Open xinoxme opened 1 year ago
I have the same message error
I have almost the identical error.
{"_action":"_ask","error":"an error occurred HTTPSConnectionPool(host='api.openai.com', port=443): Max retries exceeded with url: /v1/chat/completions (Caused by ProxyError('Cannot connect to proxy.', NewConnectionError('<urllib3.connection.HTTPSConnection object at 0x7f943d6ae970>: Failed to establish a new connection: [Errno 111] Connection refused')))","success":false}
The default config.json contains this section:
"openai_api_base": "https://api.openai.com", "proxy": { "enable": true, "http": "127.0.0.1:7890", "https": "127.0.0.1:7890" }
It looks like it's expecting a proxy to be running locally on port 7890. I tried setting the enable option to false
but I still get the error above.
Getting the same error, have not seen any feed back on this error so far.
Same here.
{"_action":"_ask","error":"an error occurred HTTPSConnectionPool(host='api.openai.com', port=443): Max retries exceeded with url: /v1/chat/completions (Caused by ProxyError('Cannot connect to proxy.', NewConnectionError('<urllib3.connection.HTTPSConnection object at 0x0000020E2C832C50>: Failed to establish a new connection: [WinError 10061] No connection could be made because the target machine actively refused it')))","success":false}
edit config.json and enter api key ! set False for proxy
api key was already in there. I set proxy to false, but that gets me: "I apologize, but it seems that I am having difficulty connecting to the OpenAI server. Please check your internet connection and try again later."
..and in the console: " list index out of range None
Following up... I cleared all conversations in the web gui and it's working properly now. Thanks for the help. The fix: set the proxy false in config.json "proxy": { "enable": false,
When I ask something, it never answers me. Any help?
edit config.json and enter api key ! set False for proxy
it did not work
Yeah same, I think I was under the impression there was some kind of reversed engendered api that I. Wouldn't need to use my own. This you need an api key.
Check out the author's other project - gpt4free - that's the reverse engineered API. This isn't that.
edit config.json and enter api key ! set False for proxy
it did not work
I had to restart the script, restart the browser, and clear all conversations to get it working after the changes.
edit config.json and enter api key ! set False for proxy
it did not work
I had to restart the script, restart the browser, and clear all conversations to get it working after the changes.
did it work?
Because of the way docker works, the config.json is burned INTO the image when you first build it, you can create a hard link between your local config.json and the one inside the docker image, or you can re-build the image with the update config.json with the steps below.
1) docker-compose down
2) make changes to config.json (nano config.json)
3) docker-compose up -d --build
-d starts it in daemon mode (background) and --build forces to rebuild the image with any updated files.
edit config.json and enter api key ! set False for proxy
it did not work
I had to restart the script, restart the browser, and clear all conversations to get it working after the changes.
did it work?
Yes. It worked fine... two days ago. When I tried it today (with no changes) I don't get any response at all. No errors, it just sticks on generating forever.
See the same problem. config.json looks fine, API key is good. cleared conversations, restarted script. no change.
set proxy: false and it works a little better now except the console shows a lot of "list index out of range" errors now. gpt-35-turbo works, but gpt-4 fails
after i changed true to false all modes stopped working
after i changed true to false all modes stopped working
Likewise in my case, I just get no response no matter if I use gpt-3.5 or gpt-4
Luisf371's comment on May 4 corrected the problem for me, after changing the proxy term to false in config.json. Thank you, I would have never figured that out!
If you have set the proxy
to false and added the API key
and it still doesn't work, then it could be an issue related to your OpenAI account since they only let you use the API for free for the first three months
after account creation. Unfortunately, the only solution to that issue is adding a payment method here. Keep in mind that you can set usage limits here.
Thank you for your open-source project. I have identified a bug When the input content is too much, it will exceed the range of the input box