为什么请求的模型是 gpt-4 但接口返回的回答里显示 gpt-3.5-turbo-0301

[dlxj]

在 chatgpt 网页上是可以正常使用 gpt4 的

curl 127.0.0.1:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer NONEED" \
  -d '{
     "model": "gpt-4",
     "messages": [{"role": "user", "content": "你会说中文吗"}],
     "temperature": 0.7
   }'

{"id":"chatcmpl-QXlha2FBbmROaXhpZUFyZUF3ZXNvbWUK","object":"chat.completion","created":0,"model":"gpt-3.5-turbo-0301","usage":{"prompt_tokens":0,"completion_tokens":0,"total_tokens":0},"choices":[{"index":0,"message":{"role":"assistant","content":"会的，我可以用中文回答问题。请随时告诉我您的问题或需求。"},"finish_reason":null}]}

[xqdoo00o]

额，这个是代码里写死的，model名字无所谓的吧