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 16, 18:15 UTC
[Bugfix] Fix run_batch logger (#7640)
Pooya Davoodi committed
Aug 23, 2024 at 13:58 UTC
6885fde317433eec52e00c14329270d742f0630d
1 file changed
+2
-5
vllm/entrypoints/openai/run_batch.py
+2
-5
@@ -6,7 +6,7 @@ import aiohttp
6
7
from vllm.engine.arg_utils import AsyncEngineArgs, nullable_str
8
from vllm.engine.async_llm_engine import AsyncLLMEngine
9
-from vllm.entrypoints.logger import RequestLogger
9
+from vllm.entrypoints.logger import RequestLogger, logger
10
# yapf: disable
11
from vllm.entrypoints.openai.protocol import (BatchRequestInput,
12
BatchRequestOutput,
@@ -16,13 +16,10 @@ from vllm.entrypoints.openai.protocol import (BatchRequestInput,
16
# yapf: enable
17
from vllm.entrypoints.openai.serving_chat import OpenAIServingChat
18
from vllm.entrypoints.openai.serving_embedding import OpenAIServingEmbedding
19
-from vllm.logger import init_logger
19
from vllm.usage.usage_lib import UsageContext
20
from vllm.utils import FlexibleArgumentParser, random_uuid
21
from vllm.version import __version__ as VLLM_VERSION
22
24
-logger = init_logger(__name__)
25
-
23
24
def parse_args():
25
parser = FlexibleArgumentParser(
@@ -184,7 +181,7 @@ async def main(args):
181
if __name__ == "__main__":
182
args = parse_args()
183
187
- logger.info("vLLM API server version %s", VLLM_VERSION)
184
+ logger.info("vLLM batch processing API version %s", VLLM_VERSION)
185
logger.info("args: %s", args)
186
187
asyncio.run(main(args))