curl访问流式非流式大模型openai api接口

参考：https://platform.openai.com/docs/api-reference/making-requests

命令行访问：
直接是vllm的openai api接口

curl http://192.168.***:10860/v1/chat/completions -H "Content-Type: application/json" -H "Authorization: EMPTY" -d '{"model": "/ai/Qwen1.5-7B-Chat","messages": [{"role": "user", "content"