
Note: 1) You can get something similar to this just with prompting, but I wasn't completely happy with the results
2) More passes and higher LoRA strengths get the best results
Supposedly works without a trigger, but I set it up to use the following:
"loose_curlz, a close-up view of a woman with a blonde<* change color to what you want*> and short <*options are short, medium and long*>length loose_curlz hair style"
It always gets the color right and usually gets the length.
I've been using strength 1.4 and 35 passes.
描述:
训练词语:
名称: 1069483_training_data.zip
大小 (KB): 9199
类型: Training Data
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
名称: loose_curlz.safetensors
大小 (KB): 18807
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success