Allen NLP is based on PyTorch NLP research library is used to provide the best and most advanced in-depth learning model in the industry in various language tasks.
Allen NLP can make it easy to design and evaluate new deep learning models, and it is almost applicable to any NLP problem. By using some basic components, you can easily run the model in the cloud or on your own laptop.
Allen NLP design concept:
Ultra modular and lightweight. You can use your favorite components to seamlessly connect with PyTorch.
Extensive testing, easy to expand. The test coverage rate is higher than 90%, and the sample model provides excellent templates.
Real filling and covering, easy and painless implementation of the correct model.
Experimental friendliness. The experimental process can be reproduced through a comprehensive record that conforms to the json specification.
AllenNLP includes high-quality models such as Semantic Role Labelling, Question and Answering (BiDAF), Entry (attention splitting), and so on.
Allen NLP is mainly constructed and maintained by Allen Artificial Intelligence Laboratory and cooperates with Washington University and other institutions.
Demo Address
