close #4453)

Browse source
This commit is contained in:
Georgi Gerganov 2023-12-14 10:35:29 +02:00
parent 873637afc7
commit 55e87c3749
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
1 changed files with 2 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
ggml.c
View file

@ -7759,10 +7759,10 @@ static void ggml_compute_forward_mul_f32(
const int ith = params->ith;
const int nth = params->nth;
// TODO: OpenCL kernel support broadcast
#ifdef GGML_USE_CLBLAST
if (src1->backend == GGML_BACKEND_GPU) {
GGML_ASSERT(ggml_are_same_shape(src0, src1));
// TODO: OpenCL kernel support full broadcast
GGML_ASSERT(ggml_can_repeat_rows(src1, src0));
if (ith == 0) {
ggml_cl_mul(src0, src1, dst);
}