From cfa3234a5b8a6ec4f65ceb080bdc232eaee9b080 Mon Sep 17 00:00:00 2001 From: Wenlong Wang Date: Sat, 13 Sep 2025 00:45:11 -0700 Subject: [PATCH] [CI][Spec Decode] Adjust threshold for flaky ngram spec decoding test again (#24771) Signed-off-by: wwl2755 --- tests/v1/e2e/test_spec_decode.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/v1/e2e/test_spec_decode.py b/tests/v1/e2e/test_spec_decode.py index 0b240b7d434e..bf90f50b1082 100644 --- a/tests/v1/e2e/test_spec_decode.py +++ b/tests/v1/e2e/test_spec_decode.py @@ -117,9 +117,9 @@ def test_ngram_correctness( print(f"ref_output: {ref_output.outputs[0].text}") print(f"spec_output: {spec_output.outputs[0].text}") - # Heuristic: expect at least 68% of the prompts to match exactly + # Heuristic: expect at least 66% of the prompts to match exactly # Upon failure, inspect the outputs to check for inaccuracy. - assert matches >= int(0.68 * len(ref_outputs)) + assert matches >= int(0.66 * len(ref_outputs)) del spec_llm torch.cuda.empty_cache() cleanup_dist_env_and_memory()