D

Distilbert Base Uncased Finetuned Sms Spam Detection

Developed by mariagrandury
A DistilBERT-based SMS spam detection model, fine-tuned on the sms_spam dataset, achieving an accuracy of 99.21%
Downloads 274
Release Time : 3/2/2022

Model Overview

This model is a fine-tuned version of DistilBERT, specifically designed to identify whether a text message is spam. It achieves efficient text classification capabilities through a lightweight architecture.

Model Features

High Accuracy
Achieves 99.21% accuracy on the evaluation set, effectively distinguishing between normal SMS and spam messages.
Lightweight Architecture
Based on the distilled architecture of DistilBERT, it reduces computational resource requirements while maintaining performance.
Fast Inference
Faster inference speed compared to the original BERT model, suitable for real-time application scenarios.

Model Capabilities

Text Classification
Spam Detection
Short Text Analysis

Use Cases

Communication Security
SMS Filtering System
Integrated into mobile communication systems to automatically filter spam messages.
Can intercept over 99% of spam messages.
User Protection
Fraud Prevention Alert
Identifies potential scam messages and warns users.
Reduces the risk of users being scammed.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase