happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

ggml-zdnn: fix incorrect ztensor shape, reduce memory padding 

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
This commit is contained in:
Aaron Teo 2025-07-24 16:22:06 +08:00
parent 828519659b
commit c1653ab639
No known key found for this signature in database
1 changed files with 3 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
ggml/src/ggml-zdnn/ggml-zdnn-rewrite.cpp
View File

@ -108,13 +108,14 @@ static void ggml_zdnn_free(struct ggml_backend_zdnn_context * ctx) {
}
struct ggml_backend_zdnn_buffer {
char name[128];
void * data;
size_t size;
zdnn_tensor_desc pre_tfm_desc;
zdnn_tensor_desc tfm_desc;
zdnn_ztensor ztensor;
char name[GGML_MAX_NAME];
};
struct ggml_backend_zdnn_buffer_context {
@ -240,7 +241,7 @@ static void ggml_zdnn_init_tensor(struct ggml_backend_zdnn_buffer * buffer, stru
ZDNN_2D,
FP32,
&buffer->pre_tfm_desc,
1, 1, tensor->ne[1], tensor->ne[0]
tensor->ne[1], tensor->ne[0]
);
} break;
default: