Few days ago, I was fine-tuning LLMs on documents filled with comments, strikethroughs, and edits. I couldn’t find the right tool to get the data in correct format, so I created a Python script.

This script allows users to extract comments and their associated text from a DOCX file.

The output can be presented in multiple formats, and the script provides the option to include additional details such as the author’s name and timestamp.

Features

  • Extract comments and the text they are associated with. It handles parent-child relationships of comments too.
  • Display in original, enhanced, or JSON format.
  • Option to include the author’s name and timestamp of the comment.

Github Repository can be found here.

<
Previous Post
Myth-Busting: Vector Stores in AI
>
Next Post
Google’s Sign-In Alert: Why the Long Notice Matters for Your Security