CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-10 10:27:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

ggml-zdnn: disable batched matmul

Browse Source 
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
This commit is contained in:
Aaron Teo
2025-07-31 12:10:14 +08:00
parent 12e6b8b65d
commit 732df731ba
1 changed files with 3 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/src/ggml-zdnn/ggml-zdnn.cpp
View File

@@ -275,7 +275,9 @@ static bool ggml_zdnn_supports_op(const ggml_backend_zdnn_device_context * ctx_d
const int64_t max_batch = ctx_dev->max_size;
return ggml_is_contiguous(src0) &&
return ggml_is_matrix(src0) &&
ggml_is_matrix(src1) &&
ggml_is_contiguous(src0) &&
ggml_is_contiguous(src1) &&
src0->type == GGML_TYPE_F32 && src1->type == GGML_TYPE_F32 &&
(ne0 <= max_batch && ne1 <= max_batch && ne10 <= max_batch);