If Flux is not supported, the Stable Diffusion webui project can be abandoned. Novices should not waste their time learning on this project.

[skyliwq]

If Flux is not supported, the Stable Diffusion webui project can be abandoned. Novices should not waste their time learning on this project.

[mnlkrp]

Atm FLUX has no LORAs, no finetuned models, no ControlNet, no anything basically.
Meanwhile SD1.5 and SDXL has it all. So no, Automatic1111 is still one of the best Stable Diffusion tools.

[cainezen]

There are many LORA(s), models fine-tuned and integrated with LORA(s), and ControlNET is supported in ComfyUI at least. So I'm not sure where you are at there.

[skyliwq]

It’s just a matter of time.

Atm FLUX 没有 LORA，没有微调模型，没有 ControlNet，基本上什么都没有。 而 SD1.5 和 SDXL 则拥有一切。所以，Automatic1111 仍然是最好的稳定扩散工具之一。

It’s just a matter of time, and moreover, there are other UIs as well

[dill-shower]

It’s just a matter of time.

Atm FLUX 没有 LORA，没有微调模型，没有 ControlNet，基本上什么都没有。 而 SD1.5 和 SDXL 则拥有一切。所以，Automatic1111 仍然是最好的稳定扩散工具之一。

It’s just a matter of time, and moreover, there are other UIs as well

This isn't a matter of time; it's about the fundamental limitations of the FLUX architecture, intentionally implemented by the developers. What they sold you as open-source and supposedly better than Stable Diffusion isn't actually that in practice. The only difference between FLUX and closed models like DALL-E is that FLUX uses your hardware for image generation instead of the company's servers.

All FLUX models you can find are distilled versions. Any LoRA created for them adds something but distorts the overall picture. Fine-tuning is impossible; all attempts to fine-tune the model so far have only resulted in quality degradation. This will continue until the developers release the original Pro model to the public. While StabilityAI has promised to eventually release their largest models as open-source, FLUX developers have never made such commitments. FLUX Pro, which is the original and most effective model, is only available through API. We can only dream that someday they'll cater to their fans and allow them to create their own models. For now, you can create LoRAs and train them for specific purposes, but this severely compromises the model's general capabilities. Your so-called open FLUX model is questionably open at best.

[highnrgappalachian]

You had this cringe take and posted it as an issue? Seriously?

[skyliwq]

This is not about whether weui and flex are easy to use. We need a comprehensive UI that can incorporate the latest technologies, such as Comfy UI and Webui Forge

[nitinmukesh]

I agree with OP. I recently moved to Forge.
Shut this project and archive it. Join the team at Forge and support it. More the better we will get new features in Forge fast. Forge is anyway Auto1111 WebUI but with support for newer models.

[E-Studios]

I agree flux isnt the checkpoint we are hoped for... out of the box it doesnt even work well... people say hands and feet work perfectly... no it aint. yes Flux looks great with realistic images but i think most used and most compatible is SD1.5/SDXL and Pony.

Anyways... yea A1111 really needs put more people together because its getting outdated each hour meanwhile ForgeUI is faster and supports even more stuff.

Forge is just like A1111 but better. (still using A1111) atm for most of the work because im lazy to set it up.

[Tower13Studios]

The diffusers in Forge creating false results, Forge is better when you wish to create with FLUX or exotic models, other than that Automatic1111 beats Forge and Comfy combined by a gazillion of landslides.

[dill-shower]

I agree with OP. I recently moved to Forge. Shut this project and archive it. Join the team at Forge and support it. More the better we will get new features in Forge fast. Forge is anyway Auto1111 WebUI but with support for newer models.

You can make a pull request and if it doesn't break anything, it will be accepted. But spending effort on supporting a hopeless model for which the developers decided not to even publish documentation is wrong. Even sd3, which was criticized so much on reddit, has more prospects and its support in webui is not ideal.

[nitinmukesh]

Hopeless model. I think you are not keeping yourself up-to date.
I am not developer who can raise PR, just an end user.

[dill-shower]

Hopeless model. I think you are not keeping yourself up-to date. I am not developer who can raise PR, just an end user.

Stable Diffusion 1.4 was released in early August 2022. In October 2022, the anime image generation model NovelAI, based on SD1.4, became publicly available (leak).

Flux1 was released on August 1, 2024. As of September 13, 2024, the main theme for all FLUX fans is that WE FINALLY LEARNED HOW TO RUN IT (not completely and not quite correctly yet) and WE ARE CREATING LoRAs (defective, broken ones), "We think we've figured out how it works" (which looks pretty sad compared to Stability AI's official publications detailing the structure of each of their models).

FLUX is as promising as an open source diffusion model – as Windows is promising as an open source version of Linux

[Hackgets]

I simply do not need flux. For example, being able to prompt natural language is great, but I am not a native English speaker, so I can't benefit from it. Open source should be open to everyone, not only for enthusiasts.

[SunGreen777]

I am writing to you using GPT as a translator. You are also writing here not in your native language. Do you want the model to understand English, Russian, Chinese, Spanish, French, and Japanese perfectly? How large will it be? What system will support it?

[divineblessing]

From being unfamiliar with ComfyUI to getting used to it, although some operations are a bit cumbersome, the Flux model is incredibly powerful. I haven't opened SD for a long time. :(

[SunGreen777]

You can write about the lack of documentation and the absence of a complete model in open access, or you can simply compare the results. I am satisfied with the current models FLUX, even if they are not perfect. Now I can say that the cat’s whiskers should be pink and the tail blue, which I couldn’t do before

[dill-shower]

You can write about the lack of documentation and the absence of a complete model in open access, or you can simply compare the results. I am satisfied with the current models Midjorney, even if they are not perfect. Now I can say that the cat’s whiskers should be pink and the tail blue, which I couldn’t do before

[highnrgappalachian]

You're welcome to use FLUX wherever you want. Incidentally, the name of this repo is 'Stable Diffusion WebUI'.

But the truth is that the few snowflakes who are 'in agreement' with the OP have an agenda to make the contributors to this monumental project feel bad and abandon it.

Incidentally, one of the guys above who is supposedly here to champion FLUX, also just made a new inflammatory topic about Web-UI.

[SunGreen777]

It’s not about incitement, FORGE works, not everything but here the work has stopped. And there is no information. The point is that hundreds of people worked on extensions for a project that someone decided to abandon, everyone will be happy if it is handed over to knowledgeable people and modernized to support FLUX.
InvokeAI doesn’t say that the model is bad, but they use it.

[ngia103]

Relax guys, lets not rush into conclusion. Maybe something happened and the creator had to take a break, maybe he is working for Stability AI or others AI projects now. Maybe next month there will be a brand new A1111 Flux. We don't know. I mean we don't even know who he is and what is he even doing.

It would be a shame to see this project abandoned, but I am still very grateful for all the hard work put in by the creator and everyone involved over the past two years. If Stability decided to release something like SD3XL then I think we will be back in the game. If not, well, SD 1.5, SDXL and Pony are still usable in next 1-2 years. Flux is great but not yet mature and also the generation time is in minutes, not in seconds like SD.