This LoRA aims to replicate the style of artist Bartolomeobari.
If you like this LoRA, please consider directly supporting the artist (don't forget that without the artist, this LoRA wouldn't exist).
Please don't pretend to be or publish in the name of the artist.
Please share you creations! I'm curious to see what you can achieve with my models!
Recommended parameters 1 (basic and universal):
Base model: Pony Diffusion XL v6, AutismMix... I've noticed that AutismMix tends to solve certain anatomical problems at the cost of the less obvious style.
CFG Scale: 6-7
Sampler: Euler a.
If you want less "faded colors" look and more contrast, you can use another sampler such as DPM++ 2m SDE Exponential which works great with hires fix.
Steps: 30 - 40 or more
LoRA Strength: 0.7 - 1.0
Inpainting or hires fix are your friend to fix faces and details, but I advise you to use with 15-20 steps.
Recommended parameters 2 (awesome on ComfyUI):
Base model: Pony Diffusion XL v6
CFG Scale: 3.5 - 4.0 (1.5 - 2.5 if you're using LCM-LoRA)
Sampler: LCM
Scheduler: simple (recommended to avoid very small noise / artifacts) / sgm_uniform
Steps: 30 (5 - 15 if you're using LCM-LoRA)
LoRA Strength: 0.8 - 1.0
It works surprisingly well even without the LCM-LoRA. Of course, adding the LCM LoRA will speed up the generation but you'll need to lower steps and CFG scale.
Then you can apply your hires workflow if you want to.
Another approach that works well too is to start generating your raw image using Euler A sampler, then refine it using LCM-LoRA with very few steps. This is preferred if you want a more complex background as LCM tends to simplify the details a lot, especially in the background.
Note: the yellow / warm color tone comes from the style itself. You can add some custom post processing nodes if you want a colder tone, assuming you're using ComfyUI.
This model comes with no warranty. Do not use this model for inappropriate purposes.
