WebML CG Teleconference – 14 February 2019 - 14:30-15:00 UTC+0

Hi All,

Below are the logistics and agenda for the monthly 30-minute call this Thursday. For links to the agenda details, please see the canonical agenda at:

https://github.com/webmachinelearning/meetings/blob/master/telcons/2019-02-14-agenda.md


Any agenda additions or changes, please let me know.


WebML CG Teleconference – 14 February 2019 - 14:30-15:00 UTC+0

Logistics

* Chair: Anssi
* Scribe: ?
* IRC : irc://irc.w3.org:6667/#webmachinelearning

* IRC web client: http://irc.w3.org/?channels=#webmachinelearning

* Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2018Nov/0000.html

  (login with your W3C account)
* Minutes: https://www.w3.org/2019/02/14-webmachinelearning-minutes.html


Agenda

#3 High level vs low level

* Revisit positions TL;DR:
  * @nsthorat: Our preference from TensorFlow is for this specification to focus on operations.
  * @RafaelCintron: WebML should include both a "loadModel-style" and "operator-style" APIs.
* Discuss implications of eager and graph execution mode mapping
* Discuss split into self-contained "what we can do now" and "what we eventually want to do" issues to keep us productive

#6 Custom operations

* Discuss and gauge support for including custom ops in scope for the initial deliverable
* Given support, decide how to proceed with investigation tasks e.g. browser implementation feasibility of built-in ops, survey ops needed to write custom ops


Thanks,

-Anssi (WebML CG chair)

Received on Monday, 11 February 2019 15:19:31 UTC