Skip to content

Prabhu204/Character-Level-CNN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

37 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Character-level-CNN

An implentation of ConvNets using Pytorch based on research paper Character-level Convolutional Networks for Text Classification

The topic classification of the text is a natural language processing.ConvNets were used for extracting infromation from raw signals, ranging from computer vision applications to speech recognition and others. Where as, text is considered as a kind of raw signal at character level and applying one-dimensional ConvNets to it.

Datasets

Dbpedia and Amazon polarity datasets were used.

Setup

Gpu: Nvidia GTX 1050 4GB

Runtime

For 10 epochs: 14h

For 25 epochs: In progress

Result

Coming soon

Usage:

Implementation of Character level CNN for text classification

[-a ALPHABET] 

[-m MAX_LENGTH]

[-p {sgd,adam}]
 
[-b BATCH_SIZE]
 
[-n NUM_EPOCHS]
 
[-l {0.01,0.001}] 
    
[-d DATASET]
  
[-g GPU] 
 
[-s SAVE_PATH]

[-t MODEL_NAME] 

[-r SAVE_RESULT] 

[-rn RESULT_NAME]

[-i IMPORT_MODEL]

Example usage:

Before start using this code please download required datastes and save them in a directory with a name Data

For training a model:

python train.py -d 'Path to dataset' -n 'num of epochs' -s 'Directory name to save trained model' -r 'Path to save result' -rn 'name_of_the_result_file.txt'

For testing a model:

python test.py -b 'enter num of batch size' -i 'path of the saved model for evaluation'

About

No description or website provided.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages