GithubHelp home page GithubHelp logo

frankblood / chinese-text-classification-pytorch Goto Github PK

View Code? Open in Web Editor NEW

This project forked from 649453932/chinese-text-classification-pytorch

0.0 1.0 0.0 42.08 MB

中文文本分类,TextCNN,TextRNN,FastText,TextRCNN,BiLSTM_Attention,DPCNN,Transformer,基于pytorch,开箱即用。

License: MIT License

Python 100.00%

chinese-text-classification-pytorch's Introduction

Chinese-Text-Classification-Pytorch

LICENSE

中文文本分类,TextCNN,TextRNN,FastText,TextRCNN,BiLSTM_Attention, DPCNN, Transformer, 基于pytorch,开箱即用。

介绍

模型介绍、数据流动过程:我的博客

数据以字为单位输入模型,预训练词向量使用 搜狗新闻 Word+Character 300d点这里下载

环境

python 3.7
pytorch 1.1
tqdm
sklearn
tensorboardX

中文数据集

我从THUCNews中抽取了20万条新闻标题,已上传至github,文本长度在20到30之间。一共10个类别,每类2万条。

类别:财经、房产、股票、教育、科技、社会、时政、体育、游戏、娱乐。

数据集划分:

数据集 数据量
训练集 18万
验证集 1万
测试集 1万

更换自己的数据集

  • 如果用字,按照我数据集的格式来格式化你的数据。
  • 如果用词,提前分好词,词之间用空格隔开,python run.py --model TextCNN --word True
  • 使用预训练词向量:utils.py的main函数可以提取词表对应的预训练词向量。

效果

模型 acc 备注
TextCNN 91.22% Kim 2014 经典的CNN文本分类
TextRNN 91.12% BiLSTM
TextRNN_Att 90.90% BiLSTM+Attention
TextRCNN 91.54% BiLSTM+池化
FastText 92.23% bow+bigram+trigram, 效果出奇的好
DPCNN 91.25% 深层金字塔CNN
Transformer 89.91% 效果较差
bert 94.83% bert + fc
ERNIE 94.61% 比bert略差(说好的中文碾压bert呢)

bert和ERNIE模型代码我放到另外一个仓库了,传送门:Bert-Chinese-Text-Classification-Pytorch,后续还会搞一些bert之后的东西,欢迎star。

使用说明

# 训练并测试:
# TextCNN
python run.py --model TextCNN

# TextRNN
python run.py --model TextRNN

# TextRNN_Att
python run.py --model TextRNN_Att

# TextRCNN
python run.py --model TextRCNN

# FastText, embedding层是随机初始化的
python run.py --model FastText --embedding random 

# DPCNN
python run.py --model DPCNN

# Transformer
python run.py --model Transformer

参数

模型都在models目录下,超参定义和模型定义在同一文件中。

对应论文

[1] Convolutional Neural Networks for Sentence Classification
[2] Recurrent Neural Network for Text Classification with Multi-Task Learning
[3] Attention-Based Bidirectional Long Short-Term Memory Networks for Relation Classification
[4] Recurrent Convolutional Neural Networks for Text Classification
[5] Bag of Tricks for Efficient Text Classification
[6] Deep Pyramid Convolutional Neural Networks for Text Categorization
[7] Attention Is All You Need

chinese-text-classification-pytorch's People

Contributors

649453932 avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.