AIbase
EN
Home
Distilbert Base Uncased Finetuned Emotion
D

Distilbert Base Uncased Finetuned Emotion

Developed by frahman
A lightweight text sentiment classification model based on DistilBERT, fine-tuned on the emotion dataset with an accuracy of 92.05%
Text Classification
Transformers
Open Source License:Apache-2.0#High-precision sentiment classification#Lightweight BERT#Text emotion analysis
Downloads 17
Release Time : 3/2/2022

Model Overview

This model is a fine-tuned version of DistilBERT, specifically designed for text sentiment classification tasks, capable of identifying emotional categories in text.

Model Features

Efficient and Lightweight
Based on the DistilBERT architecture, it is 40% smaller than standard BERT, 60% faster, while maintaining over 90% accuracy.
High Accuracy
Achieves 92.05% accuracy and 92.07% F1 score on the emotion test set.
Fast Fine-tuning
Requires only 2 training epochs to achieve excellent performance.

Model Capabilities

Text sentiment classification
Natural language understanding

Use Cases

Sentiment analysis
Social media sentiment monitoring
Analyze user emotions in social media posts
Can automatically classify into specific emotion categories
Customer feedback analysis
Automatically classify sentiment tendencies in customer reviews
Helps businesses quickly understand customer satisfaction
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