Repository navigation
|
Hi, i tries to use ollama as AI. I downloaded llama3.2 and gemma3 for ollame and used the ip:port from ollama for the Ollama API URL. Any ideas? |
Answered by
becksy82
Mar 11, 2026
Replies: 1 comment
|
I found the solution. I thought it was only necessary for online access. But it's necessary for all communication outside the container (e.g. for Open WebUI too). |
0 replies
Answer selected by
becksy82
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I found the solution.
I had configured Ollama incorrectly.
In the UnRAID template of Ollama, there is an environmental variable called ollama_host. The standard value is 127.0.0.1:11434.
For network access, it needs to be set to 0.0.0.0:11434 (or the port that you set up for the container).
I thought it was only necessary for online access. But it's necessary for all communication outside the container (e.g. for Open WebUI too).