Skip to content

LLM inference in C/C++

C++ ggml

tests : initialize all tensors in test_dsv4_hc to avoid NaNs in sentinel tensors (#25822)

fairydreaming5d5306bf3e2d
diff --git a/tests/test-backend-ops.cpp b/tests/test-backend-ops.cpp
index 5b5f196184..e3553c5e9c 100644
--- a/tests/test-backend-ops.cpp
+++ b/tests/test-backend-ops.cpp
@@ -3802,6 +3802,7 @@ struct test_dsv4_hc : public test_case {
float lo;
float hi;
if (!tensor_range(name, lo, hi)) {
+ init_tensor_uniform(t);
continue;
}