diff --git a/python/sglang/test/test_programs.py b/python/sglang/test/test_programs.py index 219ed3cf6..361bbaed0 100644 --- a/python/sglang/test/test_programs.py +++ b/python/sglang/test/test_programs.py @@ -535,7 +535,7 @@ def test_hellaswag_select(): # Compute accuracy accuracy_gen = np.mean(np.array(preds_gen) == np.array(labels)) - assert np.abs(accuracy_gen - accuracy) < 0.05 + assert np.abs(accuracy_gen - accuracy) < 0.1 assert np.abs(latency_gen - latency) < 1 return accuracy, latency