Skip to content
This repository was archived by the owner on Jul 4, 2025. It is now read-only.

Commit 9c4e5f0

Browse files
committed
test mmq
1 parent d41d3c5 commit 9c4e5f0

File tree

1 file changed

+1
-1
lines changed

1 file changed

+1
-1
lines changed

controllers/llamaCPP.cc

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -399,7 +399,7 @@ bool llamaCPP::loadModelImpl(const Json::Value &jsonBody) {
399399
}
400400
#ifdef GGML_USE_CUBLAS
401401
LOG_INFO << "Setting up GGML CUBLAS PARAMS";
402-
params.mul_mat_q = false;
402+
params.mul_mat_q = true;
403403
#endif // GGML_USE_CUBLAS
404404
if (params.model_alias == "unknown") {
405405
params.model_alias = params.model;

0 commit comments

Comments
 (0)