This project is designed to transform complex text into simpler, more understandable content, specifically targeting three main domains:
- Scientific
- Medical
- General
Built on top of google/flan-t5-large, the model is fine-tuned for text simplification using domain-specific data, and deployed through a user-friendly Streamlit interface.
- π§ Model on Hugging Face: epd007/text-simplification-final-version
- π‘ Live App (if hosted): Streamlit Demo
- βοΈ Simplifies complex sentences into easy-to-read versions
- π Domain-specific support: Scientific, Medical, General
- π€ Fine-tuned
FLAN-T5transformer model - β‘ Interactive UI built with Streamlit
- π§ͺ Trained on high-quality paired simplification datasets
Base Model: google/flan-t5-large
Fine-tuned Model: epd007/text-simplification-final-version
Training Dataset: WikiAuto
Testing Dataset: GEM/wiki_auto_asset_turk