From 499d930fe599486a55f1ccc6fb695983aef52ef9 Mon Sep 17 00:00:00 2001 From: SYLAR <125541396+lishunyang12@users.noreply.github.com> Date: Fri, 31 Jul 2026 11:16:17 +0800 Subject: [PATCH] [Bugfix] Use MiniCPM-o speech template in long audio test (#5623) Signed-off-by: lishunyang12 Co-authored-by: lishunyang12 --- tests/e2e/online_serving/test_minicpmo_4_5_expansion.py | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/tests/e2e/online_serving/test_minicpmo_4_5_expansion.py b/tests/e2e/online_serving/test_minicpmo_4_5_expansion.py index 089a304a7..6e76904b6 100644 --- a/tests/e2e/online_serving/test_minicpmo_4_5_expansion.py +++ b/tests/e2e/online_serving/test_minicpmo_4_5_expansion.py @@ -150,7 +150,16 @@ def test_text_to_audio_long_output_001(omni_server, openai_client) -> None: request_config = { "model": omni_server.model, "messages": messages, + "modalities": ["text", "audio"], "stream": True, + # Delimit the assistant answer with the TTS template so the Talker + # does not spend its codec-token budget speaking hidden reasoning. + "extra_body": { + "chat_template_kwargs": { + "use_tts_template": True, + "enable_thinking": False, + } + }, "key_words": {"audio": ["Beijing"]}, }