|
|
@@ -1,4 +1,3 @@
|
|
|
-version: "3.8"
|
|
|
services:
|
|
|
# vLLM - Qwen3 0.6B
|
|
|
vllm-0.6b:
|
|
|
@@ -6,14 +5,20 @@ services:
|
|
|
container_name: vllm-qwen3-0.6b
|
|
|
ports:
|
|
|
- "8100:8000"
|
|
|
- environment:
|
|
|
- - HF_ENDPOINT=https://hf-mirror.com
|
|
|
command: >
|
|
|
--model Qwen/Qwen3-Embedding-0.6B
|
|
|
--dtype float16
|
|
|
--port 8000
|
|
|
volumes:
|
|
|
- - ./models:/root/.cache/huggingface
|
|
|
+ # 挂载预下载的模型目录
|
|
|
+ - ./models/Qwen3-Embedding-0.6B:/app/models/Qwen3-Embedding-0.6B
|
|
|
+ # 可选：挂载缓存目录
|
|
|
+ - ./cache:/root/.cache/huggingface
|
|
|
+ environment:
|
|
|
+ # 设置 Hugging Face 国内镜像
|
|
|
+ - HF_ENDPOINT=https://hf-mirror.com
|
|
|
+ # 可选：设置 Python 包索引镜像
|
|
|
+ - PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple
|
|
|
deploy:
|
|
|
resources:
|
|
|
reservations:
|
|
|
@@ -21,7 +26,13 @@ services:
|
|
|
- driver: nvidia
|
|
|
count: all
|
|
|
capabilities: [ gpu ]
|
|
|
-
|
|
|
+ # 添加健康检查
|
|
|
+ healthcheck:
|
|
|
+ test: [ "CMD", "curl", "-f", "http://localhost:8000/health" ]
|
|
|
+ interval: 30s
|
|
|
+ timeout: 10s
|
|
|
+ retries: 3
|
|
|
+ start_period: 40s
|
|
|
# # vLLM - Qwen3 4B
|
|
|
# vllm-4b:
|
|
|
# image: vllm/vllm-openai:latest
|
luojunhui