Skip to content

This is a working demo of the pegasus summarization model trained on cnn_dailymail

License

Notifications You must be signed in to change notification settings

TheRockXu/pegasus-demo

 
 

Repository files navigation

This is a trained model of PEGASUS on cnn_dailymail dataset.

Requirements -

numpy
sentencepiece
tensorflow==2.2.0

PEGASUS library

Pre-training with Extracted Gap-sentences for Abstractive SUmmarization Sequence-to-sequence models, or PEGASUS, uses self-supervised objective Gap Sentences Generation (GSG) to train a transformer encoder-decoder model. The paper can be found on arXiv. ICML 2020 accepted.

'screen'

If you use this code or these models, please cite the following paper:

@misc{zhang2019pegasus,
    title={PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization},
    author={Jingqing Zhang and Yao Zhao and Mohammad Saleh and Peter J. Liu},
    year={2019},
    eprint={1912.08777},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

Usage

To run the demo, please download pre-trained model on cnn_dailymail from here or gigaword from here. Unzip it and put it to model/, or anywhere really if you just specify its location and where your article file is. Suppose your article is this one

python test_example.py --article example_article --model_dir model/ --model_name cnn_dailymail

You will see this output - PREDICTION >> The hacking group known as NC29 is largely believed to operate as part of Russia's security services .<n>The three countries allege that it is carrying out a persistent and ongoing cyber campaign to steal intellectual property about a possible coronavirus vaccine .

Export Model

To export a model you have trained, please place the ExportModel.ipynb inside the PEGASUS folder. Just run the script inside by specifying which data model you want to export.

About

This is a working demo of the pegasus summarization model trained on cnn_dailymail

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 57.2%
  • Jupyter Notebook 42.8%