invoke-ai
[enhancement]: Please support FLUX.1
Is there an existing issue for this?
- [X] I have searched the existing issues
Contact Details
What should this feature add?
Please support the new open-weights models from Black Forest Labs, FLUX.1.
https://huggingface.co/black-forest-labs
Alternatives
No response
Additional Content
No response
After extensively testing the versatility and accuracy of the Flux 1 model I can confidently say adding this model to Invoke AI would be a very large +. The quality of this model is like none other, creating accurate text at 98% times of the requests, having the prompt accuracy of Dall-E yet the versatality of Midjourney supporting Art-mediums from Watercolor painting to Anime/Manga Sketches. A rather impressive creation of Black Forrest indeed.
I already have Flux working in an InvokeAI workflow
would you be able & willing to create instructions as to how you did this? I think it would be very useful for other users as well as Flux is a very high-end model but at the moment most people need to use ComfyUI to use it.
I already have Flux working in an InvokeAI workflow
would you be able & willing to create instructions as to how you did this? I think it would be very useful for other users as well as Flux is a very high-end model but at the moment most people need to use ComfyUI to use it.
New Diffusers release now supports FluxPipeline: https://github.com/huggingface/diffusers/releases/tag/v0.30.0
I already have Flux working in an InvokeAI workflow
would you be able & willing to create instructions as to how you did this? I think it would be very useful for other users as well as Flux is a very high-end model but at the moment most people need to use ComfyUI to use it.
Howdy folks, I have no doubt that the Invoke team will be more than aware of Flux, what a leap forward it is and how the community is very keen to get it working with their favourite tool.
Although it's great to see the enthusiasm in the comments I would suggest showing your interest with an emoji reaction on the post, freeing up the comments for folks that have suggestions on how to help implement it.
This will help keep the signal to noise ratio high and although I'm not a maintainer - I could imagine I might find it a little stressful or feel like it's being demanded / expected (although I'm sure that's not the case).
Just my 2c and a ❤️ out to the maintainers of such a great open source tool.
Fair point
On Fri, Aug 9, 2024 at 17:41 Sam @.***> wrote:
Howdy folks, I have no doubt that the Invoke team will be more than aware of Flux, what a leap forward it is and how the community is very keen to get it working with their favourite tool.
Although it's great to see the enthusiasm in the comments I would suggest showing your interest with an emoji reaction on the post, freeing up the comments for folks that have suggestions on how to help implement it.
This will help keep the signal to noise ratio high and although I'm not a maintainer - I could imagine I might find it a little stressful or feel like it's being demanded / expected (although I'm sure that's not the case).
Just my 2c and a ❤️ out to the maintainers of such a great open source tool.
— Reply to this email directly, view it on GitHub https://github.com/invoke-ai/InvokeAI/issues/6712#issuecomment-2278790316, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJSLCLGZ2AZJZEU5E6YQZO3ZQUZRNAVCNFSM6AAAAABL5FTWSGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENZYG44TAMZRGY . You are receiving this because you commented.Message ID: @.***>
I already have Flux working in an InvokeAI workflow
so was this a lie?
Yes because they didn't screenshot the import outcome like @MDsniper did. (╯°□°)╯︵ ┻━┻
Error is
[2024-08-10 13:49:18,230]::[ModelInstallService]::ERROR --> Model install error: black-forest-labs/FLUX.1-dev
InvalidModelConfigException: Unable to determine model type for /mnt/Store/Applications/InvokeAi/models/tmpinstall_dd9x7x7x/FLUX.1-dev;
class FluxPipeline is not one of the supported classes [
StableDiffusionPipeline,
StableDiffusionInpaintPipeline,
StableDiffusionXLPipeline,
StableDiffusionXLImg2ImgPipeline,
StableDiffusionXLInpaintPipeline,
LatentConsistencyModelPipeline,
AutoencoderKL,
AutoencoderTiny,
ControlNetModel,
CLIPVisionModelWithProjection,
T2IAdapter
]
This is why we use other interfaces The devs here dont care that you want it. Nor do they care about Playground 2.5, or Kolors, or Aura, Flux, SD 3, or presets for Pony, or for so much, so much, because you aren't a paying customer. Search all you like, there is no guide to install these major models. So, download pinokio and choose from several different UI's to use them.. Even if you have 4 gigs of vram.. Some people just care enough to get things to work (and in 1 click installs). pinokio.computer (web site/download the app) will only give you one problem.. harddrive space.
For what it's worth, I've been using "Draw Things" on my Mac, which already supports Flux. The UI is a little quirky, but it's a native app and runs very well. And it's free.
bump!
well, 4 thumbs down from devs I guess. Deleting invokeAI. I'll will try pinokio as @kwal559 mentioned
Thank you @sammcj
Folks, come on it's ok to be excited and even a little impatient at times but there's no need to rip out open source maintainers that write software for you for free. They owe you nothing.
We're thinking in the simple way. If other apps can run flux with a little research and upgrades, why anybody in community not making a PR for flux?
I'm fairly sure there are PRs underway to get Flux working, e.g. https://github.com/invoke-ai/InvokeAI/pull/6739 - it just seems they're not ready yet.
When you compare InvokeAI to some of the other tools out there that have implemented it - InvokeAI is more of a product, it's large and is generally held to high standards, this is quite different from something like Comfy which is esentially a lot of different peoples scripts and packages that can be glued together in various ways - people can make great things due to the flexibility but - it's also a lot harder and much less refined.
I'm sure that if it was as simple as simply supporting the model, invoke would already have flux, but while I have no insider knowledge I'm assuming it's probably included as part of a larger piece of work with the intent of a better integrated outcome. (I'm just guessing here).
You're right. I think I got a little too exited about flux. Also that pr was opened 2 days ago. I've subscribed now and will keep a close eye on it.