floyo logo
Powered by
ThinkDiffusion
floyo logo
Powered by
ThinkDiffusion

Detailed Auto Caption

Generate high-quality captions for LoRA training and automatically resize images to SDXL/Flux-compatible resolution.

599

Generate highly detailed captions for LoRA training and automatically resize images to SDXL/Flux-compatible resolution.

Key Inputs

  • Dataset Path: The folder path containing your source images.

  • Trigger word: add a unique trigger word prefix

  • Output Path: The destination folder for resized images and generated .txt caption files.

Read more

N
Generates in about 4 mins 1 sec

Nodes & Models

Florence2ModelLoader
Florence2Run
Label (rgthree)
Text Multiline
JWImageResizeToClosestSDXLResolution
FL_ImageCaptionSaver
LoadImageListFromDir //Inspire
LoadImageListFromDir //Inspire
Florence2Run
easy promptConcat

Generate highly detailed captions for LoRA training and automatically resize images to SDXL/Flux-compatible resolution.

Key Inputs

  • Dataset Path: The folder path containing your source images.

  • Trigger word: add a unique trigger word prefix

  • Output Path: The destination folder for resized images and generated .txt caption files.

Read more

N