Skip to content

bgawrych/gluon-nlp

 
 

Repository files navigation

GluonNLP: Your Choice of Deep Learning for NLP

GluonNLP is a toolkit that enables easy text preprocessing, datasets loading and neural models building to help you speed up your Natural Language Processing (NLP) research.

News

Installation

Make sure you have Python 3.5 or newer and a recent version of MXNet (our CI server runs the testsuite with Python 3.5).

You can install MXNet and GluonNLP using pip.

GluonNLP is based on the most recent version of MXNet.

In particular, if you want to install the most recent MXNet release:

pip install --upgrade mxnet>=1.6.0

Else, if you want to install the most recent MXNet nightly build:

pip install --pre --upgrade mxnet

Then, you can install GluonNLP:

pip install gluonnlp

Please check more installation details.

Docs 📖

GluonNLP documentation is available at our website.

Community

GluonNLP is a community that believes in sharing.

For questions, comments, and bug reports, Github issues is the best way to reach us.

We now have a new Slack channel here. (register).

How to Contribute

GluonNLP community welcomes contributions from anyone!

There are lots of opportunities for you to become our contributors:

  • Ask or answer questions on GitHub issues.
  • Propose ideas, or review proposed design ideas on GitHub issues.
  • Improve the documentation.
  • Contribute bug reports GitHub issues.
  • Write new scripts to reproduce state-of-the-art results.
  • Write new examples to explain key ideas in NLP methods and models.
  • Write new public datasets (license permitting).
  • Most importantly, if you have an idea of how to contribute, then do it!

For a list of open starter tasks, check good first issues.

Also see our contributing guide on simple how-tos, contribution guidelines and more.

Resources

Check out how to use GluonNLP for your own research or projects.

If you are new to Gluon, please check out our 60-minute crash course.

For getting started quickly, refer to notebook runnable examples at Examples.

For advanced examples, check out our Scripts.

For experienced users, check out our API Notes.

Quick Start Guide

Load the Wikitext-2 dataset, for example:

>>> import gluonnlp as nlp
>>> train = nlp.data.WikiText2(segment='train')
>>> train[0:5]
['=', 'Valkyria', 'Chronicles', 'III', '=']

Build vocabulary based on the above dataset, for example:

>>> vocab = nlp.Vocab(counter=nlp.data.Counter(train))
>>> vocab
Vocab(size=33280, unk="<unk>", reserved="['<pad>', '<bos>', '<eos>']")

From the models package, apply a Standard RNN language model to the above dataset:

>>> model = nlp.model.language_model.StandardRNN('lstm', len(vocab),
...                                              200, 200, 2, 0.5, True)
>>> model
StandardRNN(
  (embedding): HybridSequential(
    (0): Embedding(33280 -> 200, float32)
    (1): Dropout(p = 0.5, axes=())
  )
  (encoder): LSTM(200 -> 200.0, TNC, num_layers=2, dropout=0.5)
  (decoder): HybridSequential(
    (0): Dense(200 -> 33280, linear)
  )
)

For example, load a GloVe word embedding, one of the state-of-the-art English word embeddings:

>>> glove = nlp.embedding.create('glove', source='glove.6B.50d')
# Obtain vectors for 'baby' in the GloVe word embedding
>>> type(glove['baby'])
<class 'mxnet.ndarray.ndarray.NDArray'>
>>> glove['baby'].shape
(50,)

Reference Paper

The bibtex entry for the reference paper of GluonNLP is:

@article{gluoncvnlp2020,
  author  = {Jian Guo and He He and Tong He and Leonard Lausen and Mu Li and Haibin Lin and Xingjian Shi and Chenguang Wang and Junyuan Xie and Sheng Zha and Aston Zhang and Hang Zhang and Zhi Zhang and Zhongyue Zhang and Shuai Zheng and Yi Zhu},
  title   = {GluonCV and GluonNLP: Deep Learning in Computer Vision and Natural Language Processing},
  journal = {Journal of Machine Learning Research},
  year    = {2020},
  volume  = {21},
  number  = {23},
  pages   = {1-7},
  url     = {http://jmlr.org/papers/v21/19-429.html}
}

New to Deep Learning or NLP?

For background knowledge of deep learning or NLP, please refer to the open source book Dive into Deep Learning.

Packages

No packages published

Languages

  • Python 95.7%
  • Shell 1.8%
  • Groovy 1.2%
  • C++ 0.6%
  • Perl 0.4%
  • Makefile 0.2%
  • Other 0.1%