-
-
Notifications
You must be signed in to change notification settings - Fork 3.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
AIOHTTP_CLIENT_TIMEOUT Does not work #3258
Comments
It would also be nice if it was higher than 5 minutes by default (e.g. 20 minutes), please consider that some people do not have high end hardware |
Which docker image are you using? And did you redeploy the compose file after changing it? (docker compose up) |
Using
|
Hmm.. Using
I did put it at 15 minutes in my PR, but it was changed back to the default (5 min) before being accepted. Edit: Could you post your docker-compose file? |
|
Hmm.. Is the keep-alive still on default setting? Can you try changing it if it's default? In Open Webui, under Settings -> General -> Advanced Settings -> Keep Alive and set it to something like 30m? I don't think this should kick in when it's generating, but it's worth checking out. Also, do you know if ollama spent more than 5 minutes loading the model? Did it start responding before it stopped? |
It starts generating before it stops
Does not fix it |
since you have a build section in the docker compose file, is it built from the folder or do you use image from the web? And are you using ollama embedded in the image or separate external install? |
Ollama is running outside of docker I mostly copied from the example docker files: |
Could you try this docker file?
and if it still doesn't work, can you give the output of
|
Hmm, it seems to be working now |
Glad to hear! I wonder what the issue was. I suspect it was either using an old image somehow or the image has old code in it. Both of those seems weird, though. Main should have the newest code, and pull should have pulled the newest image.. Edit: And since it's working, maybe you can close the bug report :) |
Bug Report
Setting
AIOHTTP_CLIENT_TIMEOUT
has no effect, it is impossible for longer responses with a larger model to complete using this software on my GPU.Description
I am using host ollama with a docker compose file, where I set
AIOHTTP_CLIENT_TIMEOUT
like this:Bug Summary:
Setting
AIOHTTP_CLIENT_TIMEOUT
has no effect. Setting it to 1 does not fail after 1 second. Setting it to 99999999 still fails in 5 minutes.Steps to Reproduce:
AIOHTTP_CLIENT_TIMEOUT
to anythingExpected Behavior:
Failing after the time specified by
AIOHTTP_CLIENT_TIMEOUT
Actual Behavior:
Ignores
AIOHTTP_CLIENT_TIMEOUT
Environment
Open WebUI Version: 0.3.5
Ollama (if applicable): 0.1.44
Operating System: Linux
Browser (if applicable): Firefox
Reproduction Details
Confirmation:
Logs and Screenshots
Browser Console Logs:
[Include relevant browser console logs, if applicable]
Docker Container Logs:
Screenshots (if applicable):
[Attach any relevant screenshots to help illustrate the issue]
Installation Method
Docker compose with host ollama
The text was updated successfully, but these errors were encountered: