/NLP_exctractive_and_abstractive_simple_implementation

Abstractive summarizers GPT2 vs XLNet vs BERT model

Primary LanguageJupyter NotebookMIT LicenseMIT

GPT2 vs XLNet vs BERT

Hi, when it comes to summarization models, there are different recommendations, some prefer BERT, some GPT2 and GPT3 while few prefer XLNet This colab notebook is simple implementation of all these 3 models so that you can easily test out, have a POC and then move ahead with whichever suits you fine. Kindly maintain the versions of library mentioned in the notebook else you might see version incompatibility issue with spacy thinc

I will also be publishing Pegasus model simple implementation(huggingface transformers) so that this can be used to your first step in detailed exploration of your NLP tasks