AIbase
EN
Home
Distilbert Base Uncased Finetuned Emotional
D

Distilbert Base Uncased Finetuned Emotional

Developed by TieIncred
A lightweight text sentiment classification model based on DistilBERT, fine-tuned on the emotion dataset with an accuracy of 93.05%
Text Classification
Transformers
Open Source License:Apache-2.0#Sentiment Classification#High Accuracy#Lightweight BERT
Downloads 663
Release Time : 10/24/2023

Model Overview

This model is a fine-tuned version of DistilBERT, specifically designed for text sentiment classification tasks. Through distillation technology, it retains 90% of BERT's performance while reducing size by 40% and improving inference speed by 60%.

Model Features

Efficient and Lightweight
Uses distillation technology, 40% smaller than original BERT with 60% faster inference speed
High Accuracy
Achieves 93.05% accuracy and 93.09% F1 score on the emotion test set
Quick Deployment
Based on PyTorch framework, supports mainstream deep learning deployment environments

Model Capabilities

Text Sentiment Classification
Short Text Analysis
Sentiment Polarity Detection

Use Cases

Social Media Analysis
Tweet Sentiment Analysis
Analyze sentiment tendencies in social media posts
Can automatically label positive/negative/neutral emotions
Customer Service
Customer Feedback Classification
Automatically classify emotions in customer feedback
Helps prioritize negative feedback
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