Modelpublic

logakarthikeyan/spam-classifcation-model-bert

sourceHugging Facemitupdated 7mo agoView on Hugging Face

0likes17downloads

Model Card

BERT for SMS Spam Detection

This model is a fine-tuned version of bert-base-uncased on the ucirvine/sms_spam dataset. It achieves high precision and recall for identifying promotional or fraudulent messages.

Training Results

During training, the model showed rapid convergence, reaching over 99% accuracy by the third epoch.

Epoch	Training Loss	Filtered Train Acc	Validation Loss	Validation Acc
1	0.0914	0.9711	0.0400	0.9901
2	0.0215	0.9951	0.0530	0.9892
3	0.0058	0.9987	0.0513	0.9910

Final Evaluation

The final model performance on the test set:

Classification Report

Class	Precision	Recall	F1-Score	Support
Ham (Not Spam)	0.99	1.00	0.99	966
Spam	0.98	0.95	0.97	149
Accuracy			0.9910	1115

Usage

python

from transformers import pipeline

classifier = pipeline("text-classification", model="your-username/your-model-name")
result = classifier("Congratulations! You've won a $1,000 Walmart gift card. Click here to claim.")
print(result)