
First time training a character-focused model, so I figured I would go for a character with a unique outfit with lots of details like Rapi.
Trained on a dataset of 48 images. Might expand in the future.
Prompting tips:
-
Can get overbaked results at high values, I recommend somewhere between 0.6 and 0.8 but whatever works for you. Go lower if you're hunting for a specific pose.
-
Always include the keyword rapi-nikke for most consistent results.
-
Use with anime/stylized checkpoints.
描述:
Base model, occasionally subject to overcooking. I recommend a weight of 0.8
Base LoRA trained on aniwave-v1. I will be experimenting with other base models in future experimental versions.
训练词语: rapi-nikke
名称: RapiLoraV1.safetensors
大小 (KB): 147570
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success