[Frontend][last/N] Move all non-OpenAI content out of the OpenAI folder. (#56369)

Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
This commit is contained in:
wang.yuqi
2026-09-11 05:33:26 +00:00
committed by GitHub
parent 5ff50f3996
commit fe4d4109c2
5 changed files with 3 additions and 85 deletions
@@ -1,82 +0,0 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
import asyncio
import random
from collections.abc import Callable
import openai
import pytest
import pytest_asyncio
from tests.utils import RemoteOpenAIServer
MODEL_NAME = "Qwen/Qwen2.5-1.5B-Instruct"
@pytest.fixture(scope="module")
def server():
args = [
# use half precision for speed and memory savings in CI environment
"--dtype",
"bfloat16",
"--max-model-len",
"8192",
"--enforce-eager",
"--max-num-seqs",
"128",
"--load-format",
"dummy",
]
with RemoteOpenAIServer(MODEL_NAME, args) as remote_server:
yield remote_server
@pytest_asyncio.fixture
async def client(server):
async with server.get_async_client() as async_client:
yield async_client
@pytest.mark.asyncio
@pytest.mark.parametrize(
ids=["completion", "chat"],
argnames=["create_func_gen", "content_body"],
argvalues=[
(lambda x: x.completions.create, {"prompt": " ".join(["A"] * 10_000)}),
(
lambda x: x.chat.completions.create,
{"messages": [{"role": "user", "content": " ".join(["A"] * 10_000)}]},
),
],
)
async def test_with_and_without_truncate(
server: RemoteOpenAIServer,
client: openai.AsyncOpenAI,
create_func_gen: Callable,
content_body: dict,
):
create_func = create_func_gen(client)
body = {"model": MODEL_NAME, **content_body, "max_tokens": 10}
num_requests = 10
truncate_prompt_tokens = [1000] * (num_requests // 2) + [None] * (
num_requests - num_requests // 2
)
random.shuffle(truncate_prompt_tokens)
bodies = [
{**body, "extra_body": {"truncate_prompt_tokens": t}}
for t in truncate_prompt_tokens
]
async def get_status_code(**kwargs):
try:
await create_func(**kwargs)
return 200
except openai.APIStatusError as e:
return e.status_code
responses = await asyncio.gather(*[get_status_code(**b) for b in bodies])
assert 500 not in responses
@@ -6,7 +6,7 @@ import asyncio
import pytest
from starlette.responses import StreamingResponse
from vllm.entrypoints.openai.sse_keep_alive import (
from vllm.entrypoints.serve.utils.sse_keep_alive import (
SSE_KEEP_ALIVE_COMMENT,
with_sse_keep_alive,
)
@@ -14,7 +14,6 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
ChatCompletionResponse,
)
from vllm.entrypoints.openai.chat_completion.serving import OpenAIServingChat
from vllm.entrypoints.openai.sse_keep_alive import with_sse_keep_alive
from vllm.entrypoints.serve.engine.protocol import ErrorResponse
from vllm.entrypoints.serve.utils.api_utils import (
load_aware_call,
@@ -22,6 +21,7 @@ from vllm.entrypoints.serve.utils.api_utils import (
with_cancellation,
)
from vllm.entrypoints.serve.utils.orca_metrics import metrics_header
from vllm.entrypoints.serve.utils.sse_keep_alive import with_sse_keep_alive
from vllm.logger import init_logger
logger = init_logger(__name__)
@@ -12,7 +12,6 @@ from vllm.entrypoints.openai.completion.protocol import (
CompletionResponse,
)
from vllm.entrypoints.openai.completion.serving import OpenAIServingCompletion
from vllm.entrypoints.openai.sse_keep_alive import with_sse_keep_alive
from vllm.entrypoints.serve.engine.protocol import ErrorResponse
from vllm.entrypoints.serve.utils.api_utils import (
load_aware_call,
@@ -20,6 +19,7 @@ from vllm.entrypoints.serve.utils.api_utils import (
with_cancellation,
)
from vllm.entrypoints.serve.utils.orca_metrics import metrics_header
from vllm.entrypoints.serve.utils.sse_keep_alive import with_sse_keep_alive
from vllm.logger import init_logger
logger = init_logger(__name__)