XLabel: eXplainable Labeling Assistant

XLabel: eXplainable Labeling Assistant

XLabel is an open-source Streamlit app that takes an explainable machine learning approach to visual-interactive data labeling.

Main repository: GitHub
Try out the app at Hugging Face Demo. No need to upload the data!


XLabel can:

  • Predict the most probable labels using Explainable Boosting Machine (EBM) from InterpretML.
  • Show the contributions of each feature towards the predicted labels.
  • Provide an option to write the labels directly into the data file (use XLabel.py) or save them in a separate file (use XLabelDL.py)
  • Support data with multiple labels and multiple classes.
  • Support data with missing values (thanks to EBM) and/or non-numeric categorical features.


  1. Prepare your data in CSV or Excel format. The corresponding columns of the labels must already exist in the file, and a few rows have already been labeled.
  2. Simply upload your file, select the number of labels and click Sample!
  3. Inspect the suggested labels and the heatmaps in the main screen, and change the labels as needed.
  4. Click the β€œSubmit Labels” button at the bottom of the page to save the labels.

Check out XLabel’s GitHub repository for more information.