Replies: 4 comments 2 replies
-
If | curl doesnt have the same error then this is probably something else (not a firewall / proxy issue). Did you specify a correct Organization or did you use org-XXXX as pictured? |
Beta Was this translation helpful? Give feedback.
-
When I test it with curl from my search head on the OS level, I am getting the correct result back: splunk@ccd01v013355:~> curl https://api.openai.com/v1/completions -H 'Content-Type: application/json' -H "Authorization: Bearer sk-U02F66OhGKDEAJChRJMpT3BlbkFJxo.........." -d '{ }' |
Beta Was this translation helpful? Give feedback.
-
The command has to be run within splunk search bar because it streams results from the pipeline through it. The error you're getting on CLI is telling you it can't find the splunk libraries because of this. you could try running it with splunk instead ./splunk cmd python /path/to/chatgpt.py I don't know if it will work that way at all but it's worth a shot. |
Beta Was this translation helpful? Give feedback.
-
More I look at this the more it does feel like a proxy issue. Do you have a proxy? Can you ask your network team why you're getting connection refused? |
Beta Was this translation helpful? Give feedback.
-
Hello,
I am getting the following error:
I guess this must have sth to do with my port opening. How can I track it further?
What I tried though was the splunk curl command, which we use a lot and really like it .-), but also there I am unable to pass the payload in the correct json form:
I tried also eval data = json_object(...) and some other combinations, but still getting same error.
The connect to check models (just header, no data) works fine, so I think the issue must be in the format of the data field.
How would I continue from here?
Best with the ta-openai-api, but if it takes me too many loops with my internal security team for port opening, I would accept also the way over the webtools-add on
Kind Regards,
Kamil
Beta Was this translation helpful? Give feedback.
All reactions