Note: This video may require joining the NVIDIA Developer Program or login

GTC Silicon Valley-2019 ID:S9628:Summarize Large Text using NLP on NVIDIA P3 Instances

Kristof Schum(AWS)
We'll describe different approaches to building a summarizer, which creates smaller versions of large documents while retaining the core message. A summarizer can automatically create snippets for the landing pages of web applications or for search engines, and it may reduce the time to read a document by an order of magnitude. We'll discuss the differences between the two approaches to creating a summarizer the statistical approach and the rephrase approach. We'll also examine the algorithm and show a notebook to pre-process the document, set up the machine learning model, and finally deploy this model for real-time invocation.

View the slides (pdf)