biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Frontend][4/n] Improve pooling entrypoints | pooling. (#39153)

Browse Source 
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
This commit is contained in:
wang.yuqi
2026-04-09 18:09:45 +08:00
committed by GitHub
parent b6c9be509e
commit 66c079ae83
43 changed files with 554 additions and 733 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
tests/entrypoints/openai/chat_completion/test_serving_chat.py
View File

@@ -567,7 +567,6 @@ def _build_serving_render(
return OpenAIServingRender(
model_config=engine.model_config,
renderer=engine.renderer,
io_processor=engine.io_processor,
model_registry=model_registry,
request_logger=None,
chat_template=CHAT_TEMPLATE,
@@ -599,7 +598,6 @@ def _build_serving_chat(engine: AsyncLLM) -> OpenAIServingChat:
class MockEngine:
model_config: MockModelConfig = field(default_factory=MockModelConfig)
input_processor: MagicMock = field(default_factory=MagicMock)
io_processor: MagicMock = field(default_factory=MagicMock)
renderer: MagicMock = field(default_factory=MagicMock)
@@ -632,7 +630,6 @@ async def test_serving_chat_returns_correct_model_name():
mock_engine.errored = False
mock_engine.model_config = MockModelConfig()
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
serving_chat = _build_serving_chat(mock_engine)
@@ -662,7 +659,6 @@ async def test_serving_chat_should_set_correct_max_tokens():
mock_engine.errored = False
mock_engine.model_config = MockModelConfig()
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
serving_chat = _build_serving_chat(mock_engine)
@@ -693,7 +689,6 @@ async def test_serving_chat_should_set_correct_max_tokens():
mock_engine.errored = False
mock_engine.model_config = mock_model_config
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
# Initialize the serving chat
@@ -737,7 +732,6 @@ async def test_serving_chat_should_set_correct_max_tokens():
mock_engine.errored = False
mock_engine.model_config = mock_model_config
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
serving_chat = _build_serving_chat(mock_engine)
@@ -779,7 +773,6 @@ async def test_serving_chat_should_set_correct_max_tokens():
mock_engine.errored = False
mock_engine.model_config = mock_model_config
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
# Initialize the serving chat
@@ -823,7 +816,6 @@ async def test_serving_chat_mistral_token_ids_prompt_is_validated():
mock_engine.errored = False
mock_engine.model_config = MockModelConfig(skip_tokenizer_init=True)
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_tokenizer = MagicMock(spec=MistralTokenizer)
mock_renderer = MistralRenderer(
@@ -863,7 +855,6 @@ async def test_serving_chat_mistral_token_ids_prompt_too_long_is_rejected():
mock_engine.errored = False
mock_engine.model_config = MockModelConfig(skip_tokenizer_init=True)
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_tokenizer = MagicMock(spec=MistralTokenizer)
mock_renderer = MistralRenderer(
@@ -906,7 +897,6 @@ async def test_serving_chat_could_load_correct_generation_config():
mock_engine.errored = False
mock_engine.model_config = mock_model_config
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
# Initialize the serving chat
@@ -952,7 +942,6 @@ async def test_serving_chat_did_set_correct_cache_salt(model_type):
mock_engine.errored = False
mock_engine.model_config = mock_model_config
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
serving_chat = _build_serving_chat(mock_engine)
@@ -1003,7 +992,6 @@ async def test_serving_chat_data_parallel_rank_extraction():
mock_engine.errored = False
mock_engine.model_config = MockModelConfig()
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
# Mock the generate method to return an async generator
@@ -1095,7 +1083,6 @@ class TestServingChatWithHarmony:
mock_engine.errored = False
mock_engine.model_config = MockModelConfig()
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
return mock_engine
@@ -1732,7 +1719,6 @@ async def test_tool_choice_validation_without_parser():
mock_engine.errored = False
mock_engine.model_config = MockModelConfig()
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
models = OpenAIServingModels(
@@ -1802,7 +1788,6 @@ async def test_streaming_n_gt1_independent_tool_parsers():
mock_engine.errored = False
mock_engine.model_config = MockModelConfig()
mock_engine.input_processor = MagicMock()
mock_engine.io_processor = MagicMock()
mock_engine.renderer = _build_renderer(mock_engine.model_config)
models = OpenAIServingModels(