DistilBERT English Fake News Classifier

This model is a fine-tuned version of distilbert-base-cased, specifically optimized for the detection of English-language disinformation and fake news.

Model Description

The model distinguishes between "Real" and "Fake" news articles.

  • Model Type: Transformer-based text classification
  • Language: English (en)
  • Base Model: distilbert-base-cased

Training Data

The model was trained on a consolidated dataset of approximately 63,000 English news articles, sourced from various datasets including:

  • WELFake
  • WebzIO

Example

input_text: "Scientists discover that lemon juice cures all viral infections instantly."
output: "fake"

input_text: "The Federal Reserve announced a interest rate hike of 25 basis points today."
output: "real"

Downloads last month
2
Safetensors
Model size
65.8M params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support