Skip to content

Data Research, Preparation And Manipulators Nodes for Model Trainers, Artists, Designers and Animators. Captions, Visualizers, Text Manipulators

License

Notifications You must be signed in to change notification settings

daxcay/ComfyUI-DataSet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

52 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

COMFY-UI-DataSet

ComfyUI-DataSet

ComfyUI-DataSet is a Data Research, Preparation And Manipulators Node for Model Trainers, Artists, Designers and Animators.

You can contact me on my discord server: https://discord.gg/Z44Zjpurjp

New documentation is coming soon

image

1. DATASET_CaptionVisualizer

This node helps anyone to visualize captions in 3 different forms: word cloud, networkgraph and frequency graph.

image

Workflow

CaptionVisualizer.json

Inputs

TextFilePathList: List of file paths ['path','path','path']
WordCloudTop: Integer will select top N data to make world cloud
NetworkGraphTop: Integer will select top N data to make network graph
FrequencyGraphTop: Integer will select top N data to make freqency graph

Output

3 Images of Word Cloud, Network graph, and Frequency graph respectively. Images are also saved in visualize folder where the caption files are stored.

image

2. DATASET_TXTFileSaver

Saves text file in an innovative way. has mode Merge | Overwrite | SaveNew | MergeAndSaveNew. Useful for caption creators and manipulators.

image

image

Inputs

content: content string type
filename: name of the file without .txt extension
directory: the directory where the file will be saved
mode: Merge | Overwrite | SaveNew | MergeAndSaveNew

If directory does not exists new directory will be created.

Modes Explaination

  1. Merge: Old text in file content appended by new text content.
  2. Overwrite: Only new text content in file (if directory is a new directory without file it will create new file).
  3. SaveNew: Make a numbered file of original name with new text content.
  4. MergeAndSaveNew: Old text in file content appended by new text content and make SaveNew.

Output

Text file in directory.

3. DATASET_TagManipulatorByImageNames

There may be a possiblity where you named your caption file with the same name as your image file (without extension πŸ˜…). This node is useful then. You can add a concept in all the caption files at once.

image

Inputs

ImageNames: List of image names

Image names can be generated using this JDCN node

image

Tags directory: directory of the caption files.
Backup: when true a backup folder with last version of file will be saved in tags directory given.
Concept: the concept you want to add in caption files.

Concept writing format

concept int,concept int,concept int

where concept can be any string and int is the position of the "," in caption file where the addition will take place.

For example,

Caption:  a young woman, in red tshirt and black pant, day 

Concept: age 24 1,sunny 2

Result: a young woman, age 24, sunny, in red tshirt and black pant, day

sunny should be after pant, but ended up after age 24 because 2nd comma position was now 3rd after adding age 24.

Concept will be written like this:

Caption:  a young woman, in red tshirt and black pant, day 

Concept: age 24 1,sunny 3

Result: a young woman, age 24, in red tshirt and black pant, sunny, day

In case you made an error you can use backup folder files.

Output

Tags directory will have something like this.

image

3. DATASET_SearchAndReplace

Helps anyone manipulate caption files to search for a given text in SearchFor and replace it with the given ReplaceWith text.

image

Inputs

TextFilePathList: text file path list

You can use this to get the file path list

image

SearchFor: search for a string you give. it can be any thing (no regex simple plain string πŸ˜…)
ReplaceWith: replaces ```SearchFor```. it can be any thing (no regex simple plain string πŸ˜…)

Output

Text file in directory. Backup of all file is saved in sr_backup folder where the caption files are stored.


Installation

  1. Using comfy-cli (https://github.com/yoland68/comfy-cli)
comfy node registry-install ComfyUI-DataSet
  1. Manual Method
  • Go to your Comfyui > Custom Nodes folder
  • Run CMD from folder path box or right click on empty area and click open in terminal.
  • Copy and Paste this command git clone https://github.com/daxcay/ComfyUI-DataSet.git
  • Then go inside ComfyUI-DataSet with cmd or open new.
  • and type pip install -r requirements.txt to install the requirements.
  1. Automatic Method with Comfy Manager
  • Inside ComfyUI > Click Manager Button on Side.
  • Click Install Custom Node and Search for DRMN and Install this node:

image

  • Restart ComfyUI and it should be good to go
  1. Recommended Plugin

    ComfyUI-JDCN (https://github.com/daxcay/ComfyUI-JDCN)


CREDITS

β—‰ Daxton Caylor - ComfyUI Node Developer

β—‰ https://github.com/rafstahelin

  • Node Request & Testing

Support for DRMN ❀️

About

Data Research, Preparation And Manipulators Nodes for Model Trainers, Artists, Designers and Animators. Captions, Visualizers, Text Manipulators

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published