Skip to content

Commit 2bf140c

Browse files
committed
fix(clip): do not imply GPUs by default
Until a better solution is found upstream, be conservative and default to GPU. ggml-org/llama.cpp#12322 ggml-org/llama.cpp#12322 (comment) Signed-off-by: Ettore Di Giacinto <[email protected]>
1 parent 12568c7 commit 2bf140c

File tree

1 file changed

+4
-1
lines changed

1 file changed

+4
-1
lines changed

backend/cpp/llama/grpc-server.cpp

+4-1
Original file line numberDiff line numberDiff line change
@@ -511,7 +511,10 @@ struct llama_server_context
511511
if (!params.mmproj.empty()) {
512512
multimodal = true;
513513
LOG_INFO("Multi Modal Mode Enabled", {});
514-
clp_ctx = clip_model_load(params.mmproj.c_str(), /*verbosity=*/ 1);
514+
clp_ctx = clip_init(params.mmproj.c_str(), clip_context_params {
515+
/* use_gpu */ false,
516+
/*verbosity=*/ 1,
517+
});
515518
if(clp_ctx == nullptr) {
516519
LOG_ERR("unable to load clip model: %s", params.mmproj.c_str());
517520
return false;

0 commit comments

Comments
 (0)