RazakAIhub/distilbert-spam-classifier

DistilBERT Spam Classifier

Model Description

Fine-tuned DistilBERT model for SMS spam detection. Trained on the SMS Spam Collection dataset. Achieves 99.1% accuracy and 96.9% F1 score.

How to Use

from transformers import pipeline

pipe = pipeline("text-classification", 
                model="RazakAIhub/distilbert-spam-classifier")

result = pipe("Congratulations! You've won a free prize!")
print(result)  # [{'label': 'spam', 'score': 0.99}]

Training Details

• Base model: distilbert-base-uncased
• Dataset: SMS Spam Collection (5,574 messages)
• Training accuracy: 99.1%
• F1 Score: 96.9%
• Epochs: 3
• Learning rate: 2e-5

Labels

• ham — legitimate message
• spam — spam message

Author

Razak Shaik — VIT-AP University, CS Final Year