mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-11-08 10:07:01 +00:00
@@ -224,16 +224,12 @@ static enum ggml_status ggml_zdnn_graph_compute(ggml_backend_t backend, struct g
|
||||
}
|
||||
|
||||
GGML_ASSERT(ok);
|
||||
|
||||
}
|
||||
|
||||
return GGML_STATUS_SUCCESS;
|
||||
}
|
||||
|
||||
static void ggml_zdnn_init_tensor(struct ggml_backend_zdnn_buffer * buffer, struct ggml_tensor * tensor) {
|
||||
// Initialize tensor descriptors for this tensor
|
||||
// TODO: Set up pre_tfm_desc, tfm_desc, and ztensor based on tensor properties
|
||||
|
||||
switch (tensor->op) {
|
||||
case GGML_OP_MUL_MAT:
|
||||
{
|
||||
|
||||
Reference in New Issue
Block a user