Dyslexia-Friendly Summarizer

A specialized text summarization model fine-tuned to produce summaries that are easier to read for people with dyslexia.

Model Details

  • Developed by: Suyog Ghimire(Team DysIsfine)
  • Model type: Text Summarization (Transformer-based)
  • Language: English
  • License: MIT
  • Finetuned from: facebook/bart-large-cnn

What This Model Does

This model takes long texts and creates shorter, simpler summaries that are easier to read. It uses clear sentences and common words to make the summary accessible for people with dyslexia.

Example

Input: "The Sun is the main source of energy for Earth. It provides light and heat that support life. Plants use sunlight to make food, which provides energy for animals. The Sun also drives weather patterns and the water cycle. Without it, Earth would be too cold to support life."

Output: "The Sun provides light and heat that support life on Earth. Plants use sunlight to make food, and this energy passes to animals. Solar energy also drives weather and the water cycle, making Earth suitable for living organisms."

How to Use

from transformers import pipeline

summarizer = pipeline("summarization", model="[your-model-name]")

text = "Your long text here..."
summary = summarizer(text, max_length=100, min_length=30, do_sample=False)
print(summary[0]['summary_text'])

Training Data

The model was fine-tuned on a custom dataset of text-summary pairs. The training data includes educational content, news articles, and informational passages with simplified summaries.

Best For

  • Educational content and learning materials
  • News articles and informational texts
  • Making long documents easier to understand
  • Accessibility in web apps and reading tools

Limitations

  • Works best with English text
  • May lose some details when simplifying
  • Should be reviewed by humans for important information (medical, legal, safety)
  • Works better with moderate-length texts

Model Card Contact

[email protected]

Downloads last month
-
Safetensors
Model size
0.4B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support