elabeler

elabeler: NLP Text Labeling Tool

elabeler is an NLP text labeling tool that enables easy labeling of text data. It features a Streamlit frontend and a FastAPI backend for exporting labeled data. This tool is designed to help users efficiently upload, label, and manage the labels.

Features

• Streamlit Frontend: An intuitive interface for uploading CSV files, labeling text data, and managing labels.
• FastAPI Backend: An API for exporting labeled data with filtering options based on timestamps, batch IDs, and batch names.
• Single File Upload: Users can upload one CSV file at a time, ensuring a clean and organized labeling process.
• Export Options: Export labeled data in multiple formats, including CSV, JSON, and Parquet.

Directory Structure

 Dockerfile.fastapi
 Dockerfile.streamlit
 app
    __init__.py
    api.py
    labeling_app.py
    requirements.txt
 data
    labeling_data.db
 docker-compose.yml
 readme.md
 tests
     __init__.py
     conftest.py
     test_api.py

Prerequisites

• Docker
• Docker Compose

Setup and Installation

1. Clone the Repository:

   git clone https://github.com/msminhas93/elabeler.git
   cd elabeler
   

2. Build and Run with Docker Compose:

   docker-compose up --build
   

3. Access the Services:
   • Streamlit app: http://localhost:8501
   • FastAPI app: http://localhost:8000

Usage

Streamlit App

• Upload CSV: Upload a CSV file containing text data.
• Label Texts: Use the interface to label each text entry.
• Navigate Pages: Use the navigation buttons to switch between pages of data.
• Export Data: Export labeled data in CSV, JSON, or Parquet format.

FastAPI API

The FastAPI service provides an endpoint to export labeled data with optional filtering:

• Endpoint: /export
• Query Parameters:
  • start_timestamp: Filter by start timestamp (ISO format).
  • end_timestamp: Filter by end timestamp (ISO format).
  • batch_id: Filter by batch ID.
  • batch_name: Filter by batch name.
  • output_format: Specify the output format (csv, json, parquet).

Example request:

curl "http://localhost:8000/export?output_format=json"

Testing

1. Install Testing Dependencies:

   Ensure pytest and httpx are installed:

   pip install pytest httpx
   

Run Tests:

1. Execute the tests using pytest:

   pytest tests/
   

Contributing

Contributions are welcome! Please submit a pull request or open an issue for any improvements or bug fixes.