Verified: 2 years ago
SafeTensor
Type | |
Stats | 1,104 |
Reviews | (131) |
Published | Jul 9, 2023 |
Base Model | |
Training | Steps: 2,400 Epochs: 8 |
Usage Tips | Clip Skip: 2 |
Trigger Words | songyi |
Hash | AutoV2 0B73E11137 |
If you like this LoRa add a ❤️ to receive future updates.
SD 使用和训练教程文档请访问:afdian ,问题反馈,交流请加群:
Prompt: best quality, songyi, 1girl, solo, long hair, looking at viewer, pink background, black hair, dress, bare shoulders, upper body, <lora:songyi_v1_2:0.7>
Sampler: DPM++ SDE Karras
Model: chilloutmix_NiPrunedFp32Fix
Clip skip: 2
Enable the Adetailer when generate the full body size image
ADetailer model: face_yolov8n
Notice:
don't use restore faces, use adetailer
add trigger words to generate images
use controlnet tile to upscale images