A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 6, 18:15 UTC
[Refactor] Relocate chat completion and anthropic tests (#36919)
Signed-off-by: sfeng33 <4florafeng@gmail.com>
Flora Feng committed
Mar 14, 2026 at 00:16 UTC
bcfdadb1bc4db9dc8fe82710d4301a1d11114a3c
14 files changed
+14
-18
.buildkite/test_areas/entrypoints.yaml
+1
-1
@@ -34,7 +34,7 @@ steps:
34
- tests/entrypoints/test_chat_utils
35
commands:
36
- export VLLM_WORKER_MULTIPROC_METHOD=spawn
37
- - pytest -v -s entrypoints/openai --ignore=entrypoints/openai/test_chat_with_tool_reasoning.py --ignore=entrypoints/openai/test_oot_registration.py --ignore=entrypoints/openai/test_tensorizer_entrypoint.py --ignore=entrypoints/openai/correctness/ --ignore=entrypoints/openai/tool_parsers/ --ignore=entrypoints/openai/responses
37
+ - pytest -v -s entrypoints/openai --ignore=entrypoints/openai/chat_completion/test_chat_with_tool_reasoning.py --ignore=entrypoints/openai/test_oot_registration.py --ignore=entrypoints/openai/test_tensorizer_entrypoint.py --ignore=entrypoints/openai/correctness/ --ignore=entrypoints/openai/tool_parsers/ --ignore=entrypoints/openai/responses
38
- pytest -v -s entrypoints/test_chat_utils.py
39
mirror:
40
amd:
.github/mergify.yml
+1
-1
@@ -334,7 +334,7 @@ pull_request_rules:
334
- or:
335
- files~=^tests/tool_use/
336
- files~=^tests/entrypoints/openai/tool_parsers/
337
- - files=tests/entrypoints/openai/test_chat_with_tool_reasoning.py
337
+ - files=tests/entrypoints/openai/chat_completion/test_chat_with_tool_reasoning.py
338
- files~=^vllm/entrypoints/openai/tool_parsers/
339
- files=docs/features/tool_calling.md
340
- files~=^examples/tool_chat_*
tests/entrypoints/anthropic/__init__.py
tests/entrypoints/anthropic/test_anthropic_messages_conversion.py
renamed
tests/entrypoints/openai/chat_completion/__init__.py
tests/entrypoints/openai/chat_completion/test_chat.py
renamed
+1
-2
@@ -14,13 +14,12 @@ import requests
14
import torch
15
from openai import BadRequestError
16
17
+from tests.utils import RemoteOpenAIServer
18
from vllm.entrypoints.openai.chat_completion.protocol import (
19
ChatCompletionRequest,
20
)
21
from vllm.sampling_params import SamplingParams
22
22
-from ...utils import RemoteOpenAIServer
23
-
23
# any model with a chat template should work here
24
MODEL_NAME = "HuggingFaceH4/zephyr-7b-beta"
25
tests/entrypoints/openai/chat_completion/test_chat_echo.py
renamed
+1
-2
@@ -7,10 +7,9 @@ import openai # use the official client for correctness check
7
import pytest
8
import pytest_asyncio
9
10
+from tests.utils import RemoteOpenAIServer
11
from vllm.config import ModelConfig
12
12
-from ...utils import RemoteOpenAIServer
13
-
13
# # any model with a chat template should work here
14
MODEL_NAME = "Qwen/Qwen2-1.5B-Instruct"
15
tests/entrypoints/openai/chat_completion/test_chat_error.py
renamed
tests/entrypoints/openai/chat_completion/test_chat_logit_bias_validation.py
renamed
+1
-2
@@ -5,10 +5,9 @@ import openai
5
import pytest
6
import pytest_asyncio
7
8
+from tests.utils import RemoteOpenAIServer
9
from vllm.config import ModelConfig
10
10
-from ...utils import RemoteOpenAIServer
11
-
11
MODEL_NAME = "Qwen/Qwen2.5-1.5B-Instruct"
12
13
tests/entrypoints/openai/chat_completion/test_chat_with_tool_reasoning.py
renamed
+1
-1
@@ -5,7 +5,7 @@ import openai # use the official client for correctness check
5
import pytest
6
import pytest_asyncio
7
8
-from ...utils import RemoteOpenAIServer
8
+from tests.utils import RemoteOpenAIServer
9
10
# a reasoning and tool calling model
11
MODEL_NAME = "Qwen/QwQ-32B"
tests/entrypoints/openai/chat_completion/test_completion_with_function_calling.py
renamed
+1
-1
@@ -10,7 +10,7 @@ import pytest
10
import pytest_asyncio
11
12
# downloading lora to test lora requests
13
-from ...utils import ROCM_ENV_OVERRIDES, ROCM_EXTRA_ARGS, RemoteOpenAIServer
13
+from tests.utils import ROCM_ENV_OVERRIDES, ROCM_EXTRA_ARGS, RemoteOpenAIServer
14
15
# any model with a chat template should work here
16
MODEL_NAME = "Qwen/Qwen3-0.6B"
tests/entrypoints/openai/chat_completion/test_enable_force_include_usage.py
renamed
+1
-1
@@ -4,7 +4,7 @@ import openai
4
import pytest
5
import pytest_asyncio
6
7
-from ...utils import RemoteOpenAIServer
7
+from tests.utils import RemoteOpenAIServer
8
9
10
@pytest.fixture(scope="module")
tests/entrypoints/openai/chat_completion/test_serving_chat.py
renamed
+6
-7
@@ -10,6 +10,12 @@ import pytest
10
import pytest_asyncio
11
from openai import OpenAI
12
13
+from tests.entrypoints.openai.utils import (
14
+ accumulate_streaming_response,
15
+ verify_chat_response,
16
+ verify_harmony_messages,
17
+)
18
+from tests.utils import RemoteOpenAIServer
19
from vllm._aiter_ops import is_aiter_found_and_supported
20
from vllm.config import MultiModalConfig
21
from vllm.entrypoints.openai.chat_completion.protocol import (
@@ -39,13 +45,6 @@ from vllm.tokenizers.registry import tokenizer_args_from_config
45
from vllm.tool_parsers import ToolParserManager
46
from vllm.v1.engine.async_llm import AsyncLLM
47
42
-from ...utils import RemoteOpenAIServer
43
-from .utils import (
44
- accumulate_streaming_response,
45
- verify_chat_response,
46
- verify_harmony_messages,
47
-)
48
-
48
GPT_OSS_MODEL_NAME = "openai/gpt-oss-20b"
49
GPT_OSS_SPECULATOR_NAME = "RedHatAI/gpt-oss-20b-speculator.eagle3"
50