distilbert-base-uncased-finetuned-ner Open-source Named Entity Recognition Model - Lightweight and Efficient for Recognizing Entity Information

Home

Distilbert Base Uncased Finetuned Ner

Developed by murdockthedude

A lightweight named entity recognition model based on DistilBERT, fine-tuned on the conll2003 dataset

Sequence Labeling

Transformers

Open Source License:Apache-2.0 #Named Entity Recognition #High-precision NER #Lightweight BERT

Downloads 15

Release Time : 6/5/2022

Model Overview

This model is a lightweight named entity recognition model based on the DistilBERT architecture, specifically designed for token classification tasks, and performs excellently after fine-tuning on the conll2003 dataset.

Model Features

Lightweight and Efficient

Based on the DistilBERT architecture, it is more lightweight than standard BERT models while maintaining high performance.

High Accuracy

Achieves 97.17% accuracy and 87.4% F1 score on the conll2003 dataset.

Fast Training

Only requires 3 epochs of training to achieve good performance.

Model Capabilities

Named Entity Recognition

Text Token Classification

Use Cases

Natural Language Processing

News Entity Extraction

Identify entities such as person names, locations, and organization names from news texts.

97.17% accuracy, 87.4% F1 score

Document Information Extraction

Extract key entity information from business documents.

Property	Details
Model Type	distilbert-base-uncased-finetuned-ner
Training Data	conll2003
Metrics	precision, recall, f1, accuracy

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	220	0.0993	0.8511	0.8780	0.8643	0.9721
No log	2.0	440	0.0732	0.8913	0.9122	0.9016	0.9783
0.1878	3.0	660	0.0681	0.8984	0.9186	0.9083	0.9797

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご

Distilbert Base Uncased Finetuned Ner

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 distilbert-base-uncased-finetuned-ner

🚀 Quick Start

📚 Documentation

Model Information

Training Procedure

Training hyperparameters

Training results

Framework versions

📄 License