
Lora trained on screencaps of Princess Vivian from Disney's Sofia The Frist series.
This model was a test to try out a tagging tool, trying to come up with a more consolidated set of tags for the dataset compared to the automatic tagging that was used in the the Princess Amber lora.
Prompts:
princess vivian, green eyes, 1girl, solo
Her eye color wasn't learned intrinsically, so make sure to include green eyes in the positive if they aren't showing up properly.
Tags that can get her dress to look closer to her canon look are maroon dress or pink dress.
Without descriptions, her dress can come up with a lot more variety of styles and colors than what she normally wears (either because I didn't plan the tags enough to distingush her from other princesses that showed up in some pictures).
描述:
Second version. Same dataset, but with pruned eye color, hair color and hairstyle tags.
训练词语: princess vivian, green eyes, 1girl, solo
名称: PrincessVivianV2.safetensors
大小 (KB): 36983
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success