Feature request Summary
Analogous to ChatGPT's integration of mixed models, I would like to have a single place for AI requests.
1, connect ollama webui via openAI api to dall-e 3 image generation
2, be able to connect ollama webui to other image generation models which run locally
Example of how dall-e image generation is presented in chatGPT interface:
Originally created by @raphaelmaenle on GitHub (Feb 10, 2024).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/696
Originally assigned to: @tjbck on GitHub.
**Feature request Summary**
Analogous to ChatGPT's integration of mixed models, I would like to have a single place for AI requests.
1, connect ollama webui via openAI api to dall-e 3 image generation
2, be able to connect ollama webui to other image generation models which run locally
Example of how dall-e image generation is presented in chatGPT interface:
This would be awesome and a feature I miss. Curious how to integrate this. Using something from automatic1111 isn't as simple as using DALLE3 with ChatGPT. You have a lot of variables, models, tweaks etc that make an easy implementation complicated.
Dalle3 yields good results because it adds prompts to your existing prompt, and hides all the complicated settings.
<!-- gh-comment-id:1943939626 -->
@jonasbg commented on GitHub (Feb 14, 2024):
This would be awesome and a feature I miss. Curious how to integrate this. Using something from automatic1111 isn't as simple as using DALLE3 with ChatGPT. You have a lot of variables, models, tweaks etc that make an easy implementation complicated.
Dalle3 yields good results because it adds prompts to your existing prompt, and hides all the complicated settings.
This would be awesome and a feature I miss. Curious how to integrate this. Using something from automatic1111 isn't as simple as using DALLE3 with ChatGPT. You have a lot of variables, models, tweaks etc that make an easy implementation complicated.
Dalle3 yields good results because it adds prompts to your existing prompt, and hides all the complicated settings.
I think the holy grail is figuring out how to replicate that. Probably involves using a specialized fine-tuned model to parse the requested generation and create parameters to achieve it.
<!-- gh-comment-id:1957815120 -->
@justinh-rahb commented on GitHub (Feb 21, 2024):
> This would be awesome and a feature I miss. Curious how to integrate this. Using something from automatic1111 isn't as simple as using DALLE3 with ChatGPT. You have a lot of variables, models, tweaks etc that make an easy implementation complicated.
>
> Dalle3 yields good results because it adds prompts to your existing prompt, and hides all the complicated settings.
I think the holy grail is figuring out how to replicate that. Probably involves using a specialized fine-tuned model to parse the requested generation and create parameters to achieve it.
My immediate plan is to add a button for image generation and then we can refine the feature from there :)
<!-- gh-comment-id:1957829520 -->
@tjbck commented on GitHub (Feb 21, 2024):
My immediate plan is to add a button for image generation and then we can refine the feature from there :)
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @raphaelmaenle on GitHub (Feb 10, 2024).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/696
Originally assigned to: @tjbck on GitHub.
Feature request Summary
Analogous to ChatGPT's integration of mixed models, I would like to have a single place for AI requests.
1, connect ollama webui via openAI api to dall-e 3 image generation
2, be able to connect ollama webui to other image generation models which run locally
Example of how dall-e image generation is presented in chatGPT interface:
@jonasbg commented on GitHub (Feb 14, 2024):
This would be awesome and a feature I miss. Curious how to integrate this. Using something from automatic1111 isn't as simple as using DALLE3 with ChatGPT. You have a lot of variables, models, tweaks etc that make an easy implementation complicated.
Dalle3 yields good results because it adds prompts to your existing prompt, and hides all the complicated settings.
@ghost commented on GitHub (Feb 21, 2024):
would love this aswell, with auto1111 please :)
@justinh-rahb commented on GitHub (Feb 21, 2024):
I think the holy grail is figuring out how to replicate that. Probably involves using a specialized fine-tuned model to parse the requested generation and create parameters to achieve it.
@tjbck commented on GitHub (Feb 21, 2024):
My immediate plan is to add a button for image generation and then we can refine the feature from there :)
@Henglund commented on GitHub (Aug 1, 2024):
1234567890
192.168.4.1.