PDF to Text Chroma Search

Python scripts that converts PDF files to text, splits them into chunks, and stores their vector representations using GPT4All embeddings in a Chroma DB. It also provides a script to query the Chroma DB for similarity search based on user input.

Requirements

• Python 3.x
• PyPDF2
• chromadb
• langchain

Installation

1. Clone the repository:

git clone https://github.com/your-username/pdf-to-text-chroma-search.git

2. Install the required dependencies:

pip install PyPDF2 chromadb langchain

Usage

Script 1: Convert PDFs to text, split into chunks, and store in Chroma DB

1. Place your PDF files in the input directory.
2. Run the following command to convert the PDFs to text, split them into chunks, and store their vector representations in the Chroma DB:

python write_script.py

Script 2: Load Chroma DB and query user input

1. Run the following command to load the Chroma DB and query user input:

python read_script.py

2. Enter your query when prompted.