mirror of
https://github.com/open-webui/open-webui.git
synced 2026-07-17 08:21:12 -05:00
[GH-ISSUE #9798] Model toggle/dont start ollama with image gen #102647
No Assignees
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: github-starred/open-webui#102647
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @pancakesareawesome-420 on GitHub (Feb 11, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/9798
Feature Request
Please provide an option to have no model selected to prevent this from happening when generating images through the chat bar.
Important Notes
When starting a new chat and clicking image generation (after setting up automatic 1111), the generation times can take over 11-30 minutes. However, if using their webpage (sd-webui), it only takes 2-5 minutes. Additionally, the model that should not be part of the image generation responds after automatic 1111 is finished.
CPU: Intel(R) Core(TM) i7-10750H CPU
Memory: 64.0 GB DDR4 2933
GPU: NVIDIA GeForce RTX 2060 (6GB VRAM)
Is your feature request related to a problem? Please describe.
Everything works excellently except for the image generation; it seems to attempt to execute the prompt simultaneously as it generates the image. Ollama loads a model as soon as the send button is pressed, which seems to be slowing down the generation tremendously. Everything is locally run off my computer
Describe the solution you'd like
I would like to use my OpenWebUI homepage to generate images instead of the sd-webui homepage.
Describe alternatives you've considered
Model toggle/dont start ollama with image gen or have it launch after.
An option to have no model selected to prevent this from occurring when generating images through the chat bar. I am not sure if that is what is actually causing the slowdown, but I observed in Task Manager that Ollama initiated one of my models simultaneously with the image generation.
Additional context
I have attached some screenshots illustrating the slowdown and comparison. As I am new to this, please let me know if additional information is required. Although the issue seems minor, I appreciate the excellent work you are doing and am curious if this could be addressed in the future. Still loving it for my house and my parents though just wanted to get the image gen setup for them and i know they arent patient enough to wait 10 mins but 2 maybe lol.