Styles are trained on synthetic data and do not copy the styles of artists!
Photo 2 - is trained on real data (Dataset from Unsplash)
[M] Merged LoRAs
Recommended LoRA Strength (Weight): 0.8-1
Preview generated without using ADetailer & Hires. fix
Description
Repeats: 20
Epoch: 12
Image Count: 20
Network Rank: 32
Network Alpha: 16
Batch size: 2
Clip skip: 1
Training steps: 2400
Resolution: 768x1024
Train Resolution: 1024x1024
Model: ponyDiffusionV6XL_v6
WD14 Captioning
SmilingWolf/wd-v1-4-moat-tagger-v2
LoRA Weight: [0.8-1]
FAQ
Comments (27)
Having a lot of success stacking Smooth 2 (0.7) with Line Art (0.4)! Thank you for your awesome contribution. Looking forward to more styles and updates.
Rainbow Style LoRA elevated my pics from "pretty good I guess?" to "wow that's great!"
Take all my buzz!
Thank you!
I'm glad you like Rainbow Style =)
A note for ComfyUI users...
I got absolutely NOTHING out of the photo examples, until I added "RAW, photo, realistic" to the prompt.
It would seem that AUTO1111 activates some LORAs without prompts better than Comfy.
I'd recommend our Hero Modeler here add those prompts to the examples.
Why have you now prevented it from being used for onsite generation!?!?! what gives! now all pony gens will look trash
Little just stopped working for me as well, lame af
@Doookie the 2.5 styles lora has also stopped working for me too at the same time, perhaps its a site wide thing for some loras and it will only be temporary, as I cannot see why multiple posters would collude to pull their online access at the exact same time. If not though this is terrible news for us with poor gpu's that want to make decent stuff
@dfsj98jfndasjfvjm604 For sure, hoping for the best
I didn't do anything, I literally just woke up. I have no idea what's going on, a few hours ago everything was working and now even I have nothing working.
It's like that with all the models I've posted.
It's probably Civitai that's having the problem, so it's not my fault.
I CAN'T DO ANYTHING ABOUT IT.
@prgfrg23 No problem, didn’t mean to come off so aggressive
It works now.
@siyijoh551455 Of course it works I changed the base model back to SDXL instead of Pony (If I had known how much trouble this model would cause, it's as beautiful as it is terrible)
Are there some technical difficulties or was it blocked for usage? I get: Failed to generate Some of your resources are not available for generation.
I didn't do anything, I literally just woke up. I have no idea what's going on, a few hours ago everything was working and now even I have nothing working.
It's like that with all the models I've posted.
It's probably Civitai that's having the problem, so it's not my fault.
I CAN'T DO ANYTHING ABOUT IT.
@prgfrg23 Hey, man! Thanks for clarifying, hopefully civitai team will resolve it. Btw great stuff, love your model.
@prgfrg23 Thanks for the clarification, hopefully the guys at civitai will sort it and it's not been banned. Havin so much fun with it
BRUH...Failed to generate
Some of your resources are not available for generation
These loras are amazing, been having a blast mixing and matching and found a lot of different styles that i really like, hope it comes back online for in site generation soon.
BTW any suggestion for anime style nocturnal low light images? despite any prompt i use there is always a lot of light coming from somewhere
I too hope everything will work again, and about the low light images, to be honest I don't know I'm trying to solve it myself
The only thing I can do is to train LoRA but not now, now I'm working on Photo 2 as soon as I'm done I'll solve the low light images issue or try to
SDXL always keeps a fixed brightness across the image, so it's impossible to get a dark nighttime image through the standard pipeline.
There is an [A1111 extension](https://github.com/Seshelle/diffusion-noise-alternatives-webui) to change the starting noise. To get a nighttime image, set blue component to max but lower green and red to about 128. That works well.
There might also be a way to train a LoRA to produce night-time images, but I have never tried it nor have I seen it done by others.
@pabana_faen intersting! will take a look at it later, tho never used auto1111 so far, thank you.
@b1te I made a LoRA to create low light images, not sure if it's what you need but you can try it anyway =)
@prgfrg23 woo amazing, with some combinations on the others i managed to create nice night outdoors images, still with a bit too much lightning, with this i think i can create about anything i want. civit ai onsite generator is being finicky lately but in some tests i did this lora seems to fix the excessive light. thank you!
I think the issue is coming from Pony Diffusion v6 as none of the pony diffs are working with it. So it's not just your product
I made the decision to return the base model to SDXL instead of Pony to make the LoRAs work in the Civitai image generator, And also please don't use these LoRAs on models other than Pony Diffusion V6 XL, this model is too different from other SDXL models
I don't understand anything at all with LoRA character everything works but with these styles it doesn't! One Image Style LoRAs work, too
Everything seems to be working right now.
I literally have a very bad headache from all this already! I am so sick of it that I want to write in my own language what I think about it all, not to translate it all with an online translator.
Sorry for the confusion and trouble you are feeling! Thank you for quickly helping things get back online. people are aware it was not your fault as other lora's went offline at the same time! know your work is very appreciated
Saad, your pony loras are one of the best! Hope you will make more soon!
Details
Files
Rainbow Style SDXL_LoRA_Pony Diffusion V6 XL.safetensors
Mirrors
Rainbow Style SDXL_LoRA_Pony Diffusion V6 XL.safetensors
Rainbow Style SDXL_LoRA_Pony Diffusion V6 XL.safetensors
Rainbow Style SDXL_LoRA_Pony Diffusion V6 XL.safetensors
Rainbow Style SDXL_LoRA_Pony Diffusion V6 XL.safetensors
Rainbow Style SDXL_LoRA_Pony Diffusion V6 XL.safetensors
hf_model_48_17.safetensors
Rainbow Style SDXL_LoRA_Pony Diffusion V6 XL.safetensors
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.








