I didn't have much luck w/ this 4-step LoRA (would love to see some examples posted) - I downloaded their 4/30/2024 8-step from Hugging Face and it works for me wrt getting crisp results.

Could you please explain the differences and usage of each of these 5 different versions separately?

how do I use this properly? also the cfg model won't show up in civitai helper, does any one have the text trigger?

Thank you for your work, but can I have more information. You realize that KSampler has more than 10 samplers and sampling methods, how to use your model?
I'm trying to generate an image with 2-10 steps gets just .... (it doesn't work very well:)
You have posted the images, please give us information on what and how. Otherwise we can guess until Chinese Easter.

Edit: 12 step CFG version

There is also a 12 step CFG version available. Would be great if that gets added here as well.

How To Use This? I'm Using This With Forge And It Doesn't Work.

So I tried using this in Automatic1111, but I'm not seeing a difference

Do I just drop the lora in the folder & select it, just as any other lora?

Or do I have to do something like installing nodes/workflow stuff?

I cannot for the life of me get the SDXL CFG 8 Step to work. I'm using comfy. Does anyone have a workflow I can glom off of?

I put sdxl 8-step version in lora file folder, but I could not see it at main interface, what's wrong?

How do you use this?

can you make it available to use on civitai generation?

can you make it available to use on civitai generation?

Please guys someone make some stubid proof video tutorial how to use this, in forge... i dont find any info about it... only see comfy ui videos...

what’s the conversion of steps and scale when this is active is normally? i’ve been using a hyper LoRA for awhile but now i can’t seem to find the right workflow. hypothetically if this LoRA was at 1 weight, CFG scale was at 1.1, and steps at 7, what would be the closest to that without this?

if u use forge/reforge/a1111, go to settings, in search put Lora, and check "Always show all networks on the Lora page (otherwise, those detected as for incompatible version of Stable Diffusion will be hidden)" and now it will show up in lora.

From my test, Euler Beta is the best, Euler A Beta looks good too, but adds artifacts/noise in some generations so is less stable.

This is a complex series of Loras to speed up SDXL.

It's the exact same concept as DMD2 but this one has many more different ones you can download. You probably shouldn't use 1 or 2-step unless you're on some kiosk or something. Higher steps will give you better quality.

This and DMD2 are directly comparable. I'll tell you in advance it's next to impossible to determine which is better, but each one brings out different things from various models.

I will try to explain this, it's not going to be simple:

The No-CFG ones (1,2,4,8 steps) mean the CFG should be 1.0 (in ComfyUi, it might be 0 in other programs, idk). If you set the CFG higher or lower, the generation time significantly increases and it starts interpreting the negative prompt too.

The CFG version 8-steps and 12-steps (not uploaded here, you can get it from Hugging Face) means the CFG is put back in, but the samplers all change and so far my experiments with these have not been a great success. CFG3-4 seems to be good for realism models even though they tell you higher. The generation time is inherently longer than the no-CFG, equivalent to setting a CFG on no-CFG loras.

Getting back to the No-CFG, you have to adjust the Lora strength based on which sampler you choose.

I tested the No-CFG 8-step Lora at 8 steps. This is again, where the CFG is supposed to be 1.0. (That said, I do prefer lower like 0.9 or 0.8 CFG even with the no-CFG ones, but it will increase generation time)

Three combinations I found working:

Euler Ancestral (Euler A, not Euler) - Lora Strength 0.90

DPM++ 2M SDE - Lora Strength 0.60

DDIM - Lora Strength 0.90 (not for realism, it mangles realism too much, but it's interesting)

All these give totally different images, but similar in quality.

The Lora strength is subjective, but if you put it too high for a certain sampler it will quickly overbake your images.

As for scheduler, a bunch of them work, start with SIMPLE, then NORMAL is very similar too. You can try them out. (Whatever you heard elsewhere will not necessarily work well on this one, so be sure to test before doing something like Karras or Exponential, those in particular are no good for this one).

That should get you started, good luck and do comment if you can improve on this.