Hello folks - this is big!
This might come to a surprise for you, but I've always wanted to create a new "Base model" for illustrious without the "Anime" style that is so prevalent in it... unfortunately I also thought that it was impossible without a complete new training.
Well, after many years of merging models, I've decided to give it a shot and try to reach that objective by merging and balancing all of my models into one general "Western Art" model (and I wasn't even serious about it, since I was sure that those aesthetic could not "fit" into one checkpoint without huge compromises - but I wanted to see what those compromises were going to be).
Of course, it started as a huge mess ... but by developing specialized positive and negative LoRAs and rebuilding the model slice by slice with my custom model merger, I've been able to teach it how to behave, until I've reached this:
A model so stable and effective that somehow defeated in combat (Benchmarking) both my "Comics" and my "Toon" specialized models.
«Wait, but merged models are usually overcooked crap... is this going to be similar?»
If you like this model, consider bidding for it to keep it available for anyone to play on Civitai.
✦ Arthemy Western Art - Quick Start ✦
1_____Size: Around 1024 x 1344
2_____Prompt: Start from the prompts present here and tune the keywords based on your needs:
Quick Prompt
best quality, absurdres, dynamic pose, solo, (toon /(style/)) or best quality, absurdres, dynamic pose, solo, (western comics /(style/))
Quick Negative Prompt
low quality, worst quality, jpeg artifacts, (no_lineart), fisheye
3_____CFG: 3.5
4_____Sampler: Euler a
5_____Steps: 25 to 50
(Higher the Steps, between these two values, the better the results)
I hope you like my models and I can't wait to see what you all are going to create with it!
____________________________________