NLP Architect is an open-source Python library for exploring the state-of-the-art deep learning topologies and techniques for natural language processing and natural language understanding. It is intended to be a space to promote research and collaboration.
The library consists of core modules (topologies), data pipelines, utilities and end-to-end model examples with training and inference scripts. Each of the models includes algorithm descriptions and results in the documentation.
Because of its current research nature, several open source deep learning frameworks are used in this repository including:
Overtime the list of models included in this space will change, though all generally run with Python 3.5+
Framework documentation on NLP model, algorithms, and modules, and instructions on how to contribute can be found here.
We recommend installing NLP Architect within a virtual environment to ensure a self-contained environment. To install NLP Architect models within an already existing virtual environment, see below installation receipes for custom model installation. The default installation will create a new local virtual environment for development purposes.
To get started using our library, clone our repository:
git clone https://github.com/NervanaSystems/nlp-architect.git
cd nlp-architect
make
Complete install:
make install
Activate the newly created virtual environment:
. .nlp_architect_env/bin/activate
Fire up your favorite IDE/text editor/terminal and start running models
Install without creating a new virtual environment:
make install_no_virt_env
The NLP Architect is released as a reference code for research purposes. It is not an official Intel product, and the level of quality and support may not be as expected from an official product. Additional algorithms and environments are planned to be added to the framework. Feedback and contributions from the open source and NLP research communities are more than welcome.
Contacting the NLP Architect development team through Github issues or email: [email protected]