ritik-vishwkarma/NLP-Emoji-Recommender

0

stars

7

commits

Python

primary language

Aug 27, 2025

updated

README

🎯 NLP Emoji Recommender

AI-powered emoji prediction system for text!
Built with FastAPI, PyTorch, and Transformers.
Ensemble of finetuned models for accurate emoji recommendations.


Demo Screenshot


🚀 Features

  • Ensemble Prediction: Combines multiple finetuned models for best results.
  • FastAPI Web App: Modern, responsive UI for interactive emoji recommendations.
  • REST API: /predict endpoint for programmatic access.
  • Configurable: Choose number of emojis and confidence threshold.
  • Health & Model Info Endpoints: For monitoring and debugging.

🖥️ Demo

Demo Screenshot


📦 Installation

git clone https://github.com/ritik-vishwkarma/NLP-Emoji-Recommender.git
cd NLP-Emoji-Recommender
python -m venv .venv
# On Windows:
.venv\Scripts\activate
# On Linux/Mac:
source .venv/bin/activate
pip install -r requirements.txt

⚡ Usage

  1. Download or place your finetuned model files in models_advanced/ directory.
  2. Run the recommender:
    python recommender.py
    
  3. Open your browser:
    http://localhost:8000
  4. API Docs:
    http://localhost:8000/docs

📝 Requirements

fastapi
uvicorn
torch>=2.0
transformers
numpy
pydantic

📚 API Endpoints

  • / : Web UI
  • /predict : POST, returns emoji recommendations for input text
  • /health : Health check
  • /models : Model info

🛠️ Model Files

Place your finetuned model files in models_advanced/:

  • best_distilbert_f1_0.4205.pth
  • best_multihead_f1_0.4010.pth
  • (Optional) best_deepmoji_f1_0.4828.pth

🤝 Contributing

Pull requests and issues welcome!


📄 License

MIT


Repo: https://github.com/ritik-vishwkarma/NLP-Emoji-Recommender

Contributors

ritik-vishwkarma/NLP-Emoji-Recommender

0

stars

7

commits

Python

primary language

Aug 27, 2025

updated

README

🎯 NLP Emoji Recommender

AI-powered emoji prediction system for text!
Built with FastAPI, PyTorch, and Transformers.
Ensemble of finetuned models for accurate emoji recommendations.


Demo Screenshot


🚀 Features

  • Ensemble Prediction: Combines multiple finetuned models for best results.
  • FastAPI Web App: Modern, responsive UI for interactive emoji recommendations.
  • REST API: /predict endpoint for programmatic access.
  • Configurable: Choose number of emojis and confidence threshold.
  • Health & Model Info Endpoints: For monitoring and debugging.

🖥️ Demo

Demo Screenshot


📦 Installation

git clone https://github.com/ritik-vishwkarma/NLP-Emoji-Recommender.git
cd NLP-Emoji-Recommender
python -m venv .venv
# On Windows:
.venv\Scripts\activate
# On Linux/Mac:
source .venv/bin/activate
pip install -r requirements.txt

⚡ Usage

  1. Download or place your finetuned model files in models_advanced/ directory.
  2. Run the recommender:
    python recommender.py
    
  3. Open your browser:
    http://localhost:8000
  4. API Docs:
    http://localhost:8000/docs

📝 Requirements

fastapi
uvicorn
torch>=2.0
transformers
numpy
pydantic

📚 API Endpoints

  • / : Web UI
  • /predict : POST, returns emoji recommendations for input text
  • /health : Health check
  • /models : Model info

🛠️ Model Files

Place your finetuned model files in models_advanced/:

  • best_distilbert_f1_0.4205.pth
  • best_multihead_f1_0.4010.pth
  • (Optional) best_deepmoji_f1_0.4828.pth

🤝 Contributing

Pull requests and issues welcome!


📄 License

MIT


Repo: https://github.com/ritik-vishwkarma/NLP-Emoji-Recommender

Contributors

Languages

Python

100.0%