Skip to content

Commit 03dccc8

Browse files
[Misc] Add vLLM version getter to utils (#5098)
1 parent a65634d commit 03dccc8

File tree

7 files changed

+13
-11
lines changed

7 files changed

+13
-11
lines changed

setup.py

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -314,7 +314,7 @@ def find_version(filepath: str) -> str:
314314

315315

316316
def get_vllm_version() -> str:
317-
version = find_version(get_path("vllm", "__init__.py"))
317+
version = find_version(get_path("vllm", "version.py"))
318318

319319
if _is_cuda():
320320
cuda_version = str(get_nvcc_cuda_version())

vllm/__init__.py

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -12,9 +12,10 @@
1212
from vllm.pooling_params import PoolingParams
1313
from vllm.sampling_params import SamplingParams
1414

15-
__version__ = "0.5.0"
15+
from .version import __version__
1616

1717
__all__ = [
18+
"__version__",
1819
"LLM",
1920
"ModelRegistry",
2021
"PromptStrictInputs",

vllm/engine/llm_engine.py

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -6,7 +6,6 @@
66

77
from transformers import GenerationConfig, PreTrainedTokenizer
88

9-
import vllm
109
from vllm.config import (CacheConfig, DecodingConfig, DeviceConfig, LoadConfig,
1110
LoRAConfig, ModelConfig, ParallelConfig,
1211
SchedulerConfig, SpeculativeConfig,
@@ -38,6 +37,7 @@
3837
from vllm.usage.usage_lib import (UsageContext, is_usage_stats_enabled,
3938
usage_message)
4039
from vllm.utils import Counter
40+
from vllm.version import __version__ as VLLM_VERSION
4141

4242
logger = init_logger(__name__)
4343
_LOCAL_LOGGING_INTERVAL_SEC = 5
@@ -169,7 +169,7 @@ def __init__(
169169
"enforce_eager=%s, kv_cache_dtype=%s, "
170170
"quantization_param_path=%s, device_config=%s, "
171171
"decoding_config=%r, seed=%d, served_model_name=%s)",
172-
vllm.__version__,
172+
VLLM_VERSION,
173173
model_config.model,
174174
speculative_config,
175175
model_config.tokenizer,

vllm/entrypoints/openai/api_server.py

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -15,7 +15,6 @@
1515
from prometheus_client import make_asgi_app
1616
from starlette.routing import Mount
1717

18-
import vllm
1918
import vllm.envs as envs
2019
from vllm.engine.arg_utils import AsyncEngineArgs
2120
from vllm.engine.async_llm_engine import AsyncLLMEngine
@@ -29,6 +28,7 @@
2928
from vllm.entrypoints.openai.serving_embedding import OpenAIServingEmbedding
3029
from vllm.logger import init_logger
3130
from vllm.usage.usage_lib import UsageContext
31+
from vllm.version import __version__ as VLLM_VERSION
3232

3333
TIMEOUT_KEEP_ALIVE = 5 # seconds
3434

@@ -93,7 +93,7 @@ async def show_available_models():
9393

9494
@app.get("/version")
9595
async def show_version():
96-
ver = {"version": vllm.__version__}
96+
ver = {"version": VLLM_VERSION}
9797
return JSONResponse(content=ver)
9898

9999

@@ -174,7 +174,7 @@ async def authentication(request: Request, call_next):
174174
raise ValueError(f"Invalid middleware {middleware}. "
175175
f"Must be a function or a class.")
176176

177-
logger.info("vLLM API server version %s", vllm.__version__)
177+
logger.info("vLLM API server version %s", VLLM_VERSION)
178178
logger.info("args: %s", args)
179179

180180
if args.served_model_name is not None:

vllm/entrypoints/openai/run_batch.py

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -5,7 +5,6 @@
55

66
import aiohttp
77

8-
import vllm
98
from vllm.engine.arg_utils import AsyncEngineArgs, nullable_str
109
from vllm.engine.async_llm_engine import AsyncLLMEngine
1110
from vllm.entrypoints.openai.protocol import (BatchRequestInput,
@@ -15,6 +14,7 @@
1514
from vllm.logger import init_logger
1615
from vllm.usage.usage_lib import UsageContext
1716
from vllm.utils import random_uuid
17+
from vllm.version import __version__ as VLLM_VERSION
1818

1919
logger = init_logger(__name__)
2020

@@ -135,7 +135,7 @@ async def main(args):
135135
if __name__ == "__main__":
136136
args = parse_args()
137137

138-
logger.info("vLLM API server version %s", vllm.__version__)
138+
logger.info("vLLM API server version %s", VLLM_VERSION)
139139
logger.info("args: %s", args)
140140

141141
asyncio.run(main(args))

vllm/usage/usage_lib.py

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -16,6 +16,7 @@
1616
import torch
1717

1818
import vllm.envs as envs
19+
from vllm.version import __version__ as VLLM_VERSION
1920

2021
_config_home = envs.VLLM_CONFIG_ROOT
2122
_USAGE_STATS_JSON_PATH = os.path.join(_config_home, "vllm/usage_stats.json")
@@ -163,9 +164,8 @@ def _report_usage_once(self, model_architecture: str,
163164
])
164165

165166
# vLLM information
166-
import vllm # delayed import to prevent circular import
167167
self.context = usage_context.value
168-
self.vllm_version = vllm.__version__
168+
self.vllm_version = VLLM_VERSION
169169
self.model_architecture = model_architecture
170170

171171
# Metadata

vllm/version.py

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
__version__ = "0.5.0"

0 commit comments

Comments
 (0)