From 8700f86a7012b2354c3f2834c1a0d2b6023b0bcd Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Nicol=C3=B2=20Lucchesi?= Date: Fri, 31 Jul 2026 01:18:53 +0200 Subject: [PATCH] [CI] Fix `tests/entrypoints/multimodal/openai/chat_completion/test_audio.py::test_chat_streaming_audio` (#50451) Signed-off-by: NickLucche --- .../multimodal/openai/chat_completion/test_audio.py | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/tests/entrypoints/multimodal/openai/chat_completion/test_audio.py b/tests/entrypoints/multimodal/openai/chat_completion/test_audio.py index a1c13d9e33..d806100e8e 100644 --- a/tests/entrypoints/multimodal/openai/chat_completion/test_audio.py +++ b/tests/entrypoints/multimodal/openai/chat_completion/test_audio.py @@ -10,7 +10,6 @@ import pytest_asyncio from tests.utils import RemoteOpenAIServer from vllm.assets.audio import AudioAsset from vllm.multimodal.utils import encode_audio_base64, encode_audio_url, fetch_audio -from vllm.platforms import current_platform MODEL_NAME = "fixie-ai/ultravox-v0_5-llama-3_2-1b" TEST_AUDIO_URLS = [ @@ -19,10 +18,6 @@ TEST_AUDIO_URLS = [ ] MAXIMUM_AUDIOS = 2 -# Disable prefix caching on ROCm to reduce non-determinism in -# streaming-vs-non-streaming comparisons. -_ROCM_ARGS = ["--no-enable-prefix-caching"] if current_platform.is_rocm() else [] - @pytest.fixture(scope="module") def server(): @@ -37,7 +32,7 @@ def server(): "--trust-remote-code", "--limit-mm-per-prompt", json.dumps({"audio": MAXIMUM_AUDIOS}), - *_ROCM_ARGS, + "--no-enable-prefix-caching", ] with RemoteOpenAIServer(MODEL_NAME, args) as remote_server: