ColorfulXL-Lightning
Model Details
ColorfulXL merged with lightning loras (2,8 steps) from bytedance, for fast inference (3-6 steps).
Diffusers version: https://huggingface.co/recoilme/ColorfulXL-Lightning
High range of resolutions supported (576 - 1280), 576*832 example:
Also try Huge resolutions (with Kohya HRFix Integrated, SD Forge)
Simple inference in 100 line of code https://github.com/recoilme/100lineSDXL
Discord (with free generation bot): https://discord.com/invite/gsvhQEfKQ5
WIP: https://runliq.com/
Due to training LORA's on the base version of SDXL, there are problems with hands and faces, sorry for that:
But you may try ADetailer/Codeformer/HiResFix and so on
Model Details
Developed by: https://aiartlab.org/
Model type: Diffusion-based text-to-image generative model
Model Description: This model is a fine-tuned model based on colorfulxl https://civarchive.com/models/185258/colorfulxl
License: This model is not permitted to be used behind API services. Please contact [email protected] for business inquires, commercial licensing, custom models, and consultation.
Contact
* For questions and comments about the model, please join https://aiartlab.org
* For future announcements / information about AiArtLab AI models, research, and events, please follow Discord: https://discord.com/invite/gsvhQEfKQ5
* For business and partnership inquiries, please contact https://t.me/recoilme
Description
ColorfulXL merged with lightning loras (2,8 steps) from bytedance, for fast inference (3-6 steps).
FAQ
Comments (3)
you have your first image as x rated, so main civit shows model as blur.
I am loving the model, it does have some prompt adherence issues, and aim's heavily at 3d animation style so its more suitable to be used on prompts and work of that nature I'm finding.
It's insanely fast, higher steps result in sharper messier images, 5 steps at CFG 2 is the sweet spot, VAE Needed older samplers, does not take much away from speed, hopefully other AI gets 3M soon.
4 stars out of 5,
I'm fussy, I do love to control parts of image more as per feedback given.
I recommend 100% that this model should be in everyone's tool kit.
Looking forward to v2 Lightning with how amazing this is, oh depth of detail is exquisite.
Your generations are amazing, Thank you!
@recoilme Thank you very much, a generated work is only as good as the model and lora paired with it, plus time and perseverance into the prompt, and Editing when needed lol
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.