cyborg lady

  <lora:cyborg:0.2>  <lora:japaneseDollLikeness_v10:0.2> <lora:koreanDollLikeness_v15:0.6>  

cyborg kpop lady, CINEMATIC STILL, CINEMATIC STILL,  







 <lora:cyborg:0.2>

cyborg lady, CINEMATIC STILL, CINEMATIC STILL,  

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