AIbase
EN
Home
Distilbert Base Uncased Finetuned Sst2
D

Distilbert Base Uncased Finetuned Sst2

Developed by winegarj
A lightweight model fine-tuned for sentiment analysis based on the DistilBERT base model, achieving an accuracy of 90.37%
Text Classification
Transformers
Open Source License:Apache-2.0#Sentiment Analysis#High Accuracy#Text Classification
Downloads 2,556
Release Time : 4/9/2022

Model Overview

This model is a fine-tuned version of DistilBERT, specifically designed for text sentiment analysis tasks. As a lightweight variant of BERT, it retains most of the performance while significantly reducing the model size.

Model Features

Efficient and Lightweight
40% smaller in size compared to the original BERT model, with 60% faster inference speed while retaining 97% of the performance
High Accuracy
Achieves an accuracy of 90.37% on the evaluation dataset
Quick Fine-tuning
Only requires 5 training epochs to achieve excellent performance

Model Capabilities

Text Classification
Sentiment Analysis
Sentence-level Feature Extraction

Use Cases

Sentiment Analysis
Product Review Analysis
Automatically determine the sentiment polarity (positive/negative) of user reviews
Accuracy exceeds 90%
Social Media Monitoring
Real-time analysis of sentiment polarity in social media posts
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase