Egami art-style trained on NoobAI-XL v-pred. It works fine at 0.8 weight. Trigger word is "egamistyle". I'm just an amateur so sorry if it's not that great. Writing hypnosis in negative prompt is recommended to avoid blurry eyes.
Training details:
dataset_count = 28
optimizer = CAME
learning_rate = 0.0003
learning_rate_scheduler = cosine
epochs = 19
steps = 560
batch size = 1
Go to "About this version" for the training details of the respective version.
Description
Expanded a little bit the dataset and did some proper preprocessing to try to remove all the compression artifacts that were present. The training images were ai-upscaled 4x and then downscaled back to a 1024x1024 pixel count. Also changed the training parameters a little bit.
Some training details:
dataset_count = 52
optimizer = CAME
learning_rate = 0.0003
learning_rate_scheduler = cosine
lora_rank = 32
lora_alpha = 1
epochs = 12
steps = 624
batch_size = 1