
This is my newest Metart Model Lora, Siya. The lora is trained on approximately 300 pics at 9436 steps. It is fairly flexible and works with numerous sd 1.5 checkpoints. When using LCM, use Euler A with 10-15 steps and a low cfg. Hires.fix is set at 4 with a denoise of 0.4.
You may need to negative prompt out nudity, as some of the tags and pics were from her nude modeling sets. Lol!
Lets see what you guys render with her! Feel free to comment and criticize (if you must).
As always, feel free to buy me a coffee:
https://www.buymeacoffee.com/refugedenied1
描述:
All new goodness!
训练词语: S4Y4
名称: 00_Saya_Maya_lora_v1.safetensors
大小 (KB): 73850
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success