
Workflow I made after coding my custom nodes in :
https://github.com/doomy23/ComfyUI-D00MYsNodes
What it does :
-
Load images and captions from a directory (add .txt or .caption files with the same name as the initial images)
-
Generate a new one with ControlNets
-
Generate new captions
-
Save image and caption in .txt file
-
Put the output folder the same as the input folder to create an infinite loop
What is the point :
-
You put a bunch of images with captions, sketches, or anything really...
-
You load the workflow
-
You press "Queue Prompt" 100x times
-
and 100 new images variations from random inputs will be generated
You can also tweak :
-
Negative prompts
-
Custom Guider CFG values
-
Change Canny ControlNet to Sketch or anything else with the correct preprocessor
Don't forget to install other other missing nodes as well.
If you have any idea of improvements let me know in the comments.
描述:
-
Removed upscaling
-
added image size
-
fixed metadata
Removed upscaling
added image size
fixed metadata
训练词语:
名称: generateRandomImagesInA_v11.zip
大小 (KB): 4
类型: Archive
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success