RNN-Classification

KO | EN

gitlite — search

#javascript #hacktoberfest #typescript #llm #go #rust #deep-learning #nodejs #awesome #kubernetes #vue #pytorch #ios #open-source

RNN-Classification

(★ 86)

classify text by rnn/lstm, based on TensorFlow r1.0

RNN-Classification 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# RNN-Classification  
本项目构建了基于RNN/LSTM模型的文本分类器，基于TensorFlow r1.0版本。

*This project constructs a text classifier based on RNN/LSTM model, based on TensorFlow r1.0.*

当前支持BASIC-RNN、GRU、LSTM、BN-LSTM。

*This project currently supports the BASIC-RNN, GRU, LSTM, and BN-LSTM models.*

本项目基于字符，即采用了character embedding的方法，故不需要分词。因此，语料用于生成字符索引，只需要包含所有用到的字符即可。

*This project, which uses the method of character embedding, is based on characters, so word segmentation is not necessary. Thus, the corpus, which only needs to contain all the characters used, is used to generate the character index.*

本项目主要针对中文文本。若直接应用到其它语言，如英文，在不修改代码的情况下，仍然基于字符(即单个字母、标点等)；如果需要基于词(即英文单词、中文词语等)，需要添加分词程序，并修改相应代码，如utils.py文件中的TextLoader/transform函数和TextLoader/preprocess函数等。

*This project is mainly for Chinese text. If it is applied directly to other languages, such as English, while not modifying the code, it is still character-based (ie, a single letter, punctuation, etc.). If you want to turn it into word-based (ie, English words, Chinese Words, etc.), you need to add the code about word segmentation, and modify the corresponding code, such as function TextLoader/transform and function TextLoader/preprocess in file utils.py and so on.*

## 运行方式
 + 命令行，运行train.py训练模型，运行test.py测试模型，所需参数用help查看。
 + 运行run.py文件，自行修改各项参数

## 代码说明
+ model.py: 根据给定的参数、用tensorflow构建分类模型
+ lstm_bn.py: 根据论文[*Recurrent Batch Normalization*](http://arxiv.org/abs/1603.09025)构建Batch Normalization的LSTM的Cell单元
+ train.py: 模型的训练
+ test.py: 模型的测试，分为sample（预测单个例子）、predict（预测文件中的所有例子）、accuracy（预测文件中的所有例子并验证）
+ cross_validation.py: 十折交叉验证
+ utils.py: 辅助函数，主要用于得到训练/测试的批数据（batch data）
+ run.py: 训练、测试、交叉验证模型，用python运行命令行，可自行修改各项参数
+ data: 训练数据、测试数据、十折交叉验证数据，第一列为文本text，第二列为类label，可修改位置并修改run.py文件或命令行运行的参数
+ save: 保存模型训练的各项参数，可修改位置并修改run.py文件或命令行运行的参数
+ utils: 语料、label对照字典等，可修改位置并修改run.py文件或命令行运行的参数
  - corpus.txt: 语料，*需要提前生成*
  - labels.pkl: 类标号与数字的对照，字典形式{类标号:从0开始的数字}，pickle形式，*需要提前生成*

RNN-Classification

RNN-Classification 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

RNN-Classification

RNN-Classification 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README