✦ Arthemy Painter Illustrious - Quick Start ✦
With version 5.0, I've recalibrated the entire model using my custom merge tools.
My main objective was to provide more "composition freedom" and try to fix the "anime" aesthetic that most of the previous versions were naturally showcasing. This model has been fine-tuned in ComfyUI so it might not be optimized for other UIs.
As always, I strongly suggest to start from the following parameters - in order to have a reliable starting point before diving into your preferred style.
1_____Size: Around 1088x1344
2_____Prompt: Start from the prompts present here:
Quick Positive Prompt
oil_painting_(medium), (chiaroscuro), perspective, fantasy, (solo), ((sketch))
[Your Prompt Here!]Negative Prompt
2D, 3D, toon, limited palette, glowing
2_____CFG: 3.5 / 5.0
3_____Steps: 50
(I've noticed that 30 steps are not usually enough to solve the image completely)
4_____Scheduler / Sampler: Euler a, DEIS Karras
I hope you like my models and I can't wait to see what you all are going to create with it!
Description
FAQ
Comments (7)
I am not done testing yet but I found this model to be very flexible and takes almost any style loras i trow at it with ease. Goes from (almost) photorealistic to different comic artist and cartoon style and anything inbetween with a little push from loras and prompting!
Yes, my approach to checkpoints is:
Checkpoints needs to have 🎵 a little bit of everything all of the time 🎶 to work correctly and to be as flexible as possible for the user and the LoRAs he/she wants to apply - while retaining a certain "general approach".
If a checkpoint can only do a specific style... it should have been a LoRA. :P
Seems good. I just wish there was a checkpoint focused on retro DND and fantasy art . Sort of like the boomer model for 1.5. Frank frazetta, old school 90s and 80s art, pulp magazine covers, etc
In general my view is:
Checkpoints needs to point towards a direction, LoRAs are the tool you use to get specific.
I'm pretty sure you can generate something in that style with this model and a well trained LoRA on the subject.
A whole checkpoint with such a specific style, would end up full of blind spots and probably not as good as a general model with a specific LoRA attached. ✨
Arthemy 100% there's a Frank Frazetta LORA, another for Ken Kelley, Keith Parkinson, there's another that's just "Retro Fantasy Style". Then make sure you're using tags like oil-textured, painterly, blended brushwork, intricate highlights, and stuff like neon, cel-shading in the negative prompt
tigon_zayne The problem is: Are this LoRA good enough?
I mean, I've tried a few LoRAs trained on those style - just for fun - and I've noticed that they have been trained on a very limited set of images or with very minimal (or too detailed) prompts, which compromised the final effect.
Aw, it's no longer for onsite generation.
Details
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.