blankpink rose

 <lora:koreanDollLikeness_v15:0.5>  <lora:blankpink rose:0.2> <lora:blankpink rose2:0.3>

rose rose rose rose rose  .

lady  lady  lady  black eyebrows black brown eyebrows brown eyebrows brown eyebrows 

 long blonde brown hair   long blonde brown hair   kpop kpop kpop  kpop  kpop 



neg

child, little girl, child, little girl, child, little girl, child, little girl,  red lip ,  red lip 

留言

發佈留言

此網誌的熱門文章

abstract abstract abstract abstract 2d illusion nude lady van gogh style lady leg purehay reading books

 abstract abstract abstract abstract 2d illusion nude lady<lora:vangogh:0.25>  <lora:2D illustration:0.25> van gogh style lady leg <lora:purehay:0.25> <lora:picasso:0.25> purehay reading books japanese abstract abstract abstract abstract 2d illusion nude lady<lora:vangogh:0.25>  van gogh style lady leg <lora:purehay:0.25> <lora:picasso:0.25> purehay reading books   <lora:irasutoya:0.25> irasutoya
閱讀更多

Mikubill/sd-webui-controlnet#194 api

  "controlnet_input_image": [img], "controlnet_module": 'openpose', "controlnet_model": 'control_sd15_openpose [fef5e48e]', This is how json objects are formated and those that are necessary. My setup works fine this way with depth, pose and scribble. Img is a base64 string for the image that goes into the ControlNet UI on the SD WebUI. EDIT: control img can be added but procrastinating on it. If interested in doing so, just need to encode the img in the second index of the output that is in numpy array format. Good. Small advice: Move API stuff to another file, like  api.py Maybe we should add some examples in the readme.
閱讀更多