LEARNING
Merge model
Simple Euler 8 to 10 steps
Text encoder and VAE to add
added LORA Lightning for 10-step generation.
https://civarchive.com/models/1987788?modelVersionId=2250142
Description
LEARNING
Merge model
Simple Euler 8 to 10 steps
Text encoder and VAE to add
FAQ
Comments (24)
So this is Flux SRPO which is a Flux.1D Fine Tune basically, and you merged it with Qwen... Which you also merged in the Lighting model...
How did you do this? is there a workflow out there that shows how this is done?
No, I simply merged it with Forge. I tried it and was very surprised that it worked.
i try, doesn't work, is there any workflow we can take a look? thank's
on comfyUi ?
I don't have a workflow, I use forge sd and not comfyui
I just posted using Comfy. Drag one of my photos into comfy to see an example Workflow and settings.
@WillieF cool
Nice model, the flux chin is not as prevalent which is a win in itself
Great Work! After using Flux.1D for months, unable to find a better one for my self-made LORA's, this one here is working really good and much better than KREA. Thank you!
flux-qwen-image-turbo works phantastic!
super happy, thank you very much for the feedback
Sorry, I am not that much into it from the technical side - was it technically possible do what ever you did with Chroma? Or merging this with Chroma?
good suggestion i think i will try
What is qwen 2.2
Any suggestions on Sampler and Scheduler?
I use Euler Simple
Thank you i have tried most now i find it works well also
@Lightrift thank you to you
Hi, I'm curious about what do you mean by merging qwen 2509 and flux? Does it mean this model consider a combination of qwen2509 edit image model and flux base model?
You cannot merge a 20 billion parameter model (Qwen) with a 12 billion parameter model (Flux), that would be like trying to breed a cat with a dog. What you have here is just Flux SPRO output as your merge did not work.
do you have examples to backup your claim or just here hating?
@ishadowxx I have been merging , training and researching AI image models for nearly 3 years now , as well as holding a degree in computer science and I am a top 15 model creator on this site, so I do know what I am talking about. Where are you coming from?
OK, maybe I'm not very knowledgeable on the subject, but these are the two models I've used, and maybe it's just that the SRPO model stands out. You seem to know more about this than I do.
Your link is neither a "Text Encoder" or "VAE" it is a lightning LORA.
"Text encoder and VAE to add"
https://civitai.com/models/1987788?modelVersionId=2250142
Do you have links to a "Text Encoder" or "VAE" you recommend?