Type | |
Stats | 1,714 7,667 2.1k |
Reviews | (238) |
Published | Nov 21, 2024 |
Base Model | |
Training | Steps: 1,050 Epochs: 10 |
Usage Tips | Clip Skip: 1 Strength: 1 |
Trigger Words | Dark |
Hash | AutoV2 5D109E9CD2 |
Warning: NoobAI users, lower strength to about 0.2—full support is still experimental.
Seriously Consider using my settings and using Wai NSFW for this. Im working on it, Theres a bunch of other issues you can see from like my oc Lora clashes... im working on it
Works well combined with phoenicians low light chiaroscuro
This IL model was trained on WAI-NSFW-Illustrious-SDXL and NoobAI checkpoints. Not all IL checkpoints will work as shown in my previews. Let me know in the comments if a checkpoint doesn’t work and Ill look into it.
Optimized for creating detailed images in low-light and dark conditions... night photos
For best results... use my special settings... these were a secret for the longest time, these technically work for Sd1.5 aswell and was originally for Sd1.5
Positive: , [:(masterpiece,best quality,very aesthetic,absurdres,newest,:1.4 ):9]
Negative: transparent background, [:lowres,(bad quality,worst quality:1.2),bad anatomy,sketch,jpeg artifacts,ugly, poorly drawn,(censor),blurry,watermark,simple background:5] [(masterpiece,best quality, very aesthetic,absurdres,newest:1.4 ):2]
There is still plenty to do with this one, ill be addressing them over time as I create and build up my datasets
This upload is part of my cadence improvement strategy. I’m releasing functional but experimental LoRAs. Expect steady updates as I refine them to meet final quality standards.