An Open-Source Library for Processing Documents in Apache Spark.
ScaleDP is a library for processing documents using Apache Spark. And this repository contains jupyter notebooks with tutorials and examples of usage ScaleDP library.
ScaleDP includes the following features:
- Load PDF documents/Images
- Extract text from PDF documents/Images
- Extract text Images/PDF documents, OCR
- Zero-shot Data extraction from the Images/PDF documents using LLM
- Object detection over Images/PDF documents
- NER/LLM over Images/PDF documents
- Visualize results on the images
- Compatible with Spark PDF Datasource
We advise to run all examples through Google Colab for the easiest setup. Google Colab allows you to run it for free. All examples were tested using Google Colab, so it should be the most stable platform. However, any other cloud provider or local run should work.
| Chapter | Notebook |
|---|---|
| 1: QuickStart with ScaleDP |  |
| 2: Example of usage different OCR | |
| 3: NER Examples | |
| 4: LLM NER Examples | |
| 5: With Spark PDF | |
- Python 3.11
- Apache Spark 3.5 or higher
- Java 8, 11
- Tesseract 5.0 or higher
- Jupyter Lab | Google Colab
