[CI] Fix tests/entrypoints/multimodal/openai/chat_completion/test_audio.py::test_chat_streaming_audio (#50451)
Signed-off-by: NickLucche <nicolo.lucchesi@mistral.ai>
This commit is contained in:
@@ -10,7 +10,6 @@ import pytest_asyncio
|
|||||||
from tests.utils import RemoteOpenAIServer
|
from tests.utils import RemoteOpenAIServer
|
||||||
from vllm.assets.audio import AudioAsset
|
from vllm.assets.audio import AudioAsset
|
||||||
from vllm.multimodal.utils import encode_audio_base64, encode_audio_url, fetch_audio
|
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"
|
MODEL_NAME = "fixie-ai/ultravox-v0_5-llama-3_2-1b"
|
||||||
TEST_AUDIO_URLS = [
|
TEST_AUDIO_URLS = [
|
||||||
@@ -19,10 +18,6 @@ TEST_AUDIO_URLS = [
|
|||||||
]
|
]
|
||||||
MAXIMUM_AUDIOS = 2
|
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")
|
@pytest.fixture(scope="module")
|
||||||
def server():
|
def server():
|
||||||
@@ -37,7 +32,7 @@ def server():
|
|||||||
"--trust-remote-code",
|
"--trust-remote-code",
|
||||||
"--limit-mm-per-prompt",
|
"--limit-mm-per-prompt",
|
||||||
json.dumps({"audio": MAXIMUM_AUDIOS}),
|
json.dumps({"audio": MAXIMUM_AUDIOS}),
|
||||||
*_ROCM_ARGS,
|
"--no-enable-prefix-caching",
|
||||||
]
|
]
|
||||||
|
|
||||||
with RemoteOpenAIServer(MODEL_NAME, args) as remote_server:
|
with RemoteOpenAIServer(MODEL_NAME, args) as remote_server:
|
||||||
|
|||||||
Reference in New Issue
Block a user