Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
jiachengyu 57ef1e5dcd | 1 year ago | |
---|---|---|
data | 1 year ago | |
models | 1 year ago | |
README.md | 1 year ago | |
binarize.py | 1 year ago | |
jumble_words.py | 1 year ago | |
predict.py | 1 year ago | |
test_words_set.py | 1 year ago | |
train.py | 1 year ago | |
vocabID.json | 1 year ago |
Data and re-implemented scripts for 2017 AAAI paper, "Robsut Wrod Reocginiton via semi-Character Recurrent Neural Network"
Last updated: Oct 8th , 2017
This repository contains the re-implementation on the following paper:
@inproceedings{DBLP:conf/aaai/SakaguchiDPD17,
author = {Keisuke Sakaguchi and
Kevin Duh and
Matt Post and
Benjamin Van Durme},
title = {Robsut Wrod Reocginiton via Semi-Character Recurrent Neural Network},
booktitle = {Proceedings of the Thirty-First {AAAI} Conference on Artificial Intelligence,
February 4-9, 2017, San Francisco, California, {USA.}},
pages = {3281--3287},
year = {2017},
url = {http://aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14332},
publisher = {{AAAI} Press},
}
.
├── README.md # this file
├── train.py # script for training
├── predict.py # predict correct word
├── models # store model files
├── binarize.py # utility function
└── jumble_words.py # script for adding jumble noise
(prerequisites)
# Tested on CentOS7 (with CUDA7.5)
> conda create -n robsut python=2.7
> pip install keras theano h5py
# change the keras backend to theano (edit $HOME/.keras/keras.json)
(training)
THEANO_FLAGS=device=gpu0,floatX=float32 python train.py
(predicting)
THEANO_FLAGS=device=gpu0,floatX=float32 python predict.py -m models/train_j-INT_n-JUMBLE_u-650_batch-20_ep-10_model.h5
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》