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 17, 03:00 UTC
usage lib get version another way (#3735)
Simon Mo committed
Mar 29, 2024 at 15:57 UTC
8b2d3cbc1bf4791329fdf400fb1b087b8b1a42f8
1 file changed
+2
-2
vllm/usage/usage_lib.py
+2
-2
@@ -11,7 +11,6 @@ from typing import Dict, Optional
11
from uuid import uuid4
12
13
import cpuinfo
14
-import pkg_resources
14
import psutil
15
import requests
16
import torch
@@ -161,8 +160,9 @@ class UsageMessage:
160
])
161
162
# vLLM information
163
+ import vllm # delayed import to prevent circular import
164
self.context = usage_context.value
165
- self.vllm_version = pkg_resources.get_distribution("vllm").version
165
+ self.vllm_version = vllm.__version__
166
self.model_architecture = model_architecture
167
168
# Metadata