AIbase
EN
Home
Albert Base V2 Imdb
A

Albert Base V2 Imdb

Developed by textattack
Text classification model fine-tuned on IMDb dataset using TextAttack framework, based on ALBERT Base v2 architecture
Text Classification
Transformers
#Text Sentiment Classification#High Accuracy (89.2%)#Optimized for Short Texts (128 tokens)
Downloads 4,579
Release Time : 3/2/2022

Model Overview

This model is specifically designed for text sequence classification tasks, excelling in scenarios like sentiment analysis. It was trained using cross-entropy loss function with optimized hyperparameters

Model Features

Efficient Lightweight Architecture
Utilizes ALBERT's parameter-sharing mechanism to significantly reduce model size while maintaining performance
Adversarial Training Optimization
Enhanced model robustness through adversarial training via TextAttack framework
High Classification Accuracy
Achieves 89.24% accuracy on IMDb validation set

Model Capabilities

Text Classification
Sentiment Analysis
Adversarial Example Detection

Use Cases

Sentiment Analysis
Movie Review Sentiment Classification
Determines positive/negative sentiment for IMDb movie reviews
89.24% accuracy on validation set
Text Security
Adversarial Attack Detection
Identifies inputs tampered with by textual adversarial attacks
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase