missuo / freegpt35 Goto Github PK

高达1.11GB，强烈建议缩减一下体积。

之前成功安装运行过。删除了重装， freegpt35-freegpt35-1 容器启动不了，一个多小时了，端口应该没有被占用。CMD显示：

✔ 981253dc047b Download complete 27.0s
✔ chatgpt-next-web 9 layers [⣿⣿⣿⣿⣿⣿⣿⣿⣿] 0B/0B Pulled 42.4s
✔ e7ced292c644 Pull complete 23.7s
✔ b32c0114bba5 Pull complete 17.9s
✔ f3748d9674b0 Pull complete 17.1s
✔ e4e3baea97d1 Pull complete 18.8s
✔ dc47e87a8622 Pull complete 22.8s
✔ 7f3fde39cb0d Pull complete 21.4s
✔ a78e3f3f84b7 Pull complete 25.7s
✔ b9d45a982e50 Pull complete 26.6s
✔ 981253dc047b Pull complete 27.0s
[+] Running 0/3

• Network freegpt35_default Created 6567.7s
• Container freegpt35-freegpt35-1 Starting 6567.7s
• Container freegpt35-chatgpt-next-web-1 Created 6566.7s

win11下wsl2里怎么安装，调用呢？ 有谁能把命令更新一下吗？

How to add proxy options to docker？
I have v2raya docker client, how do I configure the FreeGPT35 container to use v2raya's proxy

docker run -d --name free_gpt --restart=always -p 3040:3040 ghcr.io/missuo/freegpt35

docker run -d --name gpt_next_web -p 3000:3000 yidadaa/chatgpt-next-web

宿主机上执行如下正确返回：

curl http://172.17.0.2:3040/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer any_string_you_like" \
  -d '{
    "model": "gpt-3.5-turbo",
    "messages": [
      {
        "role": "user",
        "content": "你好啊!"
      }
    ],
    "stream": true
    }'

但是部署到 chatgpt-next-web (使用域名反向代理到了chatgpt-next-web 容器)中就无法使用：

The request should return the error status code when it fails, not 200.

ChatGPT-Next-Web，稍微长点的答案（500字以内）回答不完整，再次给继续命令才能接下去，在哪里可以设置呢？

使用docker-composer部署，在请求时会出现

Before everything was normal, it would print this line of log： System: Successfully refreshed session ID and token.

{"status":false,"error":{"message":"The requested endpoint was not found. please make sure to use "http://localhost:3040/v1\" as the base URL.","type":"invalid_request_error"}}

刷新间隔的考虑是什么？以及是否可以延长？

When multiple messages are sent, {"status": false, "error": {"message": "An error happened, please make sure your request is SFW, or use a jailbreak to pass the filter.", "type": "invalid_request_error"} will appear

curl http://127.0.0.1:3040/v1/chat/completions -H "Content-Type: application/json" -H "Authorization: Bearer any_string_you_like" -d '{
"model": "gpt-3.5-turbo",
"messages": [
{
"role": "user",
"content": "Hello!"
}
],
"stream": true
}'

get a empty response with:
data: {"id":"chatcmpl-**********","created":1713965171742,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":""},"index":0,"finish_reason":"stop"}]}

使用测试用例

curl http://ip:3040/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer any_string_you_like" \
  -d '{
    "model": "gpt-3.5-turbo",
    "messages": [
      {
        "role": "user",
        "content": "Hello!"
      }
    ],
    "stream": true
    }'

发现返回的内容为空

data: {"id":"chatcmpl-9jKHVqNoeYznlG9q6iHJE7SkNQMM","created":1713580134677,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":""},"index":0,"finish_reason":"stop"}]}

在其他调用该api的应用中，同样收到回复，但是内容为空，所以是OpenAI新做了限制吗？

run node app.js

Error refreshing session ID, retrying in 1 minute...
If this error persists, your country may not be supported yet.
If your country was the issue, please consider using a U.S. VPN.

• Try changing servers and changing IP addresses.
• It's best not to deploy with Node, as it may be affected by the Node version. It is recommended to use Docker or Docker Compose.
• You can try using a WARP tunnel for outbound traffic.

The above methods may not necessarily work, but please give them a try.

US IP is not guaranteed to succeed!

does the api have an option for the system prompt?

在一些需要填写配置的地方，key随便填（我之前作废的key也ok），然后endpoint填写自己部署的地址即可。速度很快。

ps 以前都是只能买号，然后用不了几天就作废了。

body-parser 需要在 package.json 中声明一下

有些应用程序使用会报错这个，有知道怎么解决的吗？

Since

client = OpenAI(
        api_key="sk-fwref", base_url=base_url
    )

    a = client.models.list()

the Client will call base_url model list inside to connect, but it cannot return a response, how to make it work with python client?

这是为什么啊??搜索不到解决方案啊..

{
"status": false,
"error": {
"message": "Error getting a new session, please try again later, if the issue persists, please open an issue on the GitHub repository.",
"type": "invalid_request_error"
}
}

回复内容很短

{"status":false,"error":{"message":"The requested endpoint was not found. please make sure to use "http://localhost:3040/v1\" as the base URL.","type":"invalid_request_error"}}

Error refreshing session ID, retrying in 1 minute...
If this error persists, your country may not be supported yet.
If your country was the issue, please consider using a U.S. VPN.
Request: POST /v1/chat/completions 2 messages (stream-disabled)

如题，感谢作者！As title go, thanks!

https://github.com/aurorax-neo/free-gpt3.5-2api

目前可以使用这个项目！

希望能自定义api key，不然容易被扫到而滥用

我让GPT总结一篇文章，然后每个字都分一条data，这样取值比较麻烦，为什么是这样返回的？是我请求不对吗

data: {"id":"chatcmpl-bjKI588Q3ykpHOEE33fcsW05hSfZ","created":1712407975839,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":"阿"},"index":0,"finish_reason":null}]}

data: {"id":"chatcmpl-bjKI588Q3ykpHOEE33fcsW05hSfZ","created":1712407975839,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":"尔"},"index":0,"finish_reason":null}]}

data: {"id":"chatcmpl-bjKI588Q3ykpHOEE33fcsW05hSfZ","created":1712407975839,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":"玛"},"index":0,"finish_reason":null}]}

data: {"id":"chatcmpl-bjKI588Q3ykpHOEE33fcsW05hSfZ","created":1712407975839,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":"・"},"index":0,"finish_reason":null}]}

data: {"id":"chatcmpl-bjKI588Q3ykpHOEE33fcsW05hSfZ","created":1712407975839,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":"卡"},"index":0,"finish_reason":null}]}

data: {"id":"chatcmpl-bjKI588Q3ykpHOEE33fcsW05hSfZ","created":1712407975839,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":"瑞"},"index":0,"finish_reason":null}]}

data: {"id":"chatcmpl-bjKI588Q3ykpHOEE33fcsW05hSfZ","created":1712407975839,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":"，"},"index":0,"finish_reason":null}]}

Error getting a new session, please try again later, if the issue persists, please open an issue on the GitHub repository

最新版本的翻译速度特别的慢

curl http://127.0.0.1:3040/v1/chat/completions \

-H "Content-Type: application/json"
-H "Authorization: Bearer any_string_you_like"
-d '{
"model": "gpt-3.5-turbo",
"messages": [
{
"role": "user",
"content": "Hello!"
}
],
"stream": true
}'
{"status":false,"error":{"message":"An error happened, please make sure your request is SFW, or use a jailbreak to bypass the filte r.","type":"invalid_request_error"}}

{"status":false,"error":{"message":"An error happened, please make sure your request is SFW, or use a jailbreak to bypass the filter.","type":"invalid_request_error"}}

Thank you for your contribution, but I encountered some problems：

I just downloaded the source code, then ran node app.js, and made a curl request, and then got this error.

Thanks~

大佬你好，可以做一个fetch的版本吗？我用axios会触发cloudflare校验，换成fetch就没问题，原因未知

{
"status": false,
"error": {
"message": "The requested endpoint (GET /v1/chat/completions) was not found. please make sure what are you doing now.",
"type": "invalid_request_error"

Does it work? If yes then how do I use it?

The functionality to set HTTP or SOCKS proxies. I need. 3qs

Hello @missuo

I noticed that this repository contains code derived from my project, PawanOsman/ChatGPT, which is licensed under AGPL 3.0. It appears that this repository is licensed under MIT, which is not compatible with AGPL 3.0 terms. Could we update the license here to align with AGPL 3.0 to ensure compliance?

Thank you for your understanding and cooperation.

Best,
Pawan Osman

I used a US VPS to deploy the container, but I still got this error message: Error refreshing session ID, retrying in 1 minute...
If this error persists, your country may not be supported yet.

{"status":false,"error":{"message":"An error happened, please make sure your request is SFW, or use a jailbreak to bypass the filter.","type":"invalid_request_error"}}%

FreeAskInternet-API on 📦 v0.1.0 via 🐍 v3.11.6 (freeaskinternet-api-3.11) took 3m11s ❯ curl http://127.0.0.1:3040/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer any_string_you_like" \
  -d '{
    "model": "gpt-3.5-turbo",
    "messages": [
      {
        "role": "user",
        "content": "Hello!"
      }
    ],
    "stream": true
    }'
data: {"id":"chatcmpl-TnQ2dqZfFq0wFzHBehcHUxPv158X","created":1713909080880,"object":"chat.completion.chunk","model":"gpt-3.5-turbo","choices":[{"delta":{"content":""},"index":0,"finish_reason":"stop"}]}

I am getting finish_reason="stop", implying that the response is prematurely ending.

How to add proxy IP?

bad_response_status_code bad response status code 404 (request id: 2024040906334616828660341240500)

隔一段时间再用后，有提示。
{
"status": false,
"error": {
"message": "An error happened, please make sure your request is SFW, or use a jailbreak to bypass the filter.",
"type": "invalid_request_error"
}
}

目前已经实现了对话，但是无法进行会话保持，比如让GPT出3道面试题，下一个问题要答案的时候，却没办法回答上一条的答案，接口中的API是否有参数可以设置，感谢！！

How to call functions?