THRILLustrious

By pAInCREAT0R

"Thumbs Up Reviews" are appreciated as it is the only way creators not in CIVITAI's creator program get their work noticed. - Thanks!

I recommend trying CREAT0R_MIX as an alternative.

THRILLustrious -Putting the THRILL and more 'real' into Illustrious!

THRILLustrious takes the IllustriousXL model and adapts it with a more realistic aesthetic. The goal of the training was to maintain as much of the core concepts of Illustrious as possible. It is a challenge at times because much of what makes Illustrious special is tied to Anime/Booru. I think THRILLustrious is striking as good of a balance between the two as possible.

Works good with multiple sampler/optimizer combos. Most of my images are using the following settings:

35-50 steps

Euler-Ancestral (Euler-A)

Beta, SGM Uniform, or Normal

Upscale - I use various methods on this. V5 is not as reliant on this for good face.

I do not use Adetailer, but I know some people like it. I always struggle with it in COMFYUI, but I did like it when I used A1111.

Version notes:

V7 - Refined with a smaller new data set then applied a .25 backward pass with V6 to preserve concepts. Minor improvements in aesthetics - especially of people.

V6 - The latest refinement of targeted training. It may be the last one for awhile while a NOOBAI model is in the works.

V5 - This is more of a true update than V4. It retains the VAE fix with some further targeted training on faces and details.

V4 - This was more o f a 'hot fix' than a full update. The baked in VAE in V3 had somehow been corrupted in a way that introduces some 'rainbow' like artifacts in certain images. V4 cleaned the VAE and seems to have resolved this issue.

V3 Real ILL - This was a generational jump forward for Thrillustrious over V2. Better overall visuals, better details, and more realism. It was the result of targeted training over many hours. IMO, one of the biggest leaps any of my models have taken between version releases. It is finally starting to compete in my mind with my PONY and SDXL models. This version is V2 very heavily trained and then merged back at .85/.15 with V1.

V2 - Just added a lot of training on a subset of about 8000 images while I was traveling. Enhanced details, faces, and overall quality.

V1 - V1 started with a completely fresh training on the base of Illustrious XL. It was merged with BETA then the resulting checkpoint was trained. This worked to continually reinforce the concepts in the base model while pushing the aesthetic toward realism. This approach diluted the residual impact of the models merged into BETA.

Beta - See below for my original approach to making Illustrious into THRILLustrious.

I do not need your buzz! If you like the model, please share some images and leave a positive review. Keep your buzz to help you create!

How BETA was made:

All training was done in Kohya starting with the Illustrious base model. No text encoder training at all - only UNET. With the recent Kohya updates, you can do higher batch and gradient sizes at faster speeds than ever. So on my 4090 I could batch 5 or 6, gradient 5 or 6 to get a virtual batch size of 25-36. This made it much faster to train than in the past. (Honestly, this was the only reason I attempted it after finishing PONY model.) There was some DareMerge of a small amount of my FULLY_REAL_EXTREME model and Redcraft Illustrious model to smooth out aesthetics and add a little more flare. About half the training was done at 1024 and half at 512. It could use another pass at higher resolution to help the faces. I preserved the clip from Illustrious 100% though and did not train encoders at all. (I figured I would mess it up.)

Disclaimer: This model is provided for educational purposes and for the advancement of AI art in general. The creator of this checkpoint assumes no responsibility or liability for the end user's outputs. Any illegal, unethical, or deep fake use of this model is the sole responsibility of the user engaged in those activities. The end user assumes all liabilities from use of the checkpoint.

No permissions are given for anyone to distribute this model on other sites. I do not give permission for it to be used on any image generation service other than CIVITAI.