Replies: 4 comments 7 replies
-
|
I want to second this. Unlike Wan, which is a video model that people use for images, this is strictly an image model. It's also pretty good. Krea already works in forge, but I think that this might be worth having too since it's both a standard image model and open source, and based on what I'm seeing, it might be pretty easy to train off of. Obviously, can't rush anything, but I'd also love to see compatibility added, so that we don't have to use comfy just to run it. |
Beta Was this translation helpful? Give feedback.
-
|
why is it so difficult to get new stuff on webui? can someone tell me? does any of the ai platform able to help code those stuff? |
Beta Was this translation helpful? Give feedback.
-
|
https://civitai.com/articles/18144/in-progress-qwen-image-lightning-in-forgewebui I working on it, need more time |
Beta Was this translation helpful? Give feedback.
-
|
Amazing project, any news on the progress? |
Beta Was this translation helpful? Give feedback.
-
|
Hi, thanks for your interest. |
Beta Was this translation helpful? Give feedback.
-
|
WebUI is dead, and I doubt it's ever coming back. |
Beta Was this translation helpful? Give feedback.
-
|
I'm not quite as skeptical that it is ever coming back because there's
simply nothing better than it right now as far as I can tell .
I think the original creator is probably done and well I don't want to
speculate overly on the politics or personal issues that may have caused
him to step back I think it is probably going to be a matter of someone
else of comparable skill taking it over and I certainly know that's not me
but i'm at least cautiously hopeful for the future
…On Fri, Aug 22, 2025 at 12:56 PM pflky ***@***.***> wrote:
WebUI is dead, and I doubt it's ever coming back.
—
Reply to this email directly, view it on GitHub
<#3002 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AHN7U7ZC7HQJW5VMRKIW6IT3O5YVZAVCNFSM6AAAAACDGR7U6OVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTIMJZGMYDQMQ>
.
You are receiving this because you are subscribed to this thread.Message
ID:
<lllyasviel/stable-diffusion-webui-forge/repo-discussions/3002/comments/14193082
@github.com>
|
Beta Was this translation helpful? Give feedback.
-
I think it's not coming back because there's simply no way to keep up. There's been literally several new architectures released since, and there still isn't even controlnet promax support, nor Flux controlnet support, and Flux was pretty much yesterdays news months ago. He devoted time to frame pack, but it quickly became obsolete and lacked LoRA support, and again there's been multiple architectures released since. To play catch up would be pretty much an impossible task at this point. And even though I prefer Forge over ComfyUI, and mostly don't like ComfyUI, there's simply no denying that ComfyUI is well ahead because of its constant support for new things. Being 1 year behind in the AI community is as good as being 10 years behind. And not only is it lacking in support for new things, it actually lost support for existing things, for example, Normal DSINE as a controlnet, which is actually the best controlnet, better than even Depth Anything. Not to mention the Flux samplers. I'm sure there's still more things I could name, but given all that, it's pretty clear catching up is impossible, and even if they tried, they'd have to also support all the architectures that are already yesterday's news, which means a lot of work for something very few people would actually bother to use, that they already can use with ComfyUI. At that point, the best you might get is maybe a completely new Forge that supports some future architecture in a year or two, that has zero support for anything that came before it, and by then ComfyUI will be soo far ahead that you'd have to wonder what the point of even bothering would be. And considering A1111 has been abandoned even longer, and Forge is based on A1111, that's again another reason that makes it all futile. Basically, Forge died with SDXL, and took its last gasps with Flux, which it never truly fully supported, and so Forge will likely remain an SD 1.5 / SDXL WebUI frozen in time. Several other WebUI variants came and went before and during A1111's rise. Things simply move too fast to keep up. At this point, the devs for Forge would ironically just be better off creating a WebUI reskin of ComfyUI to make it look and work like Forge, but even that wouldn't be so easy, because ComfyUI constantly breaks extensions from its constant updates, which means it would break with virtually every ComfyUI update, or basically weekly. And even that would kind of be impossible, because a lot of Forge extensions have no equivalent in ComfyUI. And I wouldn't even be surprised if image generation is even something many devs would even be interested in after all this time. Many probably had their fill and moved on to other interests. |
Beta Was this translation helpful? Give feedback.
-
|
Comfy UI is a for-profit company that needs more users rather than more competitors. I don't think it will freely share its code with other creative platforms. Just like WebUI has many forks, but to my knowledge, Comfy UI doesn't have any other forks |
Beta Was this translation helpful? Give feedback.
-
Yes, swarm and others, the code is open |
Beta Was this translation helpful? Give feedback.
-
|
I would like some updates |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Qwen-Image, is an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering and precise image editing.
https://huggingface.co/Qwen/Qwen-Image/tree/main
https://huggingface.co/QuantStack/Qwen-Image-GGUF/tree/main
This new model catch my attention by is absurd prompt adherence and quality, and it's open source. Hope to see on the best and fast platform available, SD Forge obviously.
Beta Was this translation helpful? Give feedback.
All reactions