This project automates the creation of AI-generated summaries for videos, allowing for easy integration into a wiki. The primary aim is to streamline the process of reviewing, editing, and publishing summaries so they can be referenced and linked in a centralized knowledge base.
The folder structure is as follows Transcripts/Uploader/Year/Month/Date/Video ID/Video ID.ext
This allows for easily finding the needed file based on the upload date and channel while still being machine navigable.
- By Date:
- Each date folder contains files related to a specific video.
- The key file to edit and publish has the
.wiki
extension.
- Read the entire summary: At least once with a critical eye to ensure accuracy and alignment with the video.
- Delete any affiliate links: While steps are taken to remove them, they may be present in the text.
- Add internal links to:
- Relevant companies
- Product lines
- Commonly referenced terms or concepts
DO NOT Remove the AI Disclaimer.
Video directory: Louis Rossmann - Video Directory
Ensure all dependencies are installed:
- Python 3.x
- Non Standard libraries as required
- Access to a running LLaMA server
- Download Data: Retrieve the necessary video files and associated metadata.
- Generate Plaintext Transcript: Create a basic transcript from the video.
- Run AI Summary: Use the AI model to generate a concise and readable summary.
- Create
.wiki
File: Format the summary for wiki integration.
- Python 3.x
- requests library
- LLaMA server access for AI summary generation
Install dependencies using:
pip install requests
- Download Data
- Generate Plaintext Transcript
- AI Summary Creation
.wiki
File Editing- Publish to Wiki
- Update Video Directory with Wiki Link
Contributions to improve this process are welcome! If you have suggestions or improvements, feel free to submit a pull request or create an issue.