[Bugfix] Correctly call cudaProfilerStop in benchmarks script (#14183)

Signed-off-by: Brayden Zhong <b8zhong@uwaterloo.ca>
2025-03-06 19:42:49 -05:00
parent ad60bbb2b2
commit c34eeec58d
6 changed files with 5 additions and 6 deletions
--- a/benchmarks/kernels/benchmark_lora.py
+++ b/benchmarks/kernels/benchmark_lora.py
@@ -153,7 +153,6 @@ def ref_group_gemm(ref_out: torch.Tensor, input: torch.Tensor,
        result = torch.nn.functional.linear(x, w)
        result *= scaling
        out_list.append(result)
-    torch.cat(out_list, dim=0)

    cat_result = torch.cat(out_list, dim=0)