Sign In

face of intercom

58
348
1.9k
15
Verified:
SafeTensor
Type
LoRA
Stats
348
1,902
Reviews
Published
Jan 20, 2024
Base Model
SD 1.5
Training
Steps: 23,184
Epochs: 252
Usage Tips
Clip Skip: 2
Strength: 0.4
Trigger Words
(((shouting, open mouth, looking down, angry,glaring,frown,furrowed brow,screen,close-up,looking at viewer, wall, television frame, display, camera, intercom))),
Hash
AutoV2
E9D6D9A1EC

In sukatto videos, there are often scenes where a character like a 'bad mom friend' steals something on her own, uses it incorrectly, and then yells at you through the intercom, saying, 'It's your fault I had a terrible experience!' I've created a LoRa for that scene. Mainly at 0.4, an image of the target character is generated on the intercom camera, but it may not work well, such as coming out of the frame or being in the foreground. Also, if you strengthen it to 0.6, an image that fits within the screen will come out, but the drawing is fixed, and there are parts where the style of the target character is broken. So please keep generating images until a good one comes out where the target character is correctly displayed.