CodeK2.AI

• 

  Markdown-formatted project overview for the ERNIE AI, showcasing the inspiration, learning outcomes and deployment of the Codek2.AI web app

• 

  The GitHub Actions workflow for "pages build and deployment" ran smoothly with all jobs—*build*, *report-build-status*, and *deploy*

🌟 Inspiration
The idea behind this project was sparked by the challenge of turning static PDF documents into live, usable web content. I was inspired by the power of AI tools like "PaddleOCR" and "ERNIE" and how they can automate tedious tasks like extracting and presenting information in a clean format.

🛠️ What I Learned
Through this project, I learned how to:

• Use "PaddleOCR-VL" to extract both text and layout from a PDF document.
• Convert extracted content into "Markdown" format.
• Use "ERNIE" to generate a basic webpage based on that content.
• Deploy a website using 'GitHub Pages'.

🧱 How I Built It
I followed these main steps:

1. Chose a sample PDF file.
2. Ran 'PaddleOCR-VL' to extract its content and layout structure.
3. Converted that into Markdown format.
4. Used the 'ERNIE model' to build a simple web page from the Markdown.
5. Published the final page using 'GitHub Pages'.

⚠️ Challenges Faced

• Understanding how to properly format the extracted layout in Markdown.
• Getting GitHub Pages to deploy correctly.
• Handling PDF files with complex formatting or images.

Built With

• github
• html
• paddleocr