Gemma OCR

A Streamlit application that extracts text from images using Google's Gemma 3-4B model via Clarifai.

Features

• Upload images in various formats (JPG, JPEG, PNG, BMP)
• Extract text from images using the Gemma 3-4B model
• Download extracted text as a file
• Responsive UI with image preview

Requirements

• Python 3.10+
• Streamlit
• Pillow
• Clarifai SDK

Setup

1. Clone this repository:

   git clone https://github.com/Sumanth077/gemma_ocr
   cd gemma_ocr
   

2. Install the dependencies:

   pip install streamlit pillow clarifai
   

3. Set your Clarifai Personal Access Token (PAT) as an environment variable:

   export CLARIFAI_PAT="your_clarifai_pat"
   

4. Run the application:

   streamlit run ocr_app/app.py