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.
- Tutorial proposal for GluonNLP is accepted at EMNLP 2019, Hong Kong.
- GluonNLP was featured in:
Make sure you have Python 2.7 or Python 3.6 and recent version of MXNet.
You can install
GluonNLP using pip.
GluonNLP is based on the most recent version of
In particular, if you want to install the most recent
pip install --upgrade mxnet>=1.4.0
Else, if you want to install the most recent
MXNet nightly build:
pip install --pre --upgrade mxnet
Then, you can install
pip install gluonnlp
Please check more installation details.
GluonNLP documentation is available at our website.
GluonNLP is a community that believes in sharing.
For questions, comments, and bug reports, Github issues is the best way to reach us.
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.
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,)
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.