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: add guards to prevent loading ztensor if transformed 

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
This commit is contained in:
Aaron Teo 2025-07-30 16:15:20 +08:00
parent 4d5edb2221
commit b7a77cf683
No known key found for this signature in database
1 changed files with 7 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
ggml/src/ggml-zdnn/ggml-zdnn-rewrite.cpp
View File

@ -133,8 +133,13 @@ static void ggml_zdnn_mul_mat_op(ggml_backend_zdnn_context * ctx, const ggml_ten
// ggml_zdnn_create_tensor(ptd_output, td_output, zt_output, output, output_dim, ZDNN_2D);
void * bias_data = (void *)calloc(ne0, ggml_element_size(output));
ggml_zdnn_load_tensor(weights_extra->ztensor, weights->data);
ggml_zdnn_load_tensor(inputs_extra->ztensor, inputs->data);
if (weights_extra->ztensor.is_transformed == false) {
ggml_zdnn_load_tensor(weights_extra->ztensor, weights->data);
}
if (inputs_extra->ztensor.is_transformed == false) {
ggml_zdnn_load_tensor(inputs_extra->ztensor, inputs->data);
}
ggml_zdnn_load_tensor(zt_bias, bias_data);
// ggml_zdnn_load_tensor(output_extra->ztensor, output->data);