Disclaimer
Financial support is very much appreciated, but please do so by donating to my Ko-Fi instead of sending me Buzz, as I currently have no uses for CivitAI Buzz!
Description
An SDXL 1.0 LoRa made by AI_Characters for producing images with an artstyle similar to the game Darkest Dungeon - with the art mainly created by Chris Bourassa.
Recommended generation parameters
Weight: 1.0 (adjust down as needed when used in conjunction with other LoRa's)
Model: SDXL 1.0 with 0.9 vae
Steps/Sampler/CFG/Resolution: 25 steps | DPM++ 2M Karras | CFG 7 | 1536x1536
'Kohya Deep Shrink' settings: 4 blocks | 2 downscale | 0 start | 0.35 end | skip = true
Positive prompt: high-definition 8k
Negative prompt: text, blur, low-resolution, low-quality, grainy, glossy, poorly drawn, amateur, poorly shot, compression artifacts
'CLIPTextEncodeSDXL' ComfyUI node resolution: 4096x4096
Trigger words (must be included in the prompt for the LoRa to take effect!)
'comicish artwork in dgst artstyle'
Further usage notes
It doesn't replicate the style perfectly as unfortunately it does not produce the typical black/not visible eyes as seen in-game. I tried a few different things to no avail.
Usage permissions
You may not sell this model, merges containing it, or images generated with it.
Credit me when using this model in a merge or hosting it on a generation service.
TL:DR Training Workflow
Train full finetune with a dataset of 20-100 images using this config on a Vast.AI rented A100 80gb GPU
Extract LoRa at 512/512 dimensions using these settings
Resize LoRa to 16 ranks using these settings
Changelog
v1.2 [FINAL]
complete rework; higher epoch count, trained with EMA, reworked captions, reworked dataset preparation, small trigger words change - overall much better likeness and quality at the cost of some slight bias and overtraining increase - final version for now
v1.1
retrained with new captions: slightly better style likeness, slightly better flexibility, less bias towards the subject matter of the training data; added "comic book" to the trigger word
v1.0
Initial release
Description
FAQ
Comments (6)
Amazing work!
Just tested it on the civitai generator. Seems to be working ok now :)
Aside from the trigger word, what do I put in the <lora:> part? None of the gallery pics seem to use this.
I hadn't played Darkest Dungeon before when I posted images using this model. Your model is honestly even better than the quality of the style in the game. Great work!
Oh, btw. We're launching a Creators Program. We pay $1 for ⚡1,000.
ima be honest, it only really works for single characters/multiple characters standing around, for interactions like fighting, touchign, kissing etc youll have to do something else
a GREAT lora tho!