[CI] Consolidate multimodal entrypoint tests. (#44819)

Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
This commit is contained in:
wang.yuqi
2026-06-08 04:48:08 -07:00
committed by GitHub
parent 980796cd07
commit 93ee4cd47f
24 changed files with 107 additions and 62 deletions
-42
View File
@@ -4,7 +4,6 @@ import weakref
import pytest
from tests.entrypoints.openai.chat_completion.test_vision import TEST_IMAGE_ASSETS
from vllm import LLM
from vllm.distributed import cleanup_dist_env_and_memory
from vllm.sampling_params import SamplingParams
@@ -76,47 +75,6 @@ def test_multi_chat(text_llm):
assert len(outputs) == 2
@pytest.fixture(scope="function")
def vision_llm():
# pytest caches the fixture so we use weakref.proxy to
# enable garbage collection
llm = LLM(
model="microsoft/Phi-3.5-vision-instruct",
max_model_len=4096,
max_num_seqs=5,
enforce_eager=True,
trust_remote_code=True,
limit_mm_per_prompt={"image": 2},
seed=0,
)
yield weakref.proxy(llm)
del llm
cleanup_dist_env_and_memory()
@pytest.mark.parametrize(
"image_urls", [[TEST_IMAGE_ASSETS[0], TEST_IMAGE_ASSETS[1]]], indirect=True
)
def test_chat_multi_image(vision_llm, image_urls: list[str]):
messages = [
{
"role": "user",
"content": [
*(
{"type": "image_url", "image_url": {"url": image_url}}
for image_url in image_urls
),
{"type": "text", "text": "What's in this image?"},
],
}
]
outputs = vision_llm.chat(messages)
assert len(outputs) >= 0
def test_llm_chat_tokenization_no_double_bos(text_llm):
"""
LLM.chat() should not add special tokens when using chat templates.