WebML WG Teleconference – 25 May 2023 - 14:00-15:00 UTC

Latest version: https://github.com/webmachinelearning/meetings/blob/main/telcons/2023-05-25-wg-agenda.md


WebML WG Teleconference – 25 May 2023 - 14:00-15:00 UTC
See the timezone table ...
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2023-05-25-wg-agenda.md#logistics>Logistics

  *   Chair: Anssi
  *   Scribe: ? (see howto<https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md>, volunteers welcome!)
  *   IRC: irc://irc.w3.org:6667/#webmachinelearning

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

  *   Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html

  *   Minutes: https://www.w3.org/2023/05/25-webmachinelearning-minutes.html


<https://github.com/webmachinelearning/meetings/blob/main/telcons/2023-05-25-wg-agenda.md#agenda>Agenda
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2023-05-25-wg-agenda.md#%E2%84%B9%EF%B8%8F-webidl-and-infra-standard-conventions>ℹ️ WebIDL and Infra standard conventions

Review and land the open PRs to prep for an upcoming stylistic update:

  *   The computeSync and computeAsync algorithms

     *    issue: webmachinelearning/webnn#316<https://github.com/webmachinelearning/webnn/issues/316>
     *    PR: webmachinelearning/webnn#329<https://github.com/webmachinelearning/webnn/pull/329>
  *   The batchNormalization algorithm

     *    issue: webmachinelearning/webnn#334<https://github.com/webmachinelearning/webnn/issues/334>
     *    PR: webmachinelearning/webnn#339<https://github.com/webmachinelearning/webnn/pull/339>
  *   The clamp algorithm

     *    issue: webmachinelearning/webnn#347<https://github.com/webmachinelearning/webnn/issues/347>
     *    PR: webmachinelearning/webnn#348<https://github.com/webmachinelearning/webnn/pull/348>
  *   The concat algorithm

     *    meta issue: webmachinelearning/webnn#210 (comment)<https://github.com/webmachinelearning/webnn/issues/210#issuecomment-1326361748>
     *    PR: webmachinelearning/webnn#366<https://github.com/webmachinelearning/webnn/pull/366>

<https://github.com/webmachinelearning/meetings/blob/main/telcons/2023-05-25-wg-agenda.md#%E2%84%B9%EF%B8%8F-new-features>ℹ️ New features

  *   Float16Array (design finished, ready to integrate into WebNN)

     *    issue: webmachinelearning/webnn#373<https://github.com/webmachinelearning/webnn/issues/373>
     *    PR: webmachinelearning/webnn#386<https://github.com/webmachinelearning/webnn/pull/386>
  *   0D scalars

     *    issue: webmachinelearning/webnn#390<https://github.com/webmachinelearning/webnn/issues/390>

<https://github.com/webmachinelearning/meetings/blob/main/telcons/2023-05-25-wg-agenda.md#%E2%84%B9%EF%B8%8F-enhancements-editorials-questions>ℹ️ Enhancements, editorials, questions

  *   Error handling of MLNamedArrayBufferViews transfer algorithm

     *    issue: webmachinelearning/webnn#351<https://github.com/webmachinelearning/webnn/issues/351>
  *   Remove the support of null value in newShape parameter of reshape build method

     *    issue: webmachinelearning/webnn#388<https://github.com/webmachinelearning/webnn/issues/388>

<https://github.com/webmachinelearning/meetings/blob/main/telcons/2023-05-25-wg-agenda.md#%E2%84%B9%EF%B8%8F-support-for-transformers>ℹ️ Support for transformers

Continue discuss transformers and related requirements and gaps. Test our improved contribution guidelines with these new ops: explore key use cases, sample models, cross-framework support, cross-platform implementability.

Status: The WG decided to start explore support for transformers in WebNN.

Next step: Identify use cases and what new ops are needed in WebNN to support key models. Proposed investigation paths Stable Diffusion, SegmentAnything, Transformers.js/HuggingFace's transformers. Contributions and pointers welcome via issue 375.

  *    issue: webmachinelearning/webnn#375<https://github.com/webmachinelearning/webnn/issues/375>