Sign In

chen zihan 陈紫函

11
116
2
Verified:
SafeTensor
Type
LoRA
Stats
116
Reviews
Published
Dec 16, 2023
Base Model
SD 1.5
Training
Steps: 2,100
Epochs: 16
Usage Tips
Clip Skip: 2
Strength: 0.9
Hash
AutoV2
C13F0C19FE
default creator card background decoration
dantea's Avatar
dantea

This model is for learning purpose, publish nsfw contents with this model is strongly inhibited.

Purposely train with higher versatility, 0.8-0.9 weight works the best.

It's easy to be used to generate images with macro view such as full body, cowboy shot, etc...

Do remember to fix the face with Face Detailer, it looks amazing.

LoRa training wasn't so difficult but it does costs me a lot time to prep the dataset and do captioning, kindly please leave a review with the photos you've generated with this model if you like it, it means a lot to me.