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 14, 05:30 UTC
[Bugfix][V0 Deprecation][CI] use async mock and await for async method (#25325)
Signed-off-by: Yang <lymailforjob@gmail.com>
Yang Liu committed
Sep 21, 2025 at 16:06 UTC
04d3752329e39d11674862be341faaab42ac4d11
1 file changed
+20
-9
tests/entrypoints/openai/test_lora_resolvers.py
+20
-9
@@ -5,7 +5,7 @@ from contextlib import suppress
5
from dataclasses import dataclass, field
6
from http import HTTPStatus
7
from typing import Optional
8
-from unittest.mock import MagicMock
8
+from unittest.mock import AsyncMock, MagicMock
9
10
import pytest
11
@@ -83,20 +83,31 @@ def register_mock_resolver():
83
def mock_serving_setup():
84
"""Provides a mocked engine and serving completion instance."""
85
mock_engine = MagicMock(spec=AsyncLLM)
86
- mock_engine.get_tokenizer.return_value = get_tokenizer(MODEL_NAME)
86
mock_engine.errored = False
87
89
- def mock_add_lora_side_effect(lora_request: LoRARequest):
88
+ tokenizer = get_tokenizer(MODEL_NAME)
89
+ mock_engine.get_tokenizer = AsyncMock(return_value=tokenizer)
90
+
91
+ async def mock_add_lora_side_effect(lora_request: LoRARequest):
92
"""Simulate engine behavior when adding LoRAs."""
93
if lora_request.lora_name == "test-lora":
94
# Simulate successful addition
93
- return
94
- elif lora_request.lora_name == "invalid-lora":
95
+ return True
96
+ if lora_request.lora_name == "invalid-lora":
97
# Simulate failure during addition (e.g. invalid format)
98
raise ValueError(f"Simulated failure adding LoRA: "
99
f"{lora_request.lora_name}")
100
+ return True
101
+
102
+ mock_engine.add_lora = AsyncMock(side_effect=mock_add_lora_side_effect)
103
+
104
+ async def mock_generate(*args, **kwargs):
105
+ for _ in []:
106
+ yield _
107
+
108
+ mock_engine.generate = MagicMock(spec=AsyncLLM.generate,
109
+ side_effect=mock_generate)
110
99
- mock_engine.add_lora.side_effect = mock_add_lora_side_effect
111
mock_engine.generate.reset_mock()
112
mock_engine.add_lora.reset_mock()
113
@@ -131,7 +142,7 @@ async def test_serving_completion_with_lora_resolver(mock_serving_setup,
142
with suppress(Exception):
143
await serving_completion.create_completion(req_found)
144
134
- mock_engine.add_lora.assert_called_once()
145
+ mock_engine.add_lora.assert_awaited_once()
146
called_lora_request = mock_engine.add_lora.call_args[0][0]
147
assert isinstance(called_lora_request, LoRARequest)
148
assert called_lora_request.lora_name == lora_model_name
@@ -157,7 +168,7 @@ async def test_serving_completion_resolver_not_found(mock_serving_setup,
168
169
response = await serving_completion.create_completion(req)
170
160
- mock_engine.add_lora.assert_not_called()
171
+ mock_engine.add_lora.assert_not_awaited()
172
mock_engine.generate.assert_not_called()
173
174
assert isinstance(response, ErrorResponse)
@@ -181,7 +192,7 @@ async def test_serving_completion_resolver_add_lora_fails(
192
response = await serving_completion.create_completion(req)
193
194
# Assert add_lora was called before the failure
184
- mock_engine.add_lora.assert_called_once()
195
+ mock_engine.add_lora.assert_awaited_once()
196
called_lora_request = mock_engine.add_lora.call_args[0][0]
197
assert isinstance(called_lora_request, LoRARequest)
198
assert called_lora_request.lora_name == invalid_model