Hi everyone,
I recently built CategorizeAI, an open-source Streamlit application that enables multi-model NLP text classification using Supervised Learning, Zero-Shot Classification, and LLM-powered classification. It provides:
- Supervised learning for experimenting, training labeled data and classifying text;
- Zero-shot classification using Hugging Face pre-trained models without prior training;
- LLM-powered classification with local Ollama models for reasoning-based categorization;
- Experiment management to track, save, and compare classification workflows;
- Batch and real-time predictions with detailed analysis and probability distributions;
- A clean, responsive UI built with Streamlit for easy interaction.
If you’re working with text classification—CategorizeAI provides a versatile framework to experiment with different approaches, compare results, and apply dynamic and adaptable classifiers models tailored to your needs.
Note: To use the app, you’ll need to run it locally. Check out the GitHub guide steps to do it.
Feel free to explore it and share your feedback, as it would be very appreciated!
Project Source: GitHub
Demo App: CategorizeAI