Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Learnings from TensorFlow Dev Summit 2018

Learnings from TensorFlow Dev Summit 2018

I shared with the GDG Seattle community my Learnings from TensorFlow Dev Summit 2018.

Margaret Maynard-Reid

April 02, 2018
Tweet

More Decks by Margaret Maynard-Reid

Other Decks in Technology

Transcript

  1. @margaretmz | #TFDevSummit | #Extended | @GDGSeattle | @WTMSeattle |

    #WomenTechMakers Some of the announcements • TensorFlow 1.7 (released the day before the Summit) • TensorFlow.js • TensorFlow for SWIFT • TensorBoard debugger GUI plugin • An official TensorFlow blog on Medium: blog.tensorflow.org • An official TensorFlow YouTube Channel: youtube.com/tensorflow • TensorFlow Hub: a shared ML library with reusable ML code called modules.
  2. @margaretmz | #TFDevSummit | #Extended | @GDGSeattle | @WTMSeattle |

    #WomenTechMakers Eager Execution • No longer need to create a session and call session.run() • Iterate more quickly and debug easier • How to enable Eager Execution? ◦ Upgrade to TensorFlow 1.7 ◦ tf.enable_eager_execution() • Live demo with an interactive notebook goo.gl/eRpP8j • Documentation here • Watch the talk online here
  3. @margaretmz | #TFDevSummit | #Extended | @GDGSeattle | @WTMSeattle |

    #WomenTechMakers TensorFlow Hub TensorFlow Hub is a library to foster the publication, discovery, and consumption of reusable parts of machine learning model A module is • a saved model • easy to insatiate • retrainable • little training data is required in order to use a module
  4. @margaretmz | #TFDevSummit | #Extended | @GDGSeattle | @WTMSeattle |

    #WomenTechMakers TensorBoard Debugger GUI Plugin • TensorBoard debugger now has a GUI plugin! • Watch the video here for the demo https://youtu.be/XcHWLsVmHvk. • You just need one line to launch the plugin - $ tensorboard --logdir /tmp/logdir --debugger_port 7000
  5. @margaretmz | #TFDevSummit | #Extended | @GDGSeattle | @WTMSeattle |

    #WomenTechMakers TensorFlow Lite • A lightweight version of TensorFlow for on device ML • Small, fast and cross platform • Announced at I/O last year, • Developer preview released Nov of 2017 • Support Android, iOS and Raspberry Pi • Future roadmap includes on-device training
  6. @margaretmz | #TFDevSummit | #Extended | @GDGSeattle | @WTMSeattle |

    #WomenTechMakers References • TensorFlow Dev Summit 2018 Schedule • TensorFlow 1.70 release notes • Video recording to all talks at the TensorFlow Dev Summit • Official TensorFlow blog on Medium • Official TensorFlow YouTube Channel • My blog post here: bit.ly/learnings-TFDevSummit