Verified: 9 months ago
SafeTensor
Type | |
Stats | 119 29 3 |
Reviews | (9) |
Published | Sep 2, 2024 |
Base Model | |
Training | Steps: 20,000 Epochs: 10 |
Usage Tips | Clip Skip: 1 |
Hash | AutoV2 682EA4D0C1 |
汉服/唐装写真LoRA,
权重0.4-0.8,
底模选择真实系就可以,示例图用的底模是majicMIX realistic 麦橘写实_v7,
AD修复开不开都可以,模型本身可以控住人脸,但是手部最好开修复,训练素材里手部模型较少,手部可能会崩坏。
LoRA of Hanfu&Tang style Photography/chinese clothes,
weight :0.4-0.8,
model:realistic model. The base model used in the example image is majicMIX realistic majicMX-v7,
AD: turn on or turn off is not significant,
The model itself can control outline of facial features, but it is best to use ad-hand. There are few close up of hand in the training materials, so the generation may be wrong.