.

Dockerfile

@@ -152,5 +152,16 @@ ADD https://github.com/krallin/tini/releases/download/${TINI_VERSION}/tini /tini
 RUN chmod +x /tini
 ENTRYPOINT ["/tini", "--"]
 
-# 仅做 CI / CPU 自检 —— 输出帮助后退出 0
-CMD ["python3", "-m", "sglang.launch_server", "--help"]
+# ---- 拷贝模型（路径可换） ----
+COPY ./Alibaba/Qwen3-8B /root/.cradle/Alibaba/Qwen3-8B
+
+# ---- 暴露端口 ----
+EXPOSE 30000
+
+# ---- 启动 SGLang 推理服务 ----
+CMD ["python3", "-m", "sglang.launch_server", \
+     "--host", "0.0.0.0", \
+     "--port", "30000", \
+     "--model-path", "/root/.cradle/Alibaba/Qwen3-8B/", \
+     "--tp", "1", \
+     "--api-key", "token-abc123"]