~/talk

Tensorflow for Janitors

on YouTube 31m
Talkat code.talks 2017 · 2017-09-28..09-29· [machine-learning]

speaker

name:Dániel Molnár
links:none on file

resources

video:https://youtu.be/OSkL7AXu6tY
slides:none on file

summary

Claims that TensorFlow is an end-to-end, production-ready stack with Python APIs, Keras bindings, TensorBoard and TF Serving and deploy targets from CPU/GPU to Raspberry Pi; demonstrates a 500 kilobyte model on a phone running ~1 fps and mentions federated learning on Android; cites a chip that sees a cat in 41 milliseconds (4× faster than a MacBook Air, 7× cheaper, 6× lower power); argues tooling, mobile-first models and community adoption make it practical.

AI-written summary · about

chapters (10)

  1. 0:00Welcome
  2. 2:12Deep learning basics
  3. 7:19Why deep learning now
  4. 13:24Neural network components
  5. 16:56TensorFlow vs others
  6. 19:47Why TensorFlow
  7. 23:17Future of distributed DL
  8. 25:48Should you try DL
  9. 27:39DL market reality
  10. 28:32Q&A

related talks

~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.