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

Convert formatting to use ruff instead of yapf + isort (#26247)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-05 15:06:22 +01:00
committed by GitHub
parent 17edd8a807
commit d6953beb91
1508 changed files with 115244 additions and 94146 deletions
Download Patch File Download Diff File Expand all files Collapse all files

66
tests/entrypoints/openai/test_lora_resolvers.py
View File

@@ -12,8 +12,7 @@ import pytest
from vllm.config.multimodal import MultiModalConfig
from vllm.entrypoints.openai.protocol import CompletionRequest, ErrorResponse
from vllm.entrypoints.openai.serving_completion import OpenAIServingCompletion
from vllm.entrypoints.openai.serving_models import (BaseModelPath,
OpenAIServingModels)
from vllm.entrypoints.openai.serving_models import BaseModelPath, OpenAIServingModels
from vllm.lora.request import LoRARequest
from vllm.lora.resolver import LoRAResolver, LoRAResolverRegistry
from vllm.transformers_utils.tokenizer import get_tokenizer
@@ -33,14 +32,14 @@ class MockHFConfig:
@dataclass
class MockModelConfig:
"""Minimal mock ModelConfig for testing."""
model: str = MODEL_NAME
tokenizer: str = MODEL_NAME
trust_remote_code: bool = False
tokenizer_mode: str = "auto"
max_model_len: int = 100
tokenizer_revision: Optional[str] = None
multimodal_config: MultiModalConfig = field(
default_factory=MultiModalConfig)
multimodal_config: MultiModalConfig = field(default_factory=MultiModalConfig)
hf_config: MockHFConfig = field(default_factory=MockHFConfig)
logits_processor_pattern: Optional[str] = None
diff_sampling_param: Optional[dict] = None
@@ -55,17 +54,21 @@ class MockModelConfig:
class MockLoRAResolver(LoRAResolver):
async def resolve_lora(self, base_model_name: str,
lora_name: str) -> Optional[LoRARequest]:
async def resolve_lora(
self, base_model_name: str, lora_name: str
) -> Optional[LoRARequest]:
if lora_name == "test-lora":
return LoRARequest(lora_name="test-lora",
lora_int_id=1,
lora_local_path="/fake/path/test-lora")
return LoRARequest(
lora_name="test-lora",
lora_int_id=1,
lora_local_path="/fake/path/test-lora",
)
elif lora_name == "invalid-lora":
return LoRARequest(lora_name="invalid-lora",
lora_int_id=2,
lora_local_path="/fake/path/invalid-lora")
return LoRARequest(
lora_name="invalid-lora",
lora_int_id=2,
lora_local_path="/fake/path/invalid-lora",
)
return None
@@ -96,8 +99,7 @@ def mock_serving_setup():
return True
if lora_request.lora_name == "invalid-lora":
# Simulate failure during addition (e.g. invalid format)
raise ValueError(f"Simulated failure adding LoRA: "
f"{lora_request.lora_name}")
raise ValueError(f"Simulated failure adding LoRA: {lora_request.lora_name}")
return True
mock_engine.add_lora = AsyncMock(side_effect=mock_add_lora_side_effect)
@@ -106,31 +108,31 @@ def mock_serving_setup():
for _ in []:
yield _
mock_engine.generate = MagicMock(spec=AsyncLLM.generate,
side_effect=mock_generate)
mock_engine.generate = MagicMock(spec=AsyncLLM.generate, side_effect=mock_generate)
mock_engine.generate.reset_mock()
mock_engine.add_lora.reset_mock()
mock_model_config = MockModelConfig()
models = OpenAIServingModels(engine_client=mock_engine,
base_model_paths=BASE_MODEL_PATHS,
model_config=mock_model_config)
models = OpenAIServingModels(
engine_client=mock_engine,
base_model_paths=BASE_MODEL_PATHS,
model_config=mock_model_config,
)
serving_completion = OpenAIServingCompletion(mock_engine,
mock_model_config,
models,
request_logger=None)
serving_completion = OpenAIServingCompletion(
mock_engine, mock_model_config, models, request_logger=None
)
serving_completion._process_inputs = AsyncMock(return_value=(MagicMock(
name="engine_request"), {}))
serving_completion._process_inputs = AsyncMock(
return_value=(MagicMock(name="engine_request"), {})
)
return mock_engine, serving_completion
@pytest.mark.asyncio
async def test_serving_completion_with_lora_resolver(mock_serving_setup,
monkeypatch):
async def test_serving_completion_with_lora_resolver(mock_serving_setup, monkeypatch):
monkeypatch.setenv("VLLM_ALLOW_RUNTIME_LORA_UPDATING", "true")
mock_engine, serving_completion = mock_serving_setup
@@ -152,14 +154,13 @@ async def test_serving_completion_with_lora_resolver(mock_serving_setup,
assert called_lora_request.lora_name == lora_model_name
mock_engine.generate.assert_called_once()
called_lora_request = mock_engine.generate.call_args[1]['lora_request']
called_lora_request = mock_engine.generate.call_args[1]["lora_request"]
assert isinstance(called_lora_request, LoRARequest)
assert called_lora_request.lora_name == lora_model_name
@pytest.mark.asyncio
async def test_serving_completion_resolver_not_found(mock_serving_setup,
monkeypatch):
async def test_serving_completion_resolver_not_found(mock_serving_setup, monkeypatch):
monkeypatch.setenv("VLLM_ALLOW_RUNTIME_LORA_UPDATING", "true")
mock_engine, serving_completion = mock_serving_setup
@@ -182,7 +183,8 @@ async def test_serving_completion_resolver_not_found(mock_serving_setup,
@pytest.mark.asyncio
async def test_serving_completion_resolver_add_lora_fails(
mock_serving_setup, monkeypatch):
mock_serving_setup, monkeypatch
):
monkeypatch.setenv("VLLM_ALLOW_RUNTIME_LORA_UPDATING", "true")
mock_engine, serving_completion = mock_serving_setup