Till startsida
University of Gothenburg
Sitemap
To content Read more about how we use cookies on gu.se

Jean-Philippe Bernardy " TypedFlow: A library for typed and higher-order deep learning"

Research profile seminar

Jean-Philippe Bernardy " TypedFlow: A library for typed and higher-order deep learning"

TensorFlow is a library for numerical computation, with specific
features for machine-learning such as gradient computation. It is
perhaps the most popular backend for deep learning applications.
TypedFlow a higher-order and typed (HOT) frontend to TensorFlow
written in Haskell, and a library of neural-network layers and
combinators.


In this talk I will:

- briefly recall what TensorFlow is and how it applies to deep
learning
- discuss the advantages of a HOT approach vs. plain TensorFlow
- expose two use-cases: the standard MNIST example and a
sequence-to-sequence network with attention model.

Lecturer: Dr Jean-Philippe Bernardy

Date: 11/1/2017

Time: 1:15 PM - 3:00 PM

Categories: Linguistics

Organizer: CLASP

Location: Dicksonsgatan 4, seminar room

Contact person: stergios chatzikyriakidis

Page Manager: Webbredaktionen|Last update: 5/23/2016
Share:

The University of Gothenburg uses cookies to provide you with the best possible user experience. By continuing on this website, you approve of our use of cookies.  What are cookies?