Type | |
Stats | 187 0 |
Reviews | (12) |
Published | Aug 30, 2023 |
Base Model | |
Training | Steps: 2,000 Epochs: 10 |
Trigger Words | ray neverland |
Hash | AutoV2 6B2BA388CC |
If you enjoyed this LoRA, please hit the like button and give it a 5-star rating. Your encouragement motivates me to keep creating LoRAs like this.
Recommended to use Hires Fix for better quality.
Model used: AbyssOrangeMix3
Recommended VAE: orangemix or kl-f8
Trigger words: ray, neverland
On all images from preview I used DPM++ 2M SDE sampler, but you can use what you like.
Upscaler (highres fix):
Detailed illust → Latenet (nearest-exact)
Denoise strength: 0.5 (0.5~0.6)Simple upscale: Swin IR, ESRGAN, Remacri etc…
Denoise strength: Can be set low. (0.35~0.6)
This LoRA was made using my new Dataset Automaker, which automatically make a dataset of 400 images for me from original anime screencaps.
Download dataset of 400 images to train your own LoRA:
https://drive.google.com/file/d/1ti_ZOLhWDoLIs0AjxqykZcjjzMKvaCLO/view?usp=sharing