README.md

metadata

license: apache-2.0
tags:
  - text-classification
  - fake-news-detection

Fake News Detection Model

This model is trained to detect fake news articles using DistilBERT.

Training Data

The model was trained on a dataset of fake and real news articles. The dataset was preprocessed to remove irrelevant information and to balance the classes.

Performance

The model was evaluated using 5-fold cross-validation. The average metrics across all folds are as follows:

Metric	Value
Accuracy	0.973
Precision	0.962
Recall	0.986
F1	0.973
ROC AUC	0.973

Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained("HugMi/M3-Assignment2")
model = AutoModelForSequenceClassification.from_pretrained("HugMi/M3-Assignment2")

def classify_text(text):
    inputs = tokenizer(text, return_tensors="pt")
    outputs = model(**inputs)
    predicted_class = outputs.logits.argmax().item()
    return predicted_class  # 0 for fake, 1 for real
---