- From: Kostiainen, Anssi <anssi.kostiainen@intel.com>
- Date: Thu, 24 Oct 2024 15:35:19 +0000
- To: "public-webmachinelearning-wg@w3.org" <public-webmachinelearning-wg@w3.org>
- Message-ID: <CCD24D58-6F44-438C-9A2E-410E1C6CA467@intel.com>
Latest version: https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md
WebML WG Teleconference – 31 October 2024 - 14:00-15:00 UTC
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#webml-wg-teleconference--31-october-2024---1400-1500-utc>
See the timezone table ...
San Francisco Thu, 31 October 2024 07:00 UTC-7 hours
Boston Thu, 31 October 2024 10:00 UTC-4 hours
London Thu, 31 October 2024 14:00 UTC+0 hours
Berlin Thu, 31 October 2024 15:00 UTC+1 hours
Helsinki Thu, 31 October 2024 16:00 UTC+2 hours
Shanghai Thu, 31 October 2024 22:00 UTC+8 hours
Tokyo Thu, 31 October 2024 23:00 UTC+9 hours
UTC Thu, 31 October 2024 14:00 UTC
Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20241031T14
Note
In most European locations our meeting starts 1 hour earlier due to the Daylight Saving Time change<https://www.timeanddate.com/time/dst/events.html>.
Logistics
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#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/2023Jun/0000.html
* Minutes: https://www.w3.org/2024/10/31-webmachinelearning-minutes.html
Agenda
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#agenda>
ℹ️ WebML Community Group Charter update
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#%E2%84%B9%EF%B8%8F-webml-community-group-charter-update>
Review and discuss the proposed WebML Community Group charter update to add new task-specific APIs (Writing Assistance APIs<https://github.com/WICG/writing-assistance-apis>, Translator API<https://github.com/WICG/translation-api>, Prompt API<https://github.com/explainers-by-googlers/prompt-api/>) introduced at TPAC 2024 in scope. Note: The scope of work for the proposed Community Group extends beyond the current scope of the Web Machine Learning Working Group.
* ⛙ webmachinelearning/charter#9<https://github.com/webmachinelearning/charter/pull/9>
* See also: 📁 TPAC 2024 slides<https://lists.w3.org/Archives/Public/www-archive/2024Sep/att-0008/TPAC_2024_Built-in_AI_APIs.pdf> for task-specific APIs
ℹ️ Device selection abstractions
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#%E2%84%B9%EF%B8%8F-device-selection-abstractions>
Discuss the generalized device selection proposal (MLOpSupportLimits) and related security considerations.
* ⨀ webmachinelearning/webnn#749<https://github.com/webmachinelearning/webnn/issues/749>
ℹ️ MLTensor
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#%E2%84%B9%EF%B8%8F-mltensor>
Discuss and address the latest review feedback for the MLTensor explainer.
* ⛙ webmachinelearning/webnn#754<https://github.com/webmachinelearning/webnn/pull/754>
ℹ️ Tensor primitives
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#%E2%84%B9%EF%B8%8F-tensor-primitives>
Discuss requirements as appropriate, possible subtopics:
* Additional primitive ops: evolve core op set w/ MLIR linalg<https://mlir.llvm.org/docs/Dialects/Linalg/>, PT Edge<https://pytorch.org/executorch/stable/ir-exir.html#edge-operators>, TOSA
* Graph expressiveness: subgraphs, control flow
* Native runtime support: fusion, pattern matcher
* See also: 📁 Slides<https://lists.w3.org/Archives/Public/www-archive/2024Sep/att-0007/Tensor_Primitive_Ops_Proposal_-_TPAC.pdf>
ℹ️ Open issues and PRs
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#%E2%84%B9%EF%B8%8F-open-issues-and-prs>
We'll discuss new issues, debrief the group on merged PRs and review open PRs since our last meeting:
* ⨀ All open issues<https://github.com/webmachinelearning/webnn/issues>
* ⛙ All open pull requests<https://github.com/webmachinelearning/webnn/pulls>
* ⛙ Recently merged PRs<https://github.com/webmachinelearning/webnn/pulls?q=is%3Apr+is%3Amerged>
🏷️ feature request<https://github.com/webmachinelearning/webnn/labels/feature%20request>
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#%EF%B8%8F-feature-request>
* LocalResponseNormalization
* ⨀ webmachinelearning/webnn#228<https://github.com/webmachinelearning/webnn/issues/228>
* ☰ Decomposition or a new operator?
🏷️ operator specific<https://github.com/webmachinelearning/webnn/labels/operator%20specific>
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#%EF%B8%8F-operator-specific>
* Remove or redesign triangular
* ⨀ webmachinelearning/webnn#768<https://github.com/webmachinelearning/webnn/issues/768>
* ☰ All backends emulate triangular, discuss specific mask behaviour.
* Consider adding int64/uint64 data type support for some reduce operators
* ⨀ webmachinelearning/webnn#694<https://github.com/webmachinelearning/webnn/issues/694>
* ⛙ webmachinelearning/webnn#695<https://github.com/webmachinelearning/webnn/pull/695>
* ☰ Clarify PR status.
* Decomposition for gatherElements, scatterElements and gatherND
* ⨀ webmachinelearning/webnn#767<https://github.com/webmachinelearning/webnn/issues/767>
* ☰ Identify and document decompositions.
🏷️ interop<https://github.com/webmachinelearning/webnn/labels/interop>
<https://github.com/webmachinelearning/meetings/blob/main/telcons/2024-10-31-wg-agenda.md#%EF%B8%8F-interop>
* Remove pool2d MLRoundingType
* ⨀ webmachinelearning/webnn#324<https://github.com/webmachinelearning/webnn/issues/324>
* ⛙ webmachinelearning/webnn#770<https://github.com/webmachinelearning/webnn/pull/770>
* ☰ Review PR.
* Rank range support
* ⨀ webmachinelearning/webnn#456<https://github.com/webmachinelearning/webnn/issues/456>
* ⨀ related: webmachinelearning/webnn#463<https://github.com/webmachinelearning/webnn/issues/463>
* ☰ Address remaining work from opSupportLimits(), others?
Received on Thursday, 24 October 2024 15:35:36 UTC