[CI] Add pytest markers to current tests and update the doc. (#577)

Signed-off-by: Alicia <115451386+congw729@users.noreply.github.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
This commit is contained in:
AliciaandHongsheng Liu authored and GitHub committed 2026-02-09 16:49:45 +08:00
1 parent cdde401813
commit d3ea94309d
37 files changed
+165 -94

No files matched your search

@@ -8,6 +8,7 @@ import pytest
from vllm import SamplingParams
from vllm.inputs import PromptType
from tests.utils import hardware_test
from vllm_omni.entrypoints.async_omni import AsyncOmni, ClientRequestState
os.environ["VLLM_WORKER_MULTIPROC_METHOD"] = "spawn"
@@ -57,6 +58,9 @@ async def generate(
return count, request_id
@pytest.mark.core_model
@pytest.mark.omni
@hardware_test(res={"cuda": "H100", "rocm": "MI325"}, num_cards=2)
@pytest.mark.asyncio
async def test_abort():
with ExitStack() as after:
@@ -113,6 +117,9 @@ async def test_abort():
await asyncio.sleep(5)
@pytest.mark.core_model
@pytest.mark.omni
@hardware_test(res={"cuda": "H100", "rocm": "MI325"}, num_cards=2)
@pytest.mark.asyncio
async def test_build_and_log_summary(monkeypatch):
from vllm_omni.entrypoints.utils import get_final_stage_id_for_e2e
@@ -22,6 +22,8 @@ from PIL import Image
from vllm.assets.image import ImageAsset
from vllm.utils.network_utils import get_open_port
from tests.utils import hardware_test
os.environ["VLLM_WORKER_MULTIPROC_METHOD"] = "spawn"
# Increase timeout for downloading assets from S3 (default 5s is too short for CI)
os.environ.setdefault("VLLM_IMAGE_FETCH_TIMEOUT", "60")
@@ -178,6 +180,9 @@ def _decode_data_url_to_image_bytes(data_url: str) -> bytes:
return base64.b64decode(b64_data)
@pytest.mark.core_model
@pytest.mark.diffusion
@hardware_test(res={"cuda": "H100", "rocm": "MI325"})
@pytest.mark.parametrize("omni_server", test_params, indirect=True)
def test_i2i_multi_image_input_qwen_image_edit_2509(
omni_server,
@@ -23,6 +23,7 @@ from PIL import Image
from safetensors.torch import save_file
from tests.conftest import OmniServer
from tests.utils import hardware_test
os.environ["VLLM_WORKER_MULTIPROC_METHOD"] = "spawn"
@@ -144,6 +145,9 @@ def _basic_payload() -> dict:
}
@pytest.mark.core_model
@pytest.mark.diffusion
@hardware_test(res={"cuda": "L4", "rocm": "MI325"})
def test_images_generations_per_request_lora_switching(omni_server: OmniServer, tmp_path: Path) -> None:
# Base generation.
base_img = _post_images(omni_server, _basic_payload())
@@ -28,6 +28,7 @@ from tests.conftest import (
merge_base64_and_convert_to_text,
modify_stage_config,
)
from tests.utils import hardware_test
from vllm_omni.platforms import current_omni_platform
models = ["Qwen/Qwen3-Omni-30B-A3B-Instruct"]
@@ -145,6 +146,9 @@ def get_max_batch_size(size_type="few"):
return batch_sizes.get(size_type, 5)
@pytest.mark.core_model
@pytest.mark.omni
@hardware_test(res={"cuda": "H100", "rocm": "MI325"}, num_cards=2)
@pytest.mark.parametrize("omni_server", test_params, indirect=True)
def test_mix_to_text_audio_001(client: openai.OpenAI, omni_server, request) -> None:
"""
@@ -215,6 +219,9 @@ def test_mix_to_text_audio_001(client: openai.OpenAI, omni_server, request) -> N
assert similarity > 0.9, "The audio content is not same as the text"
@pytest.mark.core_model
@pytest.mark.omni
@hardware_test(res={"cuda": "H100", "rocm": "MI325"}, num_cards=2)
@pytest.mark.parametrize("omni_server", test_params, indirect=True)
def test_text_to_text_audio_001(client: openai.OpenAI, omni_server) -> None:
"""