├── 2018-10-26-lyon
    └── README.md
├── 2019-09-17-fukuoka
    └── README.md
├── 2020-10-22-tpac2020
    └── README.md
├── CONTRIBUTING.md
├── LICENSE.md
├── README.md
├── scribe-howto.md
└── telcons
    ├── 2018-12-13-agenda.md
    ├── 2019-01-10-agenda.md
    ├── 2019-02-14-agenda.md
    ├── 2019-03-14-agenda.md
    ├── 2019-04-11-agenda.md
    ├── 2019-04-25-agenda.md
    ├── 2019-05-09-agenda.md
    ├── 2019-06-06-agenda.md
    ├── 2019-06-27-agenda.md
    ├── 2019-08-08-agenda.md
    ├── 2019-08-22-agenda.md
    ├── 2019-09-05-agenda.md
    ├── 2019-10-03-agenda.md
    ├── 2019-10-17-agenda.md
    ├── 2019-10-31-agenda.md
    ├── 2019-11-14-agenda.md
    ├── 2019-11-28-agenda.md
    ├── 2019-12-05-agenda.md
    ├── 2019-12-19-agenda.md
    ├── 2020-01-09-agenda.md
    ├── 2020-01-23-agenda.md
    ├── 2020-02-06-agenda.md
    ├── 2020-02-20-agenda.md
    ├── 2020-03-05-agenda.md
    ├── 2020-03-19-agenda.md
    ├── 2020-04-02-agenda.md
    ├── 2020-04-16-agenda.md
    ├── 2020-04-30-agenda.md
    ├── 2020-05-14-agenda.md
    ├── 2020-05-14-wasi-nn.pdf
    ├── 2020-05-28-agenda.md
    ├── 2020-06-11-agenda.md
    ├── 2020-06-25-agenda.md
    ├── 2020-08-20-agenda.md
    ├── 2020-09-03-agenda.md
    ├── 2020-09-17-agenda.md
    ├── 2020-10-01-agenda.md
    ├── 2020-10-29-agenda.md
    ├── 2020-11-12-agenda.md
    ├── 2020-11-26-agenda.md
    ├── 2020-12-10-agenda.md
    ├── 2021-01-07-agenda.md
    ├── 2021-01-21-agenda.md
    ├── 2021-02-04-agenda.md
    ├── 2021-02-18-agenda.md
    ├── 2021-03-04-agenda.md
    ├── 2021-03-18-agenda.md
    ├── 2021-04-01-agenda.md
    ├── 2021-04-15-agenda.md
    ├── 2021-04-29-agenda.md
    ├── 2021-05-13-agenda.md
    ├── 2021-05-27-agenda.md
    ├── 2021-06-10-agenda.md
    ├── 2021-06-24-agenda.md
    ├── 2021-08-19-agenda.md
    ├── 2021-09-02-agenda.md
    ├── 2021-09-16-agenda.md
    ├── 2021-09-30-agenda.md
    ├── 2021-10-14-agenda.md
    ├── 2021-11-11-agenda.md
    ├── 2021-11-18-agenda.md
    ├── 2021-12-02-agenda.md
    ├── 2021-12-16-agenda.md
    ├── 2022-01-12-cg-agenda.md
    ├── 2022-01-13-wg-agenda.md
    ├── 2022-01-26-cg-agenda.md
    ├── 2022-01-27-wg-agenda.md
    ├── 2022-02-09-cg-agenda.md
    ├── 2022-02-10-wg-agenda.md
    ├── 2022-02-24-wg-agenda.md
    ├── 2022-03-09-cg-agenda.md
    ├── 2022-03-10-wg-agenda.md
    ├── 2022-03-24-wg-agenda.md
    ├── 2022-04-06-cg-agenda.md
    ├── 2022-04-07-wg-agenda.md
    ├── 2022-04-21-wg-agenda.md
    ├── 2022-05-04-cg-agenda.md
    ├── 2022-05-05-wg-agenda.md
    ├── 2022-05-19-wg-agenda.md
    ├── 2022-06-02-wg-agenda.md
    ├── 2022-06-16-wg-agenda.md
    ├── 2022-06-30-wg-agenda.md
    ├── 2022-08-11-wg-agenda.md
    ├── 2022-08-25-wg-agenda.md
    ├── 2022-09-08-wg-agenda.md
    ├── 2022-09-22-wg-agenda.md
    ├── 2022-10-06-wg-agenda.md
    ├── 2022-10-20-wg-agenda.md
    ├── 2022-11-03-wg-agenda.md
    ├── 2022-11-17-wg-agenda.md
    ├── 2022-12-01-wg-agenda.md
    ├── 2022-12-15-wg-agenda.md
    ├── 2023-01-12-wg-agenda.md
    ├── 2023-02-02-wg-agenda.md
    ├── 2023-02-16-wg-agenda.md
    ├── 2023-03-02-wg-agenda.md
    ├── 2023-03-16-wg-agenda.md
    ├── 2023-03-30-wg-agenda.md
    ├── 2023-04-13-wg-agenda.md
    ├── 2023-04-27-wg-agenda.md
    ├── 2023-05-11-wg-agenda.md
    ├── 2023-05-25-wg-agenda.md
    ├── 2023-06-08-wg-agenda.md
    ├── 2023-06-22-wg-agenda.md
    ├── 2023-06-29-wg-agenda.md
    ├── 2023-08-10-wg-agenda.md
    ├── 2023-08-24-wg-agenda.md
    ├── 2023-09-07-wg-agenda.md
    ├── 2023-09-21-wg-agenda.md
    ├── 2023-10-05-wg-agenda.md
    ├── 2023-10-19-wg-agenda.md
    ├── 2023-11-02-wg-agenda.md
    ├── 2023-11-16-wg-agenda.md
    ├── 2023-11-30-wg-agenda.md
    ├── 2023-12-14-wg-agenda.md
    ├── 2024-01-11-wg-agenda.md
    ├── 2024-01-25-wg-agenda.md
    ├── 2024-02-08-wg-agenda.md
    ├── 2024-02-22-wg-agenda.md
    ├── 2024-03-07-wg-agenda.md
    ├── 2024-03-21-wg-agenda.md
    ├── 2024-04-04-wg-agenda.md
    ├── 2024-04-18-wg-agenda.md
    ├── 2024-05-02-wg-agenda.md
    ├── 2024-05-16-wg-agenda.md
    ├── 2024-05-30-wg-agenda.md
    ├── 2024-06-13-wg-agenda.md
    ├── 2024-06-27-wg-agenda.md
    ├── 2024-08-08-wg-agenda.md
    ├── 2024-08-22-wg-agenda.md
    ├── 2024-09-05-wg-agenda.md
    ├── 2024-10-17-wg-agenda.md
    ├── 2024-10-31-wg-agenda.md
    ├── 2024-11-14-wg-agenda.md
    ├── 2024-12-05-wg-agenda.md
    ├── 2024-12-19-wg-agenda.md
    ├── 2025-01-16-wg-agenda.md
    ├── 2025-01-30-wg-agenda.md
    ├── 2025-02-13-wg-agenda.md
    ├── 2025-02-26-cg-agenda.md
    ├── 2025-02-26-cg-minutes.md
    ├── 2025-02-27-wg-agenda.md
    ├── 2025-03-12-cg-agenda.md
    ├── 2025-03-12-cg-minutes.md
    ├── 2025-03-13-wg-agenda.md
    ├── 2025-03-27-wg-agenda.md
    ├── 2025-03-31-cg-agenda.md
    ├── 2025-03-31-cg-minutes.md
    ├── 2025-04-10-wg-agenda.md
    ├── 2025-04-24-wg-agenda.md
    ├── 2025-04-28-cg-agenda.md
    ├── 2025-04-28-cg-minutes.md
    ├── 2025-05-08-wg-agenda.md
    ├── 2025-05-22-wg-agenda.md
    ├── 2025-05-26-cg-agenda.md
    ├── 2025-05-26-cg-minutes.md
    └── 2025-06-05-wg-agenda.md


/2018-10-26-lyon/README.md:
--------------------------------------------------------------------------------
 1 | # Machine Learning for the Web Community Group F2F at TPAC 2018
 2 | 
 3 | ![WebML logo][logo]
 4 | 
 5 | * Date/time: 26 Oct 2018 15:30-17:30 (CEST)
 6 | * Location: Cité Centre de Congrès de Lyon https://www.w3.org/2018/10/TPAC/#venue
 7 | * Meeting Room: Saint Clair 3B, Level 2 https://www.w3.org/2018/10/TPAC/schedule.html#map
 8 | 
 9 | ## Logistics
10 | 
11 | * Chair: Anssi Kostiainen
12 | * Scribe: Myles Maxfield
13 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
14 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
15 | * Remote participation: join IRC and ask
16 | 
17 | ## Meeting minutes
18 | 
19 | * https://www.w3.org/2018/10/26-webmachinelearning-minutes.html
20 | 
21 | ## Agenda
22 | 
23 | >To propose an agenda topic, please open a new issue:
24 | >https://github.com/webmachinelearning/meetings/issues/new
25 | 
26 | #### 15:30 Welcome, introductions & announcements
27 | 
28 | * Welcome (Chair)
29 | 
30 | * Roundtable 30-sec intros: your affiliation & interests toward the group, how you’d like to contribute (All)
31 | 
32 | * Announcement: W3C Graph Data Workshop Feb 2019 (Dave)
33 | https://www.w3.org/Data/events/data-ws-2019/
34 | 
35 | #### 15:45 Community Group intro
36 | 
37 | * Getting started in the group: goals, scope, work mode (Chair)
38 | 
39 | * CG home: https://www.w3.org/community/webmachinelearning/
40 | * Charter: https://webmachinelearning.github.io/charter/
41 | * Landing page: https://webmachinelearning.github.io/
42 | 
43 | #### 16:00 Target demographic, use cases and requirements (All)
44 | 
45 | * Discuss target demographic (JS lib and fw authors) and its needs
46 | 
47 | * Solicit input on use cases, low level vs. high level API https://github.com/webmachinelearning/meetings/issues/1
48 | 
49 | * Does our current scope capture the needs of the target audience?
50 | 
51 | #### 16:15 Review mapping to platform APIs (@huningxin)
52 | 
53 | * Review ops mapping to platform APIs that may inform the Web API design:
54 | https://github.com/intel/webml-polyfill/blob/master/docs/native_mapping.md
55 | 
56 | #### 16:30 Review & discuss Web Neural Network API spec proposals (All)
57 | 
58 | * Review and discuss API spec proposals submitted as input that could inform the spec work in this group:
59 | 
60 |   * WebML API proof-of-concept by @huningxin https://github.com/intel/webml-polyfill/blob/master/docs/api.md
61 |   * Any other proposals
62 | 
63 | >To submit a proposal, please open a new issue:
64 | >https://github.com/webmachinelearning/meetings/issues/new
65 | >
66 | >The API proposals do not need to be fleshed out or complete. Any format is accepted, explainers and position papers are equally welcome.
67 | 
68 | * Gauge community's interest to work on an open-source cross-platform reference design
69 | 
70 | #### 17:30 Follow-up plan, end meeting
71 | 
72 | [logo]: https://avatars3.githubusercontent.com/u/42399997?s=100 "WebML Logo, CC0 Creative Commons"
73 | 


--------------------------------------------------------------------------------
/2020-10-22-tpac2020/README.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Virtual Meeting at TPAC 2020 – 22 October 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | ![WebML logo][logo]
 4 | 
 5 | <table>
 6 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 22 October 2020 <td> 07:00 PDT <td> UTC-7 hours
 7 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 22 October 2020 <td> 10:00 EDT <td> UTC-4 hours
 8 | <tr><td> London (United Kingdom - England) <td> Thu, 22 October 2020 <td> 15:00 BST <td> UTC+1 hours
 9 | <tr><td> Berlin (Germany) <td> Thu, 22 October 2020 <td> 16:00 CEST <td> UTC+2 hours
10 | <tr><td> Helsinki (Finland) <td> Thu, 22 October 2020 <td> 17:00 EEST <td> UTC+3 hours
11 | <tr><td> Shanghai (China) <td> Thu, 22 October 2020 <td> 22:00 CST <td> UTC+8 hours
12 | <tr><td> Tokyo (Japan) <td> Thu, 22 October 2020 <td> 23:00 JST <td> UTC+9 hours
13 | <tr><td> Corresponding UTC (GMT) <td> Thu, 22 October 2020 <td colspan=2> 14:00 UTC
14 | </table>
15 | 
16 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20201022T14
17 | 
18 | ### Registration
19 | 
20 | Please register via https://www.w3.org/2002/09/wbs/35125/TPAC2020/
21 | 
22 | ### Logistics
23 | 
24 | * Chair: Anssi
25 | * Scribe: ?
26 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
27 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
28 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended
29 | * Minutes: https://www.w3.org/2020/10/22-webmachinelearning-minutes.html
30 | 
31 | ### Agenda
32 | 
33 | #### Workshop proceedings
34 | 
35 | Discuss the [W3C Workshop on Web and Machine Learning](https://www.w3.org/2020/06/machine-learning-workshop/) key outcomes with a focus on proposed near-term and long-term standardization directions and next steps. See the proposed [Web Machine Learning Working Group Charter](https://w3c.github.io/machine-learning-charter/charter.html) ([GH repo](https://github.com/w3c/machine-learning-charter)).
36 | 
37 | References:
38 | 
39 | - [Workshop report](https://www.w3.org/2020/06/machine-learning-workshop/report.html)
40 | - [Workshop presentations](https://www.w3.org/2020/06/machine-learning-workshop/presentations.html)
41 | - [Workshop discussions on GitHub](https://github.com/w3c/machine-learning-workshop/issues)
42 | - Live Sessions:
43 |   - Intro & Opportunities and Challenges (2020-09-16): [agenda](https://github.com/w3c/machine-learning-workshop/tree/master/agenda#september-16-2020-2pm-utc-%EF%B8%8F), [minutes](https://www.w3.org/2020/06/machine-learning-workshop/minutes/20200916.html)
44 |   - Web Platform Foundations (2020-09-22): [agenda](https://github.com/w3c/machine-learning-workshop/tree/master/agenda#september-22-2020-2pm-utc-%EF%B8%8F), [minutes](https://www.w3.org/2020/06/machine-learning-workshop/minutes/20200922.html)
45 |   - Developer's Perspective (2020-09-23): [agenda](https://github.com/w3c/machine-learning-workshop/tree/master/agenda#september-23-2020-2pm-utc-%EF%B8%8F), [minutes](https://www.w3.org/2020/06/machine-learning-workshop/minutes/20200923.html)
46 |   - User's Perspective & Next Steps (2020-09-29): [agenda](https://github.com/w3c/machine-learning-workshop/tree/master/agenda#september-29-2020-2pm-utc-%EF%B8%8F), [minutes](https://www.w3.org/2020/06/machine-learning-workshop/minutes/20200929.html)
47 | 
48 | [logo]: https://avatars3.githubusercontent.com/u/42399997?s=100 "WebML Logo, CC0 Creative Commons"
49 | 


--------------------------------------------------------------------------------
/CONTRIBUTING.md:
--------------------------------------------------------------------------------
 1 | # W3C Machine Learning for the Web Community Group
 2 | 
 3 | This repository is being used for work in the W3C Machine Learning for the Web Community Group, governed by the [W3C Community License Agreement (CLA)](https://www.w3.org/community/about/agreements/cla/). To make substantive contributions, you must join the CG. 
 4 | 
 5 | If you are not the sole contributor to a contribution (pull request), please identify all 
 6 | contributors in the pull request comment.
 7 | 
 8 | To add a contributor (other than yourself, that's automatic), mark them one per line as follows:
 9 | 
10 | ```
11 | +@github_username
12 | ```
13 | 
14 | If you added a contributor by mistake, you can remove them in a comment with:
15 | 
16 | ```
17 | -@github_username
18 | ```
19 | 
20 | If you are making a pull request on behalf of someone else but you had no part in designing the 
21 | feature, you can remove yourself with the above syntax.


--------------------------------------------------------------------------------
/LICENSE.md:
--------------------------------------------------------------------------------
 1 | All Reports in this Repository are licensed by Contributors
 2 | under the 
 3 | [W3C Software and Document License](https://www.w3.org/Consortium/Legal/2015/copyright-software-and-document).  
 4 | 
 5 | Contributions to Specifications are made under the
 6 | [W3C CLA](https://www.w3.org/community/about/agreements/cla/).
 7 | 
 8 | Contributions to Test Suites are made under the
 9 | [W3C 3-clause BSD License](https://www.w3.org/Consortium/Legal/2008/03-bsd-license.html)
10 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | # W3C Web Machine Learning WG and CG Meetings
 2 | 
 3 | This repository contains meeting agendas and links to meeting minutes:
 4 | 
 5 | * [Teleconference agendas](https://github.com/webmachinelearning/meetings/tree/main/telcons)
 6 | * For meeting minutes, see `Minutes:` link in the agendas
 7 | 
 8 | Please see the Meeting Calendars for the upcoming and past meetings:
 9 | 
10 | * [WG Meeting Calendar](https://www.w3.org/groups/wg/webmachinelearning/calendar)
11 | * [CG Meeting Calendar](https://www.w3.org/groups/cg/webmachinelearning/calendar)
12 | 
13 | For instructions on how to get invites to these meetings, please see [Join the Community](https://webmachinelearning.github.io/community/#join) for instructions.
14 | 


--------------------------------------------------------------------------------
/telcons/2018-12-13-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 13 December 2018 - 14:30-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 13 Dec 2018 <td> 06:30 PST <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 13 Dec 2018 <td> 09:30 EST <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 13 Dec 2018 <td> 14:30 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 13 Dec 2018 <td> 15:30 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 13 Dec 2018 <td> 16:30 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 13 Dec 2018 <td> 22:30 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 13 Dec 2018 <td> 23:30 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 13 Dec 2018 <td colspan=2> 14:30 UTC
12 | </table>
13 | 
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC : irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2018Nov/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2018/12/13-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * High level vs low level discussion https://github.com/webmachinelearning/webnn/issues/3
26 | * Use cases review https://github.com/webmachinelearning/webnn/pull/2
27 | * WebNN POC update https://docs.google.com/presentation/d/1ZmBg264_DIpMchq7tLhj6fMi19dcUv8I_kB8viwfxYM/edit?usp=sharing
28 | 


--------------------------------------------------------------------------------
/telcons/2019-01-10-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 10 January 2019 - 14:30-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 10 January 2019 <td> 06:30 PST <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 10 January 2019 <td> 09:30 EST <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 10 January 2019 <td> 14:30 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 10 January 2019 <td> 15:30 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 10 January 2019 <td> 16:30 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 10 January 2019 <td> 22:30 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 10 January 2019 <td> 23:30 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 10 January 2019 <td colspan=2> 14:30 UTC
12 | </table>
13 | 
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC : irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2018Nov/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/01/10-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * High level vs low level follow-up https://github.com/webmachinelearning/webnn/issues/3
26 |   * Review existing proposals (thanks!), see if we can agree on the next steps. Proposals:
27 |     * a) Incubate both the operator and load model approaches in parallel
28 |     * b) Focus on the operator approach and develop the load model approach as a polyfill initially
29 |     * c) New HTML element similar to `<video>`
30 |     * d) Others?
31 |     
32 | * Web Neural Network API https://webmachinelearning.github.io/webnn/
33 |   * Use cases landed, next: derive requirements to measure API against
34 |   * Semantic segmentation use case investigation: example, requirements, native support and performance ([slides](https://docs.google.com/presentation/d/1FYiirLCQgAZbGLxm9zHDuRSsSFZpvUmwUVYQqdOpQBM/edit?usp=sharing))
35 |   * Call for editors
36 | 


--------------------------------------------------------------------------------
/telcons/2019-02-14-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 14 February 2019 - 14:30-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 14 February 2019 <td> 06:30 PST <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 14 February 2019 <td> 09:30 EST <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 14 February 2019 <td> 14:30 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 14 February 2019 <td> 15:30 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 14 February 2019 <td> 16:30 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 14 February 2019 <td> 22:30 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 14 February 2019 <td> 23:30 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 14 February 2019 <td colspan=2> 14:30 UTC
12 | </table>
13 | 
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC : irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2018Nov/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/02/14-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | #### [#3 High level vs low level](https://github.com/webmachinelearning/webnn/issues/3)
26 | 
27 | * Revisit positions TL;DR:
28 |    * > [@nsthorat](https://github.com/webmachinelearning/webnn/issues/3#issuecomment-453129272): Our preference from TensorFlow is for this specification to focus on operations.
29 |    * > [@RafaelCintron](https://github.com/webmachinelearning/webnn/issues/3#issuecomment-457886464): WebML should include both a "loadModel-style" and "operator-style" APIs.
30 | * Discuss implications of [eager and graph execution mode mapping](https://github.com/webmachinelearning/webnn/issues/3#issuecomment-459755470)
31 | * Discuss split into self-contained "what we can do now" and "what we eventually want to do" issues to keep us productive
32 | 
33 | #### [#6 Custom operations](https://github.com/webmachinelearning/webnn/issues/6)
34 | 
35 | * Discuss and gauge support for including custom ops in scope for the initial deliverable
36 | * 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
37 | 


--------------------------------------------------------------------------------
/telcons/2019-03-14-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 14 March 2019 - 14:30-15:00 UTC+0
 2 | 
 3 | Note different US time due to [DST change](https://www.timeanddate.com/time/dst/2019a.html):bangbang:
 4 | 
 5 | <table>
 6 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 14 March 2019 <td> 07:30 PST <td> UTC-7 hours :bangbang:
 7 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 14 March 2019 <td> 10:30 EST <td> UTC-4 hours :bangbang:
 8 | <tr><td> London (United Kingdom - England) <td> Thu, 14 March 2019 <td> 14:30 GMT <td> UTC+0 hours
 9 | <tr><td> Berlin (Germany) <td> Thu, 14 March 2019 <td> 15:30 CET <td> UTC+1 hours
10 | <tr><td> Helsinki (Finland) <td> Thu, 14 March 2019 <td> 16:30 EET <td> UTC+2 hours
11 | <tr><td> Shanghai (China) <td> Thu, 14 March 2019 <td> 22:30 CST <td> UTC+8 hours
12 | <tr><td> Tokyo (Japan) <td> Thu, 14 March 2019 <td> 23:30 JST <td> UTC+9 hours
13 | <tr><td> Corresponding UTC (GMT) <td> Thu, 14 March 2019 <td colspan=2> 14:30 UTC
14 | </table>
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC : irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2018Nov/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
23 | * Minutes: https://www.w3.org/2019/03/14-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### [Executing operations #11](https://github.com/webmachinelearning/webnn/issues/11)
28 | * Discuss requirements for an API for executing operations, clarify any unknowns
29 | 
30 | #### [Custom operations #6](https://github.com/webmachinelearning/webnn/issues/6)
31 | * Interim update from custom operations support on CPU ([Wasm custom ops investigation report](https://docs.google.com/presentation/d/1xOQtyAeQNdI70rICzLNKRxmzM-Dwjhh3MdzGKW29C8Q/))
32 | 
33 | #### Industry events
34 | * Discuss TensorFlow Dev Summit takeaways relevant to this group
35 | 


--------------------------------------------------------------------------------
/telcons/2019-04-11-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 11 April 2019 - 13:30-14:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 11 April 2019 <td> 06:30 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 11 April 2019 <td> 09:30 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 11 April 2019 <td> 14:30 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 11 April 2019 <td> 15:30 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 11 April 2019 <td> 16:30 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 11 April 2019 <td> 21:30 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 11 April 2019 <td> 22:30 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 11 April 2019 <td colspan=2> 13:30 UTC
12 | </table>
13 |  
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC : irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2018Nov/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/04/11-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * Foundation specification
26 |   * Discuss & review concrete API proposals ([WebNN Chromium POC](https://github.com/intel/webml-polyfill/blob/master/docs/api.md), [JIT proposal](https://github.com/webmachinelearning/webnn/issues/11#issuecomment-474363912), others)
27 |   * Kick off concrete API sketching at https://webmachinelearning.github.io/webnn/
28 | * F2F meetings
29 |   * Discuss & decide plans for [2019 TPAC F2F meeting in Fukuoka, Japan](https://www.w3.org/2019/09/TPAC/)
30 |   * Discuss possible earlier F2F in June
31 | * Telecon cadence
32 |   * Gauge support for bi-weekly call, 30/60 mins
33 | 


--------------------------------------------------------------------------------
/telcons/2019-04-25-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 25 April 2019 - 14:00-15:00 UTC+0
 2 | 
 3 | :bangbang: This call take place 30 minutes later than our previous calls.
 4 | 
 5 | <table>
 6 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 25 April 2019 <td> 07:00 PDT <td> UTC-7 hours
 7 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 25 April 2019 <td> 10:00 EDT <td> UTC-4 hours
 8 | <tr><td> London (United Kingdom - England) <td> Thu, 25 April 2019 <td> 15:00 BST <td> UTC+1 hours
 9 | <tr><td> Berlin (Germany) <td> Thu, 25 April 2019 <td> 16:00 CEST <td> UTC+2 hours
10 | <tr><td> Helsinki (Finland) <td> Thu, 25 April 2019 <td> 17:00 EEST <td> UTC+3 hours
11 | <tr><td> Shanghai (China) <td> Thu, 25 April 2019 <td> 22:00 CST <td> UTC+8 hours
12 | <tr><td> Tokyo (Japan) <td> Thu, 25 April 2019 <td> 23:00 JST <td> UTC+9 hours
13 | <tr><td> Corresponding UTC (GMT) <td> Thu, 25 April 2019 <td colspan=2> 14:00 UTC
14 | </table>
15 |  
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: Greg Whitworth
20 | * IRC : irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
23 | * Minutes: https://www.w3.org/2019/04/25-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | * Review foundation specification
28 |   * https://github.com/webmachinelearning/webnn/pull/15
29 | 


--------------------------------------------------------------------------------
/telcons/2019-05-09-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 9 May 2019 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 9 May 2019 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 9 May 2019 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 9 May 2019 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 9 May 2019 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 9 May 2019 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 9 May 2019 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 9 May 2019 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 9 May 2019 <td colspan=2> 14:00 UTC
12 | </table>
13 |  
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC : irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/05/09-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * Graph-building syntax simpler for web developers
26 |   * https://github.com/webmachinelearning/webnn/issues/16
27 | * Define the set of operations and their specification
28 |   * https://github.com/webmachinelearning/webnn/issues/17
29 | * W3C Workshop on Web & Machine Learning
30 |   *  https://www.w3.org/2018/11/machine-learning-workshop/
31 | 


--------------------------------------------------------------------------------
/telcons/2019-06-06-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 6 June 2019 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 6 June 2019 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 6 June 2019 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 6 June 2019 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 6 June 2019 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 6 June 2019 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 6 June 2019 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 6 June 2019 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 6 June 2019 <td colspan=2> 14:00 UTC
12 | </table>
13 |  
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC : irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/06/06-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * Survey graph-building APIs from native ecosystem
26 |   * https://github.com/webmachinelearning/webnn/issues/20
27 | * Graph-building syntax simpler for web developers
28 |   * https://github.com/webmachinelearning/webnn/issues/16
29 | 


--------------------------------------------------------------------------------
/telcons/2019-06-27-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 27 June 2019 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 27 June 2019 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 27 June 2019 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 27 June 2019 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 27 June 2019 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 27 June 2019 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 27 June 2019 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 27 June 2019 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 27 June 2019 <td colspan=2> 14:00 UTC
12 | </table>
13 |  
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC : irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/06/27-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * Graph-building syntax simpler for web developers - review proposed API changes
26 |   * https://github.com/webmachinelearning/webnn/issues/16#issuecomment-502921949
27 | * Survey graph-building APIs from native ecosystem - discuss TensorRT findings, survey conclusions
28 |   * https://github.com/webmachinelearning/webnn/issues/20#issuecomment-501255805
29 | * Define the set of operations and their specification
30 |   * https://github.com/webmachinelearning/webnn/issues/17
31 | * Custom operations - framework integration case study report out: WebNN backend prototype for ONNX.js
32 |   * https://github.com/webmachinelearning/webnn/issues/6#issuecomment-505877911
33 | 


--------------------------------------------------------------------------------
/telcons/2019-08-08-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 8 August 2019 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 8 August 2019 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 8 August 2019 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 8 August 2019 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 8 August 2019 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 8 August 2019 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 8 August 2019 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 8 August 2019 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 8 August 2019 <td colspan=2> 14:00 UTC
12 | </table>
13 |  
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/08/08-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * Define the set of operations and their specification
26 |   * PROPOSED RESOLUTION: The specification will reference the ONNX operations and if there are any improvements desired for ONNX the work should be there.
27 |   * Discuss and resolve objections raised for the proposed resolution
28 |   * Clarify proposed resolution based on feedback where appropriate (versioning etc.)
29 |   * https://github.com/webmachinelearning/webnn/issues/17
30 | 
31 | * Graph-building syntax simpler for web developers
32 |   * RESOLVED RESOLUTION: Craft a spec PR based on https://github.com/webmachinelearning/webnn/issues/16#issuecomment-504305802 considering feedback on the call
33 |   * Review the PR: https://github.com/webmachinelearning/webnn/pull/22
34 | 


--------------------------------------------------------------------------------
/telcons/2019-08-22-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 22 August 2019 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 22 August 2019 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 22 August 2019 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 22 August 2019 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 22 August 2019 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 22 August 2019 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 22 August 2019 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 22 August 2019 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 22 August 2019 <td colspan=2> 14:00 UTC
12 | </table>
13 |  
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/08/22-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * Define the set of operations and their specification
26 |   * Ops compatibility study, status (leads: Daniel and Nikhil)
27 |   * ONNX namespace for WebNN ops, discuss process how to make this happen (Paul?)
28 |   * https://github.com/webmachinelearning/webnn/issues/17
29 | 
30 | * Graph-building syntax simpler for web developers
31 |   * Review the PR: https://github.com/webmachinelearning/webnn/pull/22
32 |   
33 | * Custom ops
34 |   * Discuss topics (sharing memory etc.) we need to understand better (lead: Ningxin, contributors: Rapfael, James, Kai)
35 |   * https://github.com/webmachinelearning/webnn/issues/6
36 |   
37 | * WebML CG F2F meeting at TPAC
38 |   * Fukuoka, Japan, 17 & 20 Sep 2019
39 |   * About TPAC: https://www.w3.org/2019/09/TPAC/
40 |   * Registration: https://www.w3.org/2019/09/TPAC/registration.html
41 |   * WebML CG agenda (PRs welcome): https://github.com/webmachinelearning/meetings/tree/master/2019-09-17-fukuoka
42 | 


--------------------------------------------------------------------------------
/telcons/2019-09-05-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 5 September 2019 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 5 September 2019 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 5 September 2019 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 5 September 2019 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 5 September 2019 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 5 September 2019 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 5 September 2019 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 5 September 2019 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 5 September 2019 <td colspan=2> 14:00 UTC
12 | </table>
13 |  
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/09/05-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * Ops compatibility study
26 |   * Leads: Daniel and Nikhil
27 |   * https://github.com/webmachinelearning/webnn/issues/17
28 |   
29 | * Backwards compatibility of the initial set of ops
30 |   * (Branched off from ONNX namespace discussion)
31 |   * https://github.com/webmachinelearning/webnn/issues/24
32 |   
33 | * Custom ops
34 |   * Lead: Ningxin, contributors: Rafael, James, Kai
35 |   * Data layout issue, GPU-based custom ops investigation plan 
36 |   * https://github.com/webmachinelearning/webnn/issues/6
37 |   
38 | * F2F agenda building
39 |   * WebML CG (Tue & Fri): https://github.com/webmachinelearning/meetings/tree/master/2019-09-17-fukuoka
40 |   * Breakouts (Wed): https://www.w3.org/wiki/TPAC/2019/SessionIdeas (relevant e.g. WebGPU, audio/video processing ...)
41 | 
42 | * Other
43 |   * Next bi-weekly teleconference Thu 19 Sep 2019 cancelled due to F2F on the same week
44 | 


--------------------------------------------------------------------------------
/telcons/2019-10-03-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 3 October 2019 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 3 October 2019 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 3 October 2019 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 3 October 2019 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 3 October 2019 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 3 October 2019 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 3 October 2019 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 3 October 2019 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 3 October 2019 <td colspan=2> 14:00 UTC
12 | </table>
13 |  
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2019/10/03-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | * F2F recap ([agenda](https://github.com/webmachinelearning/meetings/blob/master/2019-09-17-fukuoka/README.md), minutes: [Day 1](https://www.w3.org/2019/09/17-webmachinelearning-minutes.html), [Day 2](https://www.w3.org/2019/09/20-webmachinelearning-minutes.html))
26 | 
27 | * WebNN interop investigation [next steps](https://docs.google.com/presentation/d/1KGRc1RnnYt_1JK2Pk6r2xRkD60v4F8jc4beHMv0crng/edit#slide=id.g6353211274_0_23):
28 |   * Support key ops that access hardware acceleration [#17](https://github.com/webmachinelearning/webnn/issues/17) e.g. [conv2d](https://github.com/webmachinelearning/webnn/issues/28), [matmul](https://github.com/webmachinelearning/webnn/issues/27) (lead: Dean/Apple, Nikhil)
29 |   * Support compiling and executing ops for devices, CPU or GPU [new issue] (lead: Ningxin)
30 |   * Support interop with WebAssembly and WebGPU compute shader (up for grabs)
31 |   * Support executing ops chain with opaque operands [#11](https://github.com/webmachinelearning/webnn/issues/11) (up for grabs)
32 |   * Explore custom op support by DSL [new issue] (Kai?)
33 | 


--------------------------------------------------------------------------------
/telcons/2019-10-17-agenda.md:
--------------------------------------------------------------------------------
1 | ### CANCELLED - WebML CG Teleconference – 17 October 2019 - 14:00-15:00 UTC+0
2 | 
3 | 2019-10-17 teleconference is CANCELLED.
4 | 
5 | Topics from this meeting are redirected to our next call, see [2019-10-31-agenda.md](https://github.com/webmachinelearning/meetings/blob/master/telcons/2019-10-31-agenda.md).
6 | 


--------------------------------------------------------------------------------
/telcons/2019-10-31-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 31 October 2019 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 31 October 2019 <td> 08:00 PDT :bangbang: <td> UTC-7 hours 
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 31 October 2019 <td> 11:00 EDT :bangbang: <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 31 October 2019 <td> 15:00 BST <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 31 October 2019 <td> 16:00 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 31 October 2019 <td> 17:00 EEST <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 31 October 2019 <td> 23:00 CST :bangbang: <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 1 November 2019 <td> 00:00 JST :bangbang: <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 31 October 2019 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | :bangbang: Means the place/city is observing [Daylight Saving Time](https://www.timeanddate.com/time/dst/events.html) (DST) at time shown. Between October 27 and November 3, the time difference between US and Europe will be 1 hour less than it currently is. China, Japan and Russia do not observe DST, which means the time difference between these areas and US/Europe will be one hour more than it currently is.
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
23 | * Minutes: https://www.w3.org/2019/10/31-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | * WebNN interop investigation
28 |   * Review POC requirements for buffer sharing between dedicated ML hardware and GPU (Ningxin, Paul)
29 | 
30 | * Op compatibility
31 |   * Review concrete spec proposals for conv2d [#28](https://github.com/webmachinelearning/webnn/issues/28), matMul [#27](https://github.com/webmachinelearning/webnn/issues/27)
32 | 
33 | * Explainer
34 |   * Discuss how to move [explainer.md](https://github.com/webmachinelearning/webnn/blob/master/explainer.md) forward, priorities
35 | 
36 | * On-device training (exploratory topic)
37 |   * Discuss key use cases, feasibility, platform support (Core ML 3, ONNX 1.7 ...)
38 |   * Note: currently [out of scope](https://webmachinelearning.github.io/charter/#out-of-scope), a future opportunity subject to [group agreement](https://webmachinelearning.github.io/charter/#charter-change)
39 | 


--------------------------------------------------------------------------------
/telcons/2019-11-14-agenda.md:
--------------------------------------------------------------------------------
1 | ### CANCELLED - WebML CG Teleconference – 14 November 2019 - 15:00-16:00 UTC+0
2 | 
3 | 2019-11-14 teleconference is CANCELLED.
4 | 
5 | Topics from this meeting are redirected to our next call, see [2019-11-28-agenda.md](https://github.com/webmachinelearning/meetings/blob/master/telcons/2019-11-28-agenda.md).
6 | 


--------------------------------------------------------------------------------
/telcons/2019-11-28-agenda.md:
--------------------------------------------------------------------------------
1 | ### CANCELLED - WebML CG Teleconference – 28 November 2019 - 15:00-16:00 UTC+0
2 | 
3 | 2019-11-28 teleconference is CANCELLED due to Thanksgiving.
4 | 
5 | Topics from this meeting are redirected to our next call, see [2019-12-05-agenda.md](https://github.com/webmachinelearning/meetings/blob/master/telcons/2019-12-05-agenda.md).
6 | 


--------------------------------------------------------------------------------
/telcons/2019-12-05-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 5 December 2019 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 5 December 2019 <td> 07:00 PST <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 5 December 2019 <td> 10:00 EST <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 5 December 2019 <td> 15:00 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 5 December 2019 <td> 16:00 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 5 December 2019 <td> 17:00 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 5 December 2019 <td> 23:00 CST <td> UTC+8 hours :bangbang: 
10 | <tr><td> Tokyo (Japan) <td> Fri, 6 December 2019 <td> 00:00 JST <td> UTC+9 hours :bangbang:
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 5 December 2019 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | :bangbang: China and Japan observe China Standard Time (CST) all year.
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
23 | * Minutes: https://www.w3.org/2019/12/05-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | * [investigation] buffer sharing between GPU and ML accelerator [#33](https://github.com/webmachinelearning/webnn/issues/33)
28 | 
29 | * Op compatibility: conv2d [#28](https://github.com/webmachinelearning/webnn/issues/28), matMul [#27](https://github.com/webmachinelearning/webnn/issues/27)
30 | 
31 | * Float16 type support [#26](https://github.com/webmachinelearning/webnn/issues/26) 
32 | 
33 | * Workshop
34 | 


--------------------------------------------------------------------------------
/telcons/2019-12-19-agenda.md:
--------------------------------------------------------------------------------
1 | ### WebML CG Teleconference – 19 December 2019 - 15:00-16:00 UTC+0
2 | 
3 | 2019-12-19 teleconference is CANCELLED.
4 | 
5 | Topics from this meeting are redirected to our next call, see [2020-01-09-agenda.md](https://github.com/webmachinelearning/meetings/blob/master/telcons/2020-01-09-agenda.md).
6 | 


--------------------------------------------------------------------------------
/telcons/2020-01-09-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 9 January 2020 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 9 January 2020 <td> 07:00 PST <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 9 January 2020 <td> 10:00 EST <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 9 January 2020 <td> 15:00 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 9 January 2020 <td> 16:00 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 9 January 2020 <td> 17:00 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 9 January 2020 <td> 23:00 CST <td> UTC+8 hours :bangbang: 
10 | <tr><td> Tokyo (Japan) <td> Fri, 10 January 2020 <td> 00:00 JST <td> UTC+9 hours :bangbang:
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 9 January 2020 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | :bangbang: China and Japan observe China Standard Time (CST) all year.
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
23 | * Minutes: https://www.w3.org/2020/01/09-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | * [investigation] buffer sharing between GPU and ML accelerator [#33](https://github.com/webmachinelearning/webnn/issues/33): update on next steps
28 | 
29 | * conv2d [#28](https://github.com/webmachinelearning/webnn/issues/28): review [native API mapping table](https://github.com/webmachinelearning/webnn/blob/master/op_compatibility/conv2d.md)
30 | 
31 | * Float16 type support [#26](https://github.com/webmachinelearning/webnn/issues/26): proposed resolution to add float16 tensor type
32 | 
33 | * [W3C Workshop on Web & Machine Learning](https://w3c.github.io/machine-learning-workshop/) (24-25 March 2020, Berlin): update & next steps
34 | 


--------------------------------------------------------------------------------
/telcons/2020-01-23-agenda.md:
--------------------------------------------------------------------------------
 1 | ### WebML CG Teleconference – 23 January 2020 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 23 January 2020 <td> 07:00 PST <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 23 January 2020 <td> 10:00 EST <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 23 January 2020 <td> 15:00 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 23 January 2020 <td> 16:00 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 23 January 2020 <td> 17:00 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 23 January 2020 <td> 23:00 CST <td> UTC+8 hours :bangbang: 
10 | <tr><td> Tokyo (Japan) <td> Fri, 24 January 2020 <td> 00:00 JST <td> UTC+9 hours :bangbang:
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 23 January 2020 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | :bangbang: China and Japan observe China Standard Time (CST) all year.
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
23 | * Minutes: https://www.w3.org/2020/01/23-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | * [investigation] buffer sharing between GPU and ML accelerator [#33](https://github.com/webmachinelearning/webnn/issues/33): update on VPU and DirectML prototyping
28 | 
29 | * conv2d [#28](https://github.com/webmachinelearning/webnn/issues/28): DirectML and ONNX input to [native API mapping table](https://github.com/webmachinelearning/webnn/blob/master/op_compatibility/conv2d.md)
30 | 
31 | * Float16 type support [#26](https://github.com/webmachinelearning/webnn/issues/26): [review PR](https://github.com/webmachinelearning/webnn/pull/35)
32 | 
33 | * Handling unsupported OperandType [#36](https://github.com/webmachinelearning/webnn/issues/36): discuss API design considerations for the failure path
34 | 
35 | * [W3C Workshop on Web & Machine Learning](https://w3c.github.io/machine-learning-workshop/) (24-25 March 2020, Berlin): [officially announced](https://www.w3.org/blog/news/archives/8238), [registration](https://www.w3.org/2002/09/wbs/1/machine-learning-workshop/) open until 21 Feb 2020
36 | 


--------------------------------------------------------------------------------
/telcons/2020-02-06-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 6 February 2020 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 6 February 2020 <td> 07:00 PST <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 6 February 2020 <td> 10:00 EST <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 6 February 2020 <td> 15:00 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 6 February 2020 <td> 16:00 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 6 February 2020 <td> 17:00 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 6 February 2020 <td> 23:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 7 February 2020 <td> 00:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 6 February 2020 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2020/02/06-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | #### conv2d and matMul op definitions (15 mins)
26 | 
27 | Flesh out how we proceed adding conv2d and matMul op definitions to the WebNN API spec. Proposed work mode to evolve the op definitions iteratively: first land signature and arguments definitions, refine in subsequent PRs based on compatibility study findings.
28 | 
29 | Please review and provide feedback on the respective issues prior to the call:
30 | 
31 | - [conv2d #28](https://github.com/webmachinelearning/webnn/issues/28) and [compatibility table](https://github.com/webmachinelearning/webnn/blob/master/op_compatibility/conv2d.md)
32 | - [matMul #27](https://github.com/webmachinelearning/webnn/issues/27)
33 | 
34 | #### Revisit inference API to load and run a model (15 mins)
35 | 
36 | Discuss a proposal for inference API to load and run a model.
37 | 
38 | - [issue #41](https://github.com/webmachinelearning/webnn/issues/41)
39 | - [explainer](https://github.com/jbingham/web-ml-inference)
40 | 
41 | #### Workshop program committee sync (30 mins)
42 | 
43 | Discuss the [Web & ML workshop](https://www.w3.org/2020/01/machine-learning-workshop/) with program committee. Primarily for [committee members](https://www.w3.org/2020/01/machine-learning-workshop/#committee), CG participants are welcome to join too.
44 | 


--------------------------------------------------------------------------------
/telcons/2020-02-20-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 20 February 2020 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 20 February 2020 <td> 07:00 PST <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 20 February 2020 <td> 10:00 EST <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 20 February 2020 <td> 15:00 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 20 February 2020 <td> 16:00 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 20 February 2020 <td> 17:00 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 20 February 2020 <td> 23:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 21 February 2020 <td> 00:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 20 February 2020 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2020/02/20-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | #### Announcements
26 | 
27 | 2020-02-13 update: [Web & ML workshop](https://www.w3.org/2020/01/machine-learning-workshop/) postponed sine die due to ongoing challenges around international travel
28 | 
29 | #### conv2d op definition
30 | 
31 | https://github.com/webmachinelearning/webnn/pull/43
32 | 
33 | #### Remove Compilation.finish and setPreference
34 | 
35 | https://github.com/webmachinelearning/webnn/issues/39
36 | 
37 | #### Handling unsupported OperandType
38 | 
39 | https://github.com/webmachinelearning/webnn/issues/36
40 | 
41 | #### WebML synopsis at WebGPU F2F
42 | 
43 | Discuss key WebML takeaways from WebGPU F2F:
44 | 
45 | - [WebGPU CG F2F 2020-02-13 meeting minutes](https://docs.google.com/document/d/1fsAV1SyV-eSfjnA0C4WjpXAJcmRUnkzXAh4AblSotoI/edit#heading=h.e1hdus1hfzn6)
46 | 
47 | #### Inference API to load and run a model
48 | 
49 | Discuss feedback and decide next steps for inference API to load and run a model.
50 | 
51 | - [issue #41](https://github.com/webmachinelearning/webnn/issues/41)
52 | - [explainer](https://github.com/jbingham/web-ml-inference)
53 | 


--------------------------------------------------------------------------------
/telcons/2020-03-05-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 5 March 2020 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 5 March 2020 <td> 07:00 PST <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 5 March 2020 <td> 10:00 EST <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 5 March 2020 <td> 15:00 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 5 March 2020 <td> 16:00 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 5 March 2020 <td> 17:00 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 5 March 2020 <td> 23:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 6 March 2020 <td> 00:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 5 March 2020 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | ### Logistics
15 | 
16 | * Chair: Anssi
17 | * Scribe: ?
18 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
19 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
20 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
21 | * Minutes: https://www.w3.org/2020/03/05-webmachinelearning-minutes.html
22 | 
23 | ### Agenda
24 | 
25 | #### conv2d op definition
26 | 
27 | Review revised PR
28 | 
29 | https://github.com/webmachinelearning/webnn/pull/43
30 | 
31 | #### matMul op definition
32 | 
33 | Discuss opens and decide next steps to get this op landed into the spec
34 | 
35 | https://github.com/webmachinelearning/webnn/issues/27
36 | https://github.com/webmachinelearning/webnn/blob/master/op_compatibility/matmul.md
37 | 
38 | #### Handling unsupported OperandType
39 | 
40 | Discuss how to surface "datatype not supported by underlying platform" errors through the API
41 | 
42 | https://github.com/webmachinelearning/webnn/issues/36
43 | 
44 | #### Inference API to load and run a model
45 | 
46 | Final call for feedback prior to adding this API into the group's scope
47 | 
48 | - [issue #41](https://github.com/webmachinelearning/webnn/issues/41)
49 | - [explainer](https://github.com/jbingham/web-ml-inference)
50 | 


--------------------------------------------------------------------------------
/telcons/2020-03-19-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 19 March 2020 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 19 March 2020 <td> 08:00 PST :bangbang: <td> UTC-7 hours 
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 19 March 2020 <td> 11:00 EST :bangbang: <td> UTC-4 hours 
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 19 March 2020 <td> 15:00 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 19 March 2020 <td> 16:00 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 19 March 2020 <td> 17:00 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 19 March 2020 <td> 23:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 20 March 2020 <td> 00:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 19 March 2020 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | :bangbang: Daylight Saving Time in the US: https://www.timeanddate.com/time/dst/2020a.html
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
23 | * Minutes: https://www.w3.org/2020/03/19-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### matMul op definition
28 | 
29 | Review PR
30 | 
31 | https://github.com/webmachinelearning/webnn/pull/49
32 | 
33 | #### Handling unsupported OperandType
34 | 
35 | https://github.com/webmachinelearning/webnn/issues/36
36 | 
37 | #### Inference API to load and run a model
38 | 
39 | Proposed resolution: add the load and run a model API to the group's scope, the group does not attempt to mandate a specific ML schema or format. No changes to the charter with this clarification.
40 | 
41 | - [issue #41](https://github.com/webmachinelearning/webnn/issues/41)
42 | - [explainer](https://github.com/jbingham/web-ml-inference)
43 | 


--------------------------------------------------------------------------------
/telcons/2020-04-02-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 2 April 2020 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 2 April 2020 <td> 07:00 PST :bangbang:<td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 2 April 2020 <td> 10:00 EST :bangbang:<td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 2 April 2020 <td> 15:00 GMT <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 2 April 2020 <td> 16:00 CET <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 2 April 2020 <td> 17:00 EET <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 2 April 2020 <td> 23:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 3 April 2020 <td> 00:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 2 April 2020 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | :bangbang: Rest of the world transitioned to summer time, the US is back to the normal schedule.
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
23 | * Minutes: https://www.w3.org/2020/04/02-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | #### matMul op definition
27 | 
28 | Review revised PR:
29 | https://github.com/webmachinelearning/webnn/pull/49
30 | 
31 | #### How to handle a failing compilation
32 | 
33 | Review PR:
34 | https://github.com/webmachinelearning/webnn/pull/50
35 | 
36 | Related:
37 | Handling unsupported OperandType
38 | https://github.com/webmachinelearning/webnn/issues/36
39 | 
40 | #### Inference API to load and run a model
41 | 
42 | Dicuss current status and next steps (move repo to CG?), the group resolved to add the API to the group's scope.
43 | 
44 | - [explainer](https://github.com/jbingham/web-ml-inference)
45 | 
46 | #### Virtual workshop
47 | 
48 | Gather feedback and suggestions on the plan
49 | 


--------------------------------------------------------------------------------
/telcons/2020-04-16-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 16 April 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 16 April 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 16 April 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 16 April 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 16 April 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 16 April 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 16 April 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 16 April 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 16 April 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20200416T14&p1=1440
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2019Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/))
23 | * Minutes: https://www.w3.org/2020/04/16-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### New ops for WebNN API
28 | 
29 | Discuss & decide new ops to add to the WebNN API. Currently specified are conv2d and matMul.
30 | 
31 | Priority on ops that accelerate models for tasks that address identified use cases:
32 | https://webmachinelearning.github.io/webnn/#usecases
33 | 
34 | #### Inference API to load and run a model
35 | 
36 | Discuss next steps:
37 | 
38 | - Logistics: repo migration & naming 🚲https://github.com/webmachinelearning/webnn/issues/51
39 | - Spec draft: start with spec text, use cases discussed in explainer
40 | - Prototyping: Chromium and polyfilling plan
41 | - Open: model formats
42 | 
43 | #### Virtual workshop agenda building
44 | 
45 | Gather feedback on priority topics and talks for the virtual workshop pre-recorded offline mode:
46 | 
47 | https://bit.ly/webml-workshop-talks 
48 | 
49 | The virtual workshop scope coverage target 30-70% of the cancelled in-person event.
50 | 
51 | Infrastucture discussion. [W3C's MIT host provides us Zoom](https://www.w3.org/Guide/meetings/zoom.html). Proposal to dogfood outselves with the MIT's Zoom instance to evaluate, use for our upcoming CG calls (this one still on WebEx).
52 | 


--------------------------------------------------------------------------------
/telcons/2020-04-30-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 30 April 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 30 April 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 30 April 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 30 April 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 30 April 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 30 April 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 30 April 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 30 April 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 30 April 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20200430T14&p1=1440
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - :bangbang: we're now trialing Zoom, [app install](https://zoom.us/download) recommended, the web client has issues
23 | * Minutes: https://www.w3.org/2020/04/30-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### WebNN first wave models and ops
28 | 
29 | Discuss WebNN first wave models and ops, action review
30 | 
31 | - Review PR https://github.com/webmachinelearning/webnn/pull/52
32 | - Compare with XLA to see what ops, if any, are too high level (Nikhil & Daniel)
33 | - Review opsets from a year ago to see what couldn't be ran on ONNX opsets (Daniel)
34 | - Intersection between XLA & ONNX (Paul)
35 | 
36 | Next: choose a batch of ops to add to WebNN API informed by homework
37 | 
38 | #### Model Loader API
39 | 
40 | Review spec strawman, please open issues for your feedback
41 | 
42 | - Spec: https://webmachinelearning.github.io/model-loader/
43 | - GitHub: https://github.com/webmachinelearning/model-loader
44 | 
45 | Next: work out a plan for spec definition
46 | 
47 | #### Virtual workshop topics survey
48 | 
49 | Workshop prepwork status
50 | 
51 | - Sketch out details for the "Domain-specific compilers for ML" high-level topic
52 | - Review topics survey: https://bit.ly/webml-workshop-survey
53 | - Review workshop agenda details: https://bit.ly/webml-workshop-talks
54 | 
55 | Next: review survey results and adjust topics accordingly, start outreach to people
56 | 


--------------------------------------------------------------------------------
/telcons/2020-05-14-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 14 May 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 14 May 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 14 May 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 14 May 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 14 May 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 14 May 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 14 May 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 14 May 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 14 May 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20200514T14&p1=1440
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - :bangbang: we're now trialing Zoom, [app install](https://zoom.us/download) recommended, the web client has issues
23 | * Minutes: https://www.w3.org/2020/05/14-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### WebNN first wave models and ops ONNX and XLA-HLO intercept
28 | 
29 | - Review revised PR https://github.com/webmachinelearning/webnn/pull/52
30 | - [x] Compare with XLA to see what ops, if any, are too high level (Nikhil & Daniel)
31 | - [x] Review opsets from a year ago to see what couldn't be ran on ONNX opsets (Daniel)
32 | - [ ] Intersection between XLA & ONNX (Paul)
33 | 
34 | Related [resolution](https://www.w3.org/2020/04/30-webmachinelearning-minutes.html#r01): Add element-wise add, element-wise multiply, concatenation, and reshape ops to the WebNN API
35 | 
36 | #### Element-wise add & mul, concat, reshape, gemm, and transpose ops WebNN API definitions
37 | 
38 | Review PRs for new op definitions:
39 | 
40 | - [ ] element-wise add & mul https://github.com/webmachinelearning/webnn/pull/54
41 | - [ ] concat https://github.com/webmachinelearning/webnn/pull/55
42 | - [ ] reshape https://github.com/webmachinelearning/webnn/pull/57
43 | - [x] General Matrix Multiply (gemm) and transpose https://github.com/webmachinelearning/webnn/pull/58
44 | 
45 | #### WebAssembly System Interface (WASI) machine learning module
46 | 
47 | Review and solicit feedback for the WASI-nn proposal, see the [presentation slides](https://github.com/webmachinelearning/meetings/blob/master/telcons/2020-05-14-wasi-nn.pdf)
48 | 
49 | https://github.com/WebAssembly/WASI/issues/272
50 | 
51 | (via https://github.com/webmachinelearning/webnn/issues/32)
52 | 
53 | Note: Currently this CG is not chartered to consider WASI but that could change in the future given adequate support for such a direction.
54 | 


--------------------------------------------------------------------------------
/telcons/2020-05-14-wasi-nn.pdf:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/webmachinelearning/meetings/08940382164751e87b90e815c4da22516ab42ff8/telcons/2020-05-14-wasi-nn.pdf


--------------------------------------------------------------------------------
/telcons/2020-05-28-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 28 May 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 28 May 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 28 May 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 28 May 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 28 May 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 28 May 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 28 May 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 28 May 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 28 May 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20200528T14&p1=1440
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended, the web client has issues
23 | * Minutes: https://www.w3.org/2020/05/28-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Intersection between XLA & ONNX
28 | 
29 | Report status of the investigation on intersection between XLA & ONNX (Paul)
30 | 
31 | Reflect findings into the [first-wave models](https://github.com/webmachinelearning/webnn/blob/master/op_compatibility/first_wave_models.md).
32 | 
33 | #### ReLU and max op WebNN API definitions
34 | 
35 | PR reviewed and merged:
36 | 
37 | https://github.com/webmachinelearning/webnn/pull/62
38 | 
39 | #### LeakyReLU and min op WebNN API definitions
40 | 
41 | Review PR:
42 | 
43 | https://github.com/webmachinelearning/webnn/pull/63
44 | 
45 | #### Variations of supported convolution type
46 | 
47 | Discuss two variants of the convolution operation we should consider:
48 | - Grouped convolution
49 | - Transposed convolution
50 | 
51 | https://github.com/webmachinelearning/webnn/issues/60
52 | 
53 | #### Noise suppression use case 
54 | 
55 | Review PR:
56 | 
57 | https://github.com/webmachinelearning/webnn/pull/61
58 | 
59 | Any new ops that should be reflected into the first-wave models table?
60 | 
61 | #### Model Loader API
62 | 
63 | Discuss progress on the spec and prototype (Jonathan)
64 | 


--------------------------------------------------------------------------------
/telcons/2020-06-11-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 11 June 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 11 June 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 11 June 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 11 June 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 11 June 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 11 June 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 11 June 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 11 June 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 11 June 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20200611T14&p1=1440
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended, the web client has issues
23 | * Minutes: https://www.w3.org/2020/06/11-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Announced: W3C Virtual Workshop on Web and Machine Learning
28 | 
29 | [Announcement](https://www.w3.org/blog/news/archives/8575), [apply to speak (due July 3rd, 2020)](https://www.w3.org/2020/06/machine-learning-workshop/speakers.html#submit)
30 | 
31 | #### LeakyReLU and min ops
32 | 
33 | Review PR:
34 | 
35 | https://github.com/webmachinelearning/webnn/pull/63
36 | 
37 | 
38 | #### 2D pooling and reduce ops
39 | 
40 | Review PR:
41 | 
42 | https://github.com/webmachinelearning/webnn/pull/64
43 | 
44 | #### Grouped conv2d
45 | 
46 | Review PR:
47 | 
48 | https://github.com/webmachinelearning/webnn/pull/65
49 | 
50 | #### Noise suppression
51 | 
52 | Noise suppression use case adopted by WebNN API:
53 | 
54 | https://webmachinelearning.github.io/webnn/#usecase-noise-suppression
55 | 
56 | Discuss next steps outlined at:
57 | 
58 | https://github.com/webmachinelearning/webnn/issues/66
59 | 


--------------------------------------------------------------------------------
/telcons/2020-06-25-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 25 June 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 25 June 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 25 June 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 25 June 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 25 June 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 25 June 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 25 June 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 25 June 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 25 June 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20200625T14&p1=1440
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended, the web client has issues
23 | * Minutes: https://www.w3.org/2020/06/25-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Set inputs and outputs by string key
28 | 
29 | New feature to allow a graph to declare the keys of its inputs and outputs.
30 | 
31 | Review PR:
32 | https://github.com/webmachinelearning/webnn/pull/69
33 | 
34 | #### Add softmax and relevant element-wise binary and unary ops
35 | 
36 | Softmax required by the first-wave models.
37 | 
38 | Review PR:
39 | https://github.com/webmachinelearning/webnn/pull/68
40 | 
41 | #### Noise suppression
42 | 
43 | Discuss next steps outlined at:
44 | 
45 | https://github.com/webmachinelearning/webnn/issues/66
46 | 
47 | #### WebNN API Explainer
48 | 
49 | Discuss and clarify the spec and explainer relationship.
50 | 
51 | https://github.com/webmachinelearning/webnn/blob/master/explainer.md
52 | 
53 | Related:
54 | https://github.com/webmachinelearning/webnn/issues/70
55 | 


--------------------------------------------------------------------------------
/telcons/2020-08-20-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 20 August 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 20 August 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 20 August 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 20 August 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 20 August 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 20 August 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 20 August 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 20 August 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 20 August 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20200820T14&p1=1440
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended
23 | * Minutes: https://www.w3.org/2020/08/20-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Welcome back after the break!
28 | 
29 | #### Announcement: WebML CG meeting at TPAC
30 | 
31 | Proposed WebML CG Meeting at TPAC 2020 on October 22 15:00 UTC:
32 | 
33 | https://www.w3.org/wiki/TPAC/2020/GroupMeetings
34 | 
35 | TPAC is an opportunity to interact with the broader W3C community.
36 | 
37 | Suggestions for topics for joint meetings with other W3C groups: WebGPU, Wasm, Audio?
38 | 
39 | #### WebNN polyfill and samples
40 | 
41 | The group agreed to add WebNN API polyfill and samples to its deliverables:
42 | 
43 | https://github.com/webmachinelearning/webnn/issues/81
44 | 
45 | https://github.com/webmachinelearning/webnn/issues/70
46 | 
47 | Introduction to the initial contribution from Intel proposed to be used as a starting point, latest status, issues, plan, areas that welcome further contributions etc. (Ningxin)
48 | 
49 | #### First-wave models and ops delta with WebNN API definition
50 | 
51 | Review the delta between the [first-wave models](https://github.com/webmachinelearning/webnn/blob/master/op_compatibility/first_wave_models.md) ops and [WebNN API](https://webmachinelearning.github.io/webnn/#api) surface:
52 | 
53 | - [ ] clamp
54 | - [ ] globalAveragePool
55 | - [ ] gru
56 | - [ ] sigmoid
57 | - [ ] split
58 | - [ ] squeeze
59 | 
60 | Discuss which ops to add to the spec definition considering e.g. major platform support and performance implications.
61 | 
62 | #### Noise suppression
63 | 
64 | Discuss next steps outlined at:
65 | 
66 | https://github.com/webmachinelearning/webnn/issues/66
67 | 
68 | Possible joint meeting at TPAC with Audio WG to help chart path forward?
69 | 
70 | 


--------------------------------------------------------------------------------
/telcons/2020-09-03-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 3 September 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 3 September 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 3 September 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 3 September 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 3 September 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 3 September 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 3 September 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 3 September 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 3 September 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20200903T14
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended
23 | * Minutes: https://www.w3.org/2020/09/03-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Workshop presentations and GH discussions
28 | 
29 | Workshop presentations are actively being discussed on GitHub. CG participants feedback has been appreciated, further engagement welcome.
30 | 
31 | https://www.w3.org/2020/06/machine-learning-workshop/presentations.html
32 | 
33 | https://github.com/w3c/machine-learning-workshop/issues
34 | 
35 | #### WebNN polyfill and samples
36 | 
37 | Discuss review feedback and suggestions for the foundational implementation
38 | 
39 | https://github.com/webmachinelearning/webnn-polyfill/pull/1
40 | 
41 | Discuss and review LeNet sample for handwritten digit recognition using WebNN
42 | 
43 | https://github.com/webmachinelearning/webnn-samples/pull/1
44 | 
45 | #### GRU and corresponding ops
46 | 
47 | Fill the operator gaps to support noise suppression first-wave models
48 | 
49 | Adding GRU and GRUcell operators to support GRU recurrent network. Defining a cell operator in addition to the network operator for added customization flexibility e.g. to support stacked cell recurrent network, etc.
50 | 
51 | Review PR:
52 | https://github.com/webmachinelearning/webnn/pull/83
53 | 
54 | Proposal to close noise suppression issue with PR #83:
55 | 
56 | https://github.com/webmachinelearning/webnn/issues/66
57 | 
58 | #### First-wave models and ops delta with WebNN API definition
59 | 
60 | Review the delta between the [first-wave models](https://github.com/webmachinelearning/webnn/blob/master/op_compatibility/first_wave_models.md) ops and [WebNN API](https://webmachinelearning.github.io/webnn/#api) surface:
61 | 
62 | - [ ] clamp
63 | - [ ] globalAveragePool - Lowering to reducewindow, add and div?
64 | - [x] gru
65 | - [x] sigmoid
66 | - [ ] split - Lowering to slice?
67 | - [x] squeeze
68 | 
69 | Discuss which ops to add to the spec definition considering e.g. major platform support and performance implications.
70 | 


--------------------------------------------------------------------------------
/telcons/2020-09-17-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 17 September 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 17 September 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 17 September 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 17 September 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 17 September 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 17 September 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 17 September 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 17 September 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 17 September 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20200917T14
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended
23 | * Minutes: https://www.w3.org/2020/09/17-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Model Execution API
28 | 
29 | Discuss proposed improvements to the existing execution API:
30 | 
31 | https://github.com/webmachinelearning/webnn/issues/87
32 | 
33 | #### Packing operations for gemm / matmul
34 | 
35 | Discuss optional packing ops and related optimization opportunities:
36 | 
37 | https://github.com/webmachinelearning/webnn/issues/86
38 | 
39 | #### Fingerprinting
40 | 
41 | Discuss possible fingerprinting vectors and mitigations:
42 | 
43 | https://github.com/webmachinelearning/webnn/issues/85
44 | 
45 | 
46 | #### WebNN polyfill and samples
47 | 
48 | Continue discuss review feedback and suggestions for the foundational implementation and LeNet sample:
49 | 
50 | https://github.com/webmachinelearning/webnn-polyfill/pull/1
51 | 
52 | https://github.com/webmachinelearning/webnn-samples/pull/1
53 | 


--------------------------------------------------------------------------------
/telcons/2020-10-01-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 1 October 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 1 October 2020 <td> 07:00 PDT <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 1 October 2020 <td> 10:00 EDT <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 1 October 2020 <td> 15:00 BST <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 1 October 2020 <td> 16:00 CEST <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 1 October 2020 <td> 17:00 EEST <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 1 October 2020 <td> 22:00 CST <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 1 October 2020 <td> 23:00 JST <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 1 October 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20201001T14
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended
23 | * Minutes: https://www.w3.org/2020/10/01-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Model Execution API
28 | 
29 | Continue discuss proposed improvements to the execution API and associated API surface:
30 | 
31 | https://github.com/webmachinelearning/webnn/issues/87
32 | 
33 | #### WebNN code editor
34 | 
35 | Review and solicit feedback on the WebNN code editor:
36 | 
37 | https://github.com/webmachinelearning/webnn-samples/pull/11
38 | 
39 | https://github.com/webmachinelearning/webnn-samples/issues/4
40 | 
41 | Staging: https://huningxin.github.io/webnn-samples/code/
42 | 
43 | #### WebNN API explainer
44 | 
45 | TAG review readiness depends on the explainer document:
46 | 
47 | https://github.com/webmachinelearning/webnn/blob/master/explainer.md
48 | 
49 | Discuss how to approach the following explainer sections:
50 | 
51 | - Key scenarios
52 | - Detailed design discussion
53 | - Considered alternatives
54 | 
55 | #### Workshop proceedings at TPAC 2020
56 | 
57 | WebML CG Virtual Meeting at TPAC 2020 discusses the W3C Workshop on Web and Machine Learning key outcomes with a focus on proposed near-term and long-term standardization directions and next steps.
58 | 
59 | WebML CG participants are invited.
60 | 
61 | >- WebML CG Virtual Meeting at TPAC 2020
62 | >- 22 October 2020 - 14:00-15:00 UTC+0
63 | >- Agenda: https://github.com/webmachinelearning/meetings/tree/master/2020-10-22-tpac2020
64 | 
65 | #### Proposals for future work
66 | 
67 | New repo created to capture proposals for future work that are not in scope of any group currently:
68 | https://github.com/webmachinelearning/proposals
69 | 


--------------------------------------------------------------------------------
/telcons/2020-10-29-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 29 October 2020 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 29 October 2020 <td> 07:00 <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 29 October 2020 <td> 10:00 <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 29 October 2020 <td> 14:00 :bangbang: <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 29 October 2020 <td> 15:00 :bangbang: <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 29 October 2020 <td> 16:00 :bangbang: <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 29 October 2020 <td> 22:00 <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 29 October 2020 <td> 23:00 <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 29 October 2020 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | :bangbang: On Sunday, October 25, 2020, clocks will be set back 1 hour in most of Europe as Daylight Saving Time (DST) ends. (https://www.timeanddate.com/news/time/europe-dst-end-2020.html)
15 | 
16 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20201029T14
17 | 
18 | ### Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ?
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
24 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended
25 | * Minutes: https://www.w3.org/2020/10/29-webmachinelearning-minutes.html
26 | 
27 | ### Agenda
28 | 
29 | #### WebNN API explainer
30 | 
31 | TAG review readiness depends on the explainer document:
32 | 
33 | https://github.com/webmachinelearning/webnn/blob/master/explainer.md
34 | 
35 | Review PR:
36 | 
37 | https://github.com/webmachinelearning/webnn/pull/109
38 | 
39 | #### Support the execution of the sub-graph scenario
40 | 
41 | https://github.com/webmachinelearning/webnn/issues/105
42 | 
43 | #### Proposed new ops Mirrorpad, SquaredDifference, Pow, TransposeConv
44 | 
45 | Please review and provide feedback:
46 | 
47 | https://github.com/webmachinelearning/webnn/issues/108
48 | 
49 | #### Secure Context
50 | 
51 | Review PR:
52 | 
53 | https://github.com/webmachinelearning/webnn/pull/101
54 | 
55 | #### WebNN API ergonomics improvements
56 | 
57 | Subtopic: Chained API for the Operands
58 | 
59 | https://github.com/webmachinelearning/webnn/issues/106
60 | 
61 | Subtopic: Options dictionary for functions with lots of args
62 | 
63 | https://github.com/webmachinelearning/webnn/issues/90
64 | 
65 | 
66 | 


--------------------------------------------------------------------------------
/telcons/2020-11-12-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 12 November 2020 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 12 November 2020 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 12 November 2020 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 12 November 2020 <td> 15:00 <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 12 November 2020 <td> 16:00 <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 12 November 2020 <td> 17:00 <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 12 November 2020 <td> 23:00 :bangbang: <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 13 November 2020 <td> 00:00 :bangbang: <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 12 November 2020 <td colspan=2> 15:00 UTC :bangbang:
12 | </table>
13 | 
14 | :bangbang: Daylight Saving Time ended in the EU and US. China and Japan are now +1 hour (https://www.timeanddate.com/news/time/europe-dst-end-2020.html).
15 | 
16 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20201112T15
17 | 
18 | ### Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ?
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
24 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended
25 | * Minutes: https://www.w3.org/2020/11/12-webmachinelearning-minutes.html
26 | 
27 | ### Agenda
28 | 
29 | #### WebNN API TAG spec review submission
30 | 
31 | Final review of the TAG spec review request before submission:
32 | 
33 | https://github.com/webmachinelearning/webnn/issues/89#issuecomment-722415240
34 | 
35 | Please provide your feedback in the issue ahead the meeting.
36 | 
37 | Subtopic: Self-Review Questionnaire: Security and Privacy
38 | 
39 | Contribute your suggested responses to the questionnaire:
40 | 
41 | https://github.com/webmachinelearning/webnn/issues/119
42 | 
43 | #### Support the execution of the sub-graph scenario
44 | 
45 | Discuss and provide feedback on the preferred builder pattern:
46 | 
47 | https://github.com/webmachinelearning/webnn/issues/105
48 | 
49 | #### Proposed new ops Mirrorpad, SquaredDifference, Pow, TransposeConv
50 | 
51 | Review low-level ops decomposition and gaps:
52 | 
53 | https://github.com/webmachinelearning/webnn/issues/108
54 | 
55 | #### Specify the ModelBuilder.createModel
56 | 
57 | Discuss naming and how to spec the ModelBuilder.createModel:
58 | 
59 | https://github.com/webmachinelearning/webnn/issues/107
60 | 
61 | https://webmachinelearning.github.io/webnn/#dom-modelbuilder-createmodel
62 | 
63 | #### Chained API for the Operands
64 | 
65 | Follow-up on the mixin interface proposal:
66 | 
67 | https://github.com/webmachinelearning/webnn/issues/106
68 | 
69 | #### WG Charter feedback
70 | 
71 | Status check and discussion on WG Charter issues; please file any issues at:
72 | 
73 | https://github.com/w3c/machine-learning-charter/issues
74 | 


--------------------------------------------------------------------------------
/telcons/2020-11-26-agenda.md:
--------------------------------------------------------------------------------
1 | ## WebML CG Teleconference – 26 November 2020 - 15:00-16:00 UTC+0
2 | 
3 | ### This meeting is cancelled due to the Thanksgiving in the US.
4 | 


--------------------------------------------------------------------------------
/telcons/2021-01-21-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 21 January 2021 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 21 January 2021 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 21 January 2021 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 21 January 2021 <td> 15:00 <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 21 January 2021 <td> 16:00 <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 21 January 2021 <td> 17:00 <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 21 January 2021 <td> 23:00 <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 8 January 2021 <td> 00:00 <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 21 January 2021 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20210121T15
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html (login with your [W3C account](https://www.w3.org/Help/Account/)) - [Zoom app install](https://zoom.us/download) recommended
23 | * Minutes: https://www.w3.org/2021/01/21-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### webnn-native
28 | 
29 | Present the proposed standalone native implementation of the WebNN API spec called `webnn-native` that complements the JS implementation `webnn-polyfill`. Both the implementations help inform the API spec development, validate implementability atop native APIs (initial targets OpenVINO and DirectML, extensible with other backends), and provide a performance benchmark. (Ningxin)
30 | 
31 | Discuss the proposal to adopt `webnn-native` as a CG deliverable similarly to `webnn-polyfill`. No CG Charter changes needed, is considered "Other Software" per the existing charter text.
32 | 
33 | https://github.com/webmachinelearning/webnn/issues/129
34 | 
35 | #### WebNN conv2d layout parameter TensorFlow incompatibility
36 | 
37 | Discuss and gather further feedback on TF, TF Lite, and TF.js preference for input and filter layout approach in WebNN API:
38 | 
39 | https://github.com/webmachinelearning/webnn/issues/125
40 | 
41 | #### Numerical precision in conformance testing
42 | 
43 | Discuss and solicit insights on difference between TF.js WebGL and CPU/Wasm backends:
44 | 
45 | https://github.com/webmachinelearning/webnn-polyfill/pull/32
46 | 
47 | #### NSNet2 sample and TF.js memory leak
48 | 
49 | Check TF.js upstream blocker status and discuss any other opens blocking PR:
50 | 
51 | https://github.com/webmachinelearning/webnn-samples/pull/22
52 | 
53 | https://github.com/tensorflow/tfjs/issues/4378
54 | 
55 | #### Proposals for future work
56 | 
57 | Continue discuss proposals submitted for consideration for future work:
58 | 
59 | - Data processing: https://github.com/webmachinelearning/proposals/issues/1
60 | 
61 | - Operation-specific APIs: https://github.com/webmachinelearning/proposals/issues/2
62 | 
63 | #### Security and Privacy considerations
64 | 
65 | Last Call to review the proposed responses to the Self-Review Questionnaire for Security and Privacy: 
66 | 
67 | https://github.com/webmachinelearning/webnn/issues/119
68 | 


--------------------------------------------------------------------------------
/telcons/2021-02-04-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 4 February 2021 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 4 February 2021 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 4 February 2021 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 4 February 2021 <td> 15:00 <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 4 February 2021 <td> 16:00 <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 4 February 2021 <td> 17:00 <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 4 February 2021 <td> 23:00 <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 5 February 2021 <td> 00:00 <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 4 February 2021 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20210204T15
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
23 | * Minutes: https://www.w3.org/2021/02/04-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### TAG review feedback dissemination
28 | 
29 | Discuss the first-pass TAG review feedback, formulate responses to key questions from the TAG. (All)
30 | 
31 | Spec feedback:
32 | 
33 | https://github.com/w3ctag/design-reviews/issues/570#issuecomment-768875996
34 | 
35 | Explainer feedback:
36 | 
37 | https://github.com/w3ctag/design-reviews/issues/570#issuecomment-768151707
38 | 
39 | https://github.com/w3ctag/design-reviews/issues/570#issuecomment-768861645
40 | 
41 | #### WG Charter update
42 | 
43 | (Update deferred due to W3C Staff scheduling conflict.)
44 | 
45 | WG Charter update and next steps. (@anssiko, @dontcallmedom)
46 | 
47 | https://github.com/w3c/machine-learning-charter/
48 | 
49 | #### Security and Privacy self-review submission
50 | 
51 | Privacy and Security feedback:
52 | 
53 | https://github.com/webmachinelearning/webnn/issues/119#issuecomment-772565085
54 | 
55 | Review PR of the initial submission (All, editors):
56 | 
57 | https://github.com/webmachinelearning/webnn/pull/132
58 | 
59 | #### Proposals for future work
60 | 
61 | Continue discuss proposals submitted for consideration for future work:
62 | 
63 | - Data processing: https://github.com/webmachinelearning/proposals/issues/1 (@WenheLI)
64 | 
65 | - Operation-specific APIs: https://github.com/webmachinelearning/proposals/issues/2 (@jonathanbingham)
66 | 
67 | 


--------------------------------------------------------------------------------
/telcons/2021-02-18-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 18 February 2021 - 15:00-16:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 18 February 2021 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 18 February 2021 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 18 February 2021 <td> 15:00 <td> UTC+0 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 18 February 2021 <td> 16:00 <td> UTC+1 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 18 February 2021 <td> 17:00 <td> UTC+2 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 18 February 2021 <td> 23:00 <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Fri, 19 February 2021 <td> 00:00 <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 18 February 2021 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20210218T15
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
23 | * Minutes: https://www.w3.org/2021/02/18-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### WG Charter update
28 | 
29 | WG Charter update and next steps from W3C Staff: (@dontcallmedom)
30 | 
31 | https://github.com/w3c/machine-learning-charter/
32 | 
33 | #### TAG review feedback
34 | 
35 | Continue resolve (key) TAG review issues, please provide your comments via GH:
36 | 
37 | https://github.com/webmachinelearning/webnn/issues?q=is%3Aissue+is%3Aopen+label%3Atag-tracker
38 | 
39 | Discuss explainer feedback and iterate toward agreement on changes:
40 | 
41 | https://github.com/w3ctag/design-reviews/issues/570#issuecomment-768151707
42 | 
43 | https://github.com/w3ctag/design-reviews/issues/570#issuecomment-768861645
44 | 
45 | #### Privacy IG feedback and the Security and Privacy self-review
46 | 
47 | Discuss the early W3C Privacy Interest Group (PING) feedback and update PR #132 accordingly:
48 | 
49 | https://github.com/webmachinelearning/webnn/issues/119#issuecomment-772565085
50 | 
51 | https://github.com/webmachinelearning/webnn/pull/132
52 | 
53 | #### Operation-specific APIs proposal next steps
54 | 
55 | Gauge interest to incubate this proposal further, prototyping interest? Readiness to establish a dedicated CG repo? (@jonathanbingham, @pyu10055)
56 | 
57 | https://github.com/webmachinelearning/proposals/issues/2
58 | 
59 | #### Next up
60 | 
61 | We expect an update on webnn-native from @huningxin on our next scheduled call on 4th Mar 2021. You may want to start watching:
62 | 
63 | https://github.com/webmachinelearning/webnn-native
64 | 


--------------------------------------------------------------------------------
/telcons/2021-04-01-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 1 April 2021 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 1 April 2021 <td> 07:00 <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 1 April 2021 <td> 10:00 <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 1 April 2021 <td> 15:00 <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 1 April 2021 <td> 16:00 <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 1 April 2021 <td> 17:00 <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 1 April 2021 <td> 22:00 <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 1 April 2021 <td> 23:00 <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 1 April 2021 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20210401T14
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: http://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
23 | * Minutes: https://www.w3.org/2021/04/01-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Operation-specific APIs proposal
28 | 
29 | Background: operation-specific APIs proposal (Ping, Jonathan):
30 | https://github.com/webmachinelearning/proposals/issues/2
31 | 
32 | Discuss and review work-in-progress features that satisfy the requirements of this proposal:
33 | 
34 | - WebAssembly scenario of the op level execution use case (Ningxin):
35 | https://github.com/webmachinelearning/webnn/issues/156
36 | 
37 | - Clarify at which point weights are used in the compilation (Rafael):
38 | https://github.com/webmachinelearning/webnn/issues/157
39 | 
40 | - Discuss how the caller using an op-specific API can resource upload and download (Chai)
41 | 
42 | Any other gaps?
43 | 
44 | #### TAG review feedback - open PRs
45 | 
46 | NamedOutput mechanism clarification 
47 | - issue #140: https://github.com/webmachinelearning/webnn/issues/140
48 | - PR #147: https://github.com/webmachinelearning/webnn/pull/147
49 | 
50 | #### TAG review feedback - open issues without associated PRs
51 | 
52 | Review and provide your comments:
53 | 
54 | - Define a common term for logical tensor changes? https://github.com/webmachinelearning/webnn/issues/150
55 | - Isomorphic JS story, worker scope exposure? https://github.com/webmachinelearning/webnn/issues/142
56 | - Ergonomics of the JS examples https://github.com/webmachinelearning/webnn/issues/139
57 | - String enum for activations https://github.com/webmachinelearning/webnn/issues/138
58 | 
59 | (All open tag-tracker issues without associated PR:
60 | https://github.com/webmachinelearning/webnn/issues?q=is%3Aissue+is%3Aopen+label%3Atag-tracker+-linked%3Apr+)
61 | 
62 | #### Privacy review feedback - normative change proposals
63 | 
64 | Review the proposal to make WebNN API a policy-controlled feature:
65 | https://github.com/webmachinelearning/webnn/issues/145
66 | 


--------------------------------------------------------------------------------
/telcons/2021-04-15-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 15 April 2021 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 15 April 2021 <td> 07:00 <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 15 April 2021 <td> 10:00 <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 15 April 2021 <td> 15:00 <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 15 April 2021 <td> 16:00 <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 15 April 2021 <td> 17:00 <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 15 April 2021 <td> 22:00 <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 15 April 2021 <td> 23:00 <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 15 April 2021 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20210415T14
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
23 | * Minutes: https://www.w3.org/2021/04/15-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Web Machine Learning WG
28 | 
29 | Discuss changes to the proposed Web Machine Learning Working Group charter:
30 | 
31 | https://lists.w3.org/Archives/Public/public-webmachinelearning/2021Apr/0002.html
32 | 
33 | Please contact Dom if any of these changes feel problematic, before April 16.
34 | 
35 | #### Privacy review feedback normative changes
36 | 
37 | [privacy-tracker] Make WebNN API a policy-controlled feature:
38 | https://github.com/webmachinelearning/webnn/issues/145
39 | 
40 | Review PR:
41 | https://github.com/webmachinelearning/webnn/pull/159
42 | 
43 | #### Operation-specific APIs proposal
44 | 
45 | Background: operation-specific APIs proposal (Ping, Jonathan):
46 | https://github.com/webmachinelearning/proposals/issues/2
47 | 
48 | Continue refine work-in-progress features that satisfy the requirements of this proposal:
49 | 
50 | - WebAssembly scenario of the op level execution use case (Ningxin):
51 | https://github.com/webmachinelearning/webnn/issues/156
52 | 
53 | - Clarify at which point weights are used in the compilation (Rafael):
54 | https://github.com/webmachinelearning/webnn/issues/157
55 | 
56 |    - Review the PR to address #157: WIP (Ningxin)
57 | 
58 | - Discuss how the caller using an op-specific API can resource upload and download (Chai)
59 | 
60 | #### TAG review feedback - open issues without associated PRs
61 | 
62 | Review proposed resolutions and provide your comments:
63 | 
64 | - Define a common term for logical tensor changes?
65 |     - Issue: https://github.com/webmachinelearning/webnn/issues/150
66 |     - Proposed resolution: TBD
67 | - Isomorphic JS story, worker scope exposure?
68 |     - Issue: https://github.com/webmachinelearning/webnn/issues/142
69 |     - Proposed resolution: Expose ML interface to Worker in addition to its default Window context
70 | - Ergonomics of the JS examples
71 |     - Issue: https://github.com/webmachinelearning/webnn/issues/139
72 |     - Proposed resolution: Close with a note the primary API consumer is a JS framework
73 | - String enum for activations
74 |     - Issue: https://github.com/webmachinelearning/webnn/issues/138
75 |     - Proposed resolution: Close with a note the current organization of activation functions improves developer ergonomics 
76 | 
77 | 
78 | 


--------------------------------------------------------------------------------
/telcons/2021-05-27-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 27 May 2021 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 27 May 2021 <td> 07:00 <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 27 May 2021 <td> 10:00 <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 27 May 2021 <td> 15:00 <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 27 May 2021 <td> 16:00 <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 27 May 2021 <td> 17:00 <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 27 May 2021 <td> 22:00 <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 27 May 2021 <td> 23:00 <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 27 May 2021 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20210527T14
15 | 
16 | ### Logistics
17 | 
18 | * Chair: Anssi
19 | * Scribe: ?
20 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
21 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
22 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
23 | * Minutes: https://www.w3.org/2021/05/27-webmachinelearning-minutes.html
24 | 
25 | ### Agenda
26 | 
27 | #### Security and Privacy
28 | 
29 | Review and discuss initial Security and Privacy Considerations:
30 | 
31 |  - Issue: https://github.com/webmachinelearning/webnn/issues/122
32 |  - PR: https://github.com/webmachinelearning/webnn/pull/170
33 |  - Related issues:
34 |     - [privacy-tracker] Self-Review Questionnaire: https://github.com/webmachinelearning/webnn/issues/119 
35 |     - [privacy-tracker] Fingerprinting via matmul: https://github.com/webmachinelearning/webnn/issues/85
36 | 
37 | Discuss related WebGPU/GL Security and Privacy Considerations, applicability in this context:
38 |  - WebGPU: https://gpuweb.github.io/gpuweb/#security
39 |  - WebGL: https://www.khronos.org/registry/webgl/specs/latest/1.0/#4
40 | 
41 | 
42 | #### Operation-specific APIs proposal
43 | 
44 | Continue progress remaining features that satisfy the requirements of the operation-specific APIs proposal:
45 | 
46 | https://github.com/webmachinelearning/proposals/issues/2
47 | 
48 | Discuss design considerations and review proposed solutions to enable both efficient graph execution and imperative eager execution with a cohesive WebNN API. Related topics:
49 | 
50 | - Support download data asynchronously
51 |     - PR: WIP https://github.com/webmachinelearning/webnn/pull/166
52 | - Layout format conversion
53 |     - Issue: WIP
54 | - Clarify at which point weights are used in the compilation
55 |     - Issue: https://github.com/webmachinelearning/webnn/issues/157
56 | - AI accelerator device selection
57 |     - Issue: https://github.com/webmachinelearning/webnn/issues/169
58 | 
59 | #### Model Loader API update
60 | 
61 | - Discuss Model Loader API Chrome OS Origin Trial plans (Jonathan)
62 | - Discuss Wasm runner for TF Lite and how it could possibly be used for WebNN benchmarking (Jonathan)
63 | 
64 | #### TAG review
65 | 
66 | - [tag-tracker] Define a common term for logical tensor changes?
67 |     - Issue: https://github.com/webmachinelearning/webnn/issues/150
68 |     - PR: WIP
69 | - [tag-tracker] Ergonomics of the JS examples
70 |     - Issue: https://github.com/webmachinelearning/webnn/issues/139
71 |     - Resolution: WIP - pending concrete TAG feedback to inform the resolution
72 | - [tag-tracker] String enum for activations
73 |     - Issue: https://github.com/webmachinelearning/webnn/issues/138
74 |     - Proposed resolution: WIP - review informative note to be incorporated into the spec to explain current design principles around activations
75 | 
76 | 
77 | 


--------------------------------------------------------------------------------
/telcons/2021-11-11-agenda.md:
--------------------------------------------------------------------------------
1 | # WebML WG Teleconference – 11 November 2021 - 15:00-16:00 UTC+0
2 | 
3 | :warning: This meeting is cancelled to avoid overlap with [Remembrance Day](https://en.wikipedia.org/wiki/Remembrance_Day).
4 | 


--------------------------------------------------------------------------------
/telcons/2021-12-16-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 16 December 2021 - 15:00-16:00 UTC+0
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 16 December 2021 <td> 07:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 16 December 2021 <td> 10:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 16 December 2021 <td> 15:00 <td> UTC+0 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 16 December 2021 <td> 16:00 <td> UTC+1 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 16 December 2021 <td> 17:00 <td> UTC+2 hours
10 | <tr><td> Shanghai (China) <td> Thu, 16 December 2021 <td> 23:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Fri, 17 December 2021 <td> 00:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 16 December 2021 <td colspan=2> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20211216T15
16 |   </details>
17 |   
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? ([howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md))
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2021/12/16-webmachinelearning-minutes.html
26 |   
27 | ---
28 | 
29 | ## Agenda
30 |   
31 | ### ℹ️ Announcements
32 |   
33 | Given the recent progress with the Model Loader API, we plan to restart WebML Community Group meetings early 2022, Sydney friendly timeslot, respond to the poll by 15 Dec 2021: https://doodle.com/poll/arstwzhv6ew6t36p
34 |   
35 | Model Loader API Intent to Prototype in Chromium: https://groups.google.com/a/chromium.org/g/blink-dev/c/w7Oq2_x2D8U/m/8BhYLGCIBwAJ
36 |   
37 | ### ℹ️ TPAC meeting follow-up (cont'd)
38 | 
39 | We had a broad range of exciting discussion at TPAC with guest appearances. Thank you all for your contributions!
40 | 
41 | Let's reflect on our TPAC meeting learnings and discuss what next steps the group should take informed by these new insights and perspectives.
42 | 
43 | TPAC agenda: https://github.com/webmachinelearning/meetings/issues/18
44 | 
45 | TPAC minutes: [Day 1](https://www.w3.org/2021/10/26-webmachinelearning-minutes.html), [Day 2](https://www.w3.org/2021/10/27-webmachinelearning-minutes.html), [Day 3](https://www.w3.org/2021/10/28-webmachinelearning-minutes.html)
46 | 
47 | #### ℹ️ Ethical issues in using Machine Learning on the Web
48 |   
49 | - Introduce James Fletcher, BBC Lead, Responsible AI/ML
50 | - Initial thoughts and draft structure for 2022 Ethics work
51 | - Comments and feedback
52 | - Call to connect with experts in ethics to contribute to this work
53 | - See [presentation slides](https://lists.w3.org/Archives/Public/www-archive/2021Dec/att-0004/1216_W3C_Ethical_WebML_Intro.pdf)
54 | 
55 |   
56 | ### ℹ️ WebNN issues review
57 | 
58 | - Integration with real-time video processing https://github.com/webmachinelearning/webnn/issues/226
59 | - Should restrict the sync APIs to only exist in Workers? https://github.com/webmachinelearning/webnn/issues/229
60 | - Should WebNN support async APIs? https://github.com/webmachinelearning/webnn/issues/230
61 | - Path to Candidate Recommendation
62 |    - CR readiness tracker https://github.com/webmachinelearning/webnn/issues/240
63 |    - Wide review tracker https://github.com/webmachinelearning/webnn/issues/239
64 | 


--------------------------------------------------------------------------------
/telcons/2022-01-12-cg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML CG Teleconference – 12 January 2022 - 05:00-06:00 UTC+0
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Tue, 11 January 2022 <td> 21:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Wed, 12 January 2022 <td> 00:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Wed, 12 January 2022 <td> 05:00 <td> UTC+0 hours
 8 | <tr><td> Berlin (Germany) <td> Wed, 12 January 2022 <td> 06:00 <td> UTC+1 hours
 9 | <tr><td> Helsinki (Finland) <td> Wed, 12 January 2022 <td> 07:00 <td> UTC+2 hours
10 | <tr><td> Shanghai (China) <td> Wed, 12 January 2022 <td> 13:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Wed, 13 January 2022 <td> 14:00 <td> UTC+9 hours
12 | <tr><td> Sydney (Australia) <td> Wed, 12 January 2022 <td> 16:00 <td> UTC+11 hours
13 | <tr><td> Corresponding UTC (GMT) <td> Wed, 12 January 2022 <td colspan=2> 05:00 UTC
14 | </table>
15 | 
16 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20220112T05
17 |   </details>
18 |   
19 | ## Logistics
20 | 
21 | * Chair: Anssi
22 | * Scribe: ? ([howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md))
23 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
24 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
25 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
26 | * Minutes: https://www.w3.org/2022/01/12-webmachinelearning-minutes.html
27 |   
28 | ---
29 | 
30 | ## Agenda
31 | 
32 | We're restarting these WebML Community Group (CG) teleconferences to discuss the Model Loader API incubation.
33 |   
34 | We expect this Community Group (CG) to collaborate closely with the WebML Working Group (WG) currently focused on standardizing the WebNN API. Model Loader-WebNN API coordination topics will be discussed on the CG teleconferences primarily.
35 | 
36 | #### ℹ️ Introductions
37 | 
38 | #### ℹ️ Model Loader API
39 |   
40 | * Spec and implementation progress
41 |   * [Updated explainer](https://github.com/webmachinelearning/model-loader/blob/main/explainer.md) and [early spec draft](https://webmachinelearning.github.io/model-loader/)
42 |   * [Chromium prototype](https://chromium-review.googlesource.com/c/chromium/src/+/3341136)
43 | * Open questions
44 | * Dependencies, coordination topics with WebNN API
45 |   
46 | #### ℹ️ Meeting cadence
47 | 


--------------------------------------------------------------------------------
/telcons/2022-01-13-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 13 January 2022 - 15:00-16:00 UTC+0
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 13 January 2022 <td> 07:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 13 January 2022 <td> 10:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 13 January 2022 <td> 15:00 <td> UTC+0 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 13 January 2022 <td> 16:00 <td> UTC+1 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 13 January 2022 <td> 17:00 <td> UTC+2 hours
10 | <tr><td> Shanghai (China) <td> Thu, 13 January 2022 <td> 23:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Fri, 14 January 2022 <td> 00:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 13 January 2022 <td colspan=2> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20220113T15
16 |   </details>
17 |   
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2022/01/13-webmachinelearning-minutes.html
26 |   
27 | ---
28 | 
29 | ## Agenda
30 | 
31 | #### ℹ️ WebNN API feature and change requests
32 | 
33 | Follow up on work-in-progress requests:
34 | 
35 | - Integration with real-time video processing
36 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/226
37 |    - Status: proof-of-concept work planned for Q1, please provide your feedback in the issue
38 | 
39 | - Should restrict the sync APIs to only exist in Workers?
40 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/229
41 |    - Status: needs resolution
42 | 
43 | - Should WebNN support async APIs?
44 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/230
45 |    - Status: needs resolution
46 | 
47 | #### ℹ️ WebNN API open pull requests
48 | 
49 | Discuss/resolve issues in open PRs pending reviews:
50 | 
51 | - Open PRs: https://github.com/webmachinelearning/webnn/pulls
52 | 
53 | #### ℹ️ WebNN API Candidate Recommendation
54 | 
55 | - CR readiness tracker: https://github.com/webmachinelearning/webnn/issues/240
56 | - Wide review tracker: https://github.com/webmachinelearning/webnn/issues/239
57 | - CR blockers (WIP): https://github.com/webmachinelearning/webnn/issues?q=label%3Acr
58 | 
59 | - Status: review the CR plan, continue CR issue triage
60 |   
61 | #### ℹ️ Security review
62 |   
63 | Discuss security review feedback from Chrome Security team:
64 | 
65 | - General Security Questions: https://github.com/webmachinelearning/webnn/issues/241
66 | - Guidelines/philosophy for new operations, including security principles: https://github.com/webmachinelearning/webnn/issues/242
67 | - Op metadata that helps avoid implementation mistakes: https://github.com/webmachinelearning/webnn/issues/243
68 | - A conformance suite with disallowed intra-op examples would be helpful for hardening: https://github.com/webmachinelearning/webnn/issues/244
69 | 


--------------------------------------------------------------------------------
/telcons/2022-01-26-cg-agenda.md:
--------------------------------------------------------------------------------
1 | # WebML CG Teleconference – 26 January 2022 - 05:00-06:00 UTC+0
2 | 
3 | We're rescheduling to [9 February 2022](https://github.com/webmachinelearning/meetings/blob/main/telcons/2022-02-09-cg-agenda.md) to allow our Australian friends enjoy [Australia Day](https://en.wikipedia.org/wiki/Australia_Day)! 🇦🇺
4 | 


--------------------------------------------------------------------------------
/telcons/2022-01-27-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 27 January 2022 - 15:00-16:00 UTC+0
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 27 January 2022 <td> 07:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 27 January 2022 <td> 10:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 27 January 2022 <td> 15:00 <td> UTC+0 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 27 January 2022 <td> 16:00 <td> UTC+1 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 27 January 2022 <td> 17:00 <td> UTC+2 hours
10 | <tr><td> Shanghai (China) <td> Thu, 27 January 2022 <td> 23:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Fri, 28 January 2022 <td> 00:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 27 January 2022 <td colspan=2> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20220127T15
16 |   </details>
17 |   
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2022/01/27-webmachinelearning-minutes.html
26 |   
27 | ---
28 | 
29 | ## Agenda
30 |   
31 | #### ℹ️ Security review
32 |   
33 | >* Discuss security review feedback from Chrome Security team
34 | 
35 |    - [ ] issue: General Security Questions: https://github.com/webmachinelearning/webnn/issues/241
36 |    - [ ] issue: Guidelines/philosophy for new operations, including security principles: https://github.com/webmachinelearning/webnn/issues/242
37 |    - [ ] issue: Op metadata that helps avoid implementation mistakes: https://github.com/webmachinelearning/webnn/issues/243
38 |    - [ ] issue: A conformance suite with disallowed intra-op examples would be helpful for hardening: https://github.com/webmachinelearning/webnn/issues/244
39 | 
40 | #### ℹ️ Integration with real-time video processing
41 | 
42 | >* Review proposed prototype next steps
43 | >* Review proposed use case for spec inclusion
44 | >* Discuss progress of feature work in other specs related to the media pipeline: mediacapture-transform, WebGPU, WebCodecs (see issue #226 for pointers)
45 | 
46 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/226
47 |    - [ ] proposed prototype next steps: https://github.com/webmachinelearning/webnn/issues/226#issuecomment-1016104142
48 |    - [ ] proposed use case PR: https://github.com/webmachinelearning/webnn/pull/249
49 | 
50 | #### ℹ️ Double-precision baseline implementation of WebNN operations for testing
51 | 
52 | >* Review the double-precision baseline implementation of WebNN operations for web-platform-tests purposes
53 | 
54 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/245
55 |    - [ ] staging repo: https://github.com/huningxin/webnn-baseline
56 | 
57 | #### ℹ️ WebNN API Candidate Recommendation
58 | 
59 | >* Time permitting, check our progress against the CR milestone
60 | 
61 | - CR readiness tracker: https://github.com/webmachinelearning/webnn/issues/240
62 | - Wide review tracker: https://github.com/webmachinelearning/webnn/issues/239
63 | - CR blockers (WIP): https://github.com/webmachinelearning/webnn/issues?q=label%3Acr
64 | 


--------------------------------------------------------------------------------
/telcons/2022-02-09-cg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML CG Teleconference – 9 February 2022 - 05:00-06:00 UTC+0
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Tue, 8 February 2022 <td> 21:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Wed, 9 February 2022 <td> 00:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Wed, 9 February 2022 <td> 05:00 <td> UTC+0 hours
 8 | <tr><td> Berlin (Germany) <td> Wed, 9 February 2022 <td> 06:00 <td> UTC+1 hours
 9 | <tr><td> Helsinki (Finland) <td> Wed, 9 February 2022 <td> 07:00 <td> UTC+2 hours
10 | <tr><td> Shanghai (China) <td> Wed, 9 February 2022 <td> 13:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Wed, 9 February 2022 <td> 14:00 <td> UTC+9 hours
12 | <tr><td> Sydney (Australia) <td> Wed, 9 February 2022 <td> 16:00 <td> UTC+11 hours
13 | <tr><td> Corresponding UTC (GMT) <td> Wed, 9 February 2022 <td colspan=2> 05:00 UTC
14 | </table>
15 | 
16 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20220209T05
17 |   </details>
18 |   
19 | ## Logistics
20 | 
21 | * Chair: Anssi
22 | * Scribe: ? ([howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md))
23 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
24 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
25 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
26 | * Minutes: https://www.w3.org/2022/02/09-webmachinelearning-minutes.html
27 |   
28 | ---
29 | 
30 | ## Agenda
31 | 
32 | ###  🔄 Model Loader API
33 | 
34 | #### ℹ️ Benchmark discussion
35 | 
36 | - Continued from 12 Jan meeting: open questions, proposed next steps
37 | 
38 | #### ℹ️ Hardware acceleration
39 | 
40 | - What will the interface look like when hardware acceleration is supported?
41 |   - What if the data of a tensor is on GPU/TPU?
42 |   - Shall we let users know which ML accelerator is available on the platform? (Performance v.s. Privacy)
43 | 
44 | #### ℹ️ Versioning
45 | 
46 | - How to version the API? (supported ops and formats)
47 | 
48 | #### ℹ️ Streaming inputs
49 | 
50 | - Dedicated interface for streaming inputs (e.g. buffered inputs for video, audio)?
51 | 
52 | #### ℹ️ Model format
53 | 
54 | - Prototype vs standard format requirements
55 | - "The Model Loader API needs a standard format supported across browsers and devices for broad interoperability." condition for standards adoption into WebML WG
56 | 
57 | #### ℹ️ References
58 |   * [WebML CG Teleconference – 12 Jan 2022 minutes](https://www.w3.org/2022/01/12-webmachinelearning-minutes.html)
59 |   * [Updated explainer](https://github.com/webmachinelearning/model-loader/blob/main/explainer.md) and [early spec draft](https://webmachinelearning.github.io/model-loader/)
60 |   * [Chromium prototype](https://chromium-review.googlesource.com/c/chromium/src/+/3341136)
61 |   
62 | 


--------------------------------------------------------------------------------
/telcons/2022-02-10-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 10 February 2022 - 15:00-16:00 UTC+0
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 10 February 2022 <td> 07:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 10 February 2022 <td> 10:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 10 February 2022 <td> 15:00 <td> UTC+0 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 10 February 2022 <td> 16:00 <td> UTC+1 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 10 February 2022 <td> 17:00 <td> UTC+2 hours
10 | <tr><td> Shanghai (China) <td> Thu, 10 February 2022 <td> 23:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Fri, 11 February 2022 <td> 00:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 10 February 2022 <td colspan=2> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20220210T15
16 |   </details>
17 |   
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2022/02/10-webmachinelearning-minutes.html
26 | 
27 | 
28 | 
29 | ---
30 | 
31 | #### ℹ️ Security review response
32 | 
33 | >* Review the group's proposed response to security questions from Chrome Security team. Please record your feedback and comments in the issues and/or draft PR. Agree on the next step that is an "official" Security wide review.
34 | 
35 |    - [ ] issue: General Security Questions: https://github.com/webmachinelearning/webnn/issues/241
36 |    - [ ] PR: Update Security Considerations per review feedback: https://github.com/webmachinelearning/webnn/pull/251
37 |    - [ ] All security-tracker issues: https://github.com/webmachinelearning/webnn/issues?q=label%3Asecurity-tracker+
38 | 
39 | #### ℹ️ Privacy review refresh
40 | 
41 | >* Initial privacy review was completed a year ago. Discuss/identify any privacy-impacting changes since the initial review that need PING's attention.
42 | 
43 |    - [x] issue: Initial Privacy Review: https://github.com/webmachinelearning/webnn/issues/119
44 |    - [x] spec: Privacy Considerations: https://www.w3.org/TR/webnn/#privacy
45 |    - [ ] All privacy-tracker issues: https://github.com/webmachinelearning/webnn/issues?q=label%3Aprivacy-tracker+
46 | 
47 | #### ℹ️ Double-precision baseline implementation of WebNN operations for testing
48 | 
49 | >* Review proposed baseline implementation of WebNN operations ("webnn-baseline") used for demonstrating interop. Implements 41 WebNN ops for first-wave models in JS double-precision Number and calculations without depending on 3rd party libs.
50 | 
51 |    - [ ] PR: WebNN baseline implementation for first-wave ops https://github.com/webmachinelearning/webnn-baseline/pull/1
52 |    - [x] GH repo (webnn-baseline is a CG deliverable): https://github.com/webmachinelearning/webnn-baseline
53 | 
54 | 
55 | #### ℹ️ WebNN API open pull requests
56 | 
57 | >* Discuss/resolve issues in open PRs pending reviews:
58 | 
59 |    - Open PRs: https://github.com/webmachinelearning/webnn/pulls
60 | 


--------------------------------------------------------------------------------
/telcons/2022-02-24-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 24 February 2022 - 15:00-16:00 UTC+0
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 24 February 2022 <td> 07:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 24 February 2022 <td> 10:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 24 February 2022 <td> 15:00 <td> UTC+0 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 24 February 2022 <td> 16:00 <td> UTC+1 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 24 February 2022 <td> 17:00 <td> UTC+2 hours
10 | <tr><td> Shanghai (China) <td> Thu, 24 February 2022 <td> 23:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Fri, 25 February 2022 <td> 00:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 24 February 2022 <td colspan=2> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20220224T15
16 |   </details>
17 |   
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2022/02/24-webmachinelearning-minutes.html
26 | 
27 | 
28 | 
29 | ---
30 | 
31 | #### ℹ️ Announcements
32 | 
33 | - Survey about TPAC 2022
34 | - Update on Ethical Considerations
35 | 
36 | #### ℹ️ Security review response ready for review
37 | 
38 | >* Review the group's response to security questions from Chrome Security team. Please record your feedback and comments in the issues and/or draft PR. Agree on the next step that is an "official" Security wide review.
39 | 
40 | - [ ] issue: General Security Questions: https://github.com/webmachinelearning/webnn/issues/241
41 | - [ ] PR: Update Security Considerations per review feedback: https://github.com/webmachinelearning/webnn/pull/251
42 | - [ ] All security-tracker issues: https://github.com/webmachinelearning/webnn/issues?q=label%3Asecurity-tracker+
43 | 
44 | #### ℹ️ Candidate Recommendation technical scope
45 | 
46 | CR expectation is that the spec is considered technically complete and fit for purpose. Review the current open and proposed CR issues targeting this milestone. Discuss any blockers or new information that might suggest we need to adjust our scope and/or timeline.
47 | 
48 | Current open CR issues:
49 | - [ ] The baseline implementation of WebNN ops https://github.com/webmachinelearning/webnn/issues/245
50 | - [ ] Should WebNN support async APIs? https://github.com/webmachinelearning/webnn/issues/230
51 | - [ ] Should restrict the sync APIs to only exist in Workers? https://github.com/webmachinelearning/webnn/issues/229
52 | - [ ] Integration with real-time video processing https://github.com/webmachinelearning/webnn/issues/226
53 | - [ ] Add method steps to operations https://github.com/webmachinelearning/webnn/issues/210
54 | 
55 | (All "cr" issues https://github.com/webmachinelearning/webnn/labels/cr)
56 | 
57 | Proposed CR issues:
58 | - [ ] WebNN should support int8 quantized models https://github.com/webmachinelearning/webnn/issues/128
59 | - [ ] WebNN / WebGPU interop (currently tracked as part of #226) https://github.com/gpuweb/gpuweb/issues/2500
60 | 
61 | #### ℹ️ WebNN API open pull requests
62 | 
63 | >* Discuss/resolve issues in open PRs pending reviews:
64 | 
65 |    - Open PRs: https://github.com/webmachinelearning/webnn/pulls
66 | 


--------------------------------------------------------------------------------
/telcons/2022-03-09-cg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML CG Teleconference – 9 March 2022 - 05:00-06:00 UTC+0
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Tue, 8 March 2022 <td> 21:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Wed, 9 March 2022 <td> 00:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Wed, 9 March 2022 <td> 05:00 <td> UTC+0 hours
 8 | <tr><td> Berlin (Germany) <td> Wed, 9 March 2022 <td> 06:00 <td> UTC+1 hours
 9 | <tr><td> Helsinki (Finland) <td> Wed, 9 March 2022 <td> 07:00 <td> UTC+2 hours
10 | <tr><td> Shanghai (China) <td> Wed, 9 March 2022 <td> 13:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Wed, 9 March 2022 <td> 14:00 <td> UTC+9 hours
12 | <tr><td> Sydney (Australia) <td> Wed, 9 March 2022 <td> 16:00 <td> UTC+11 hours
13 | <tr><td> Corresponding UTC (GMT) <td> Wed, 9 March 2022 <td colspan=2> 05:00 UTC
14 | </table>
15 | 
16 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20220309T05
17 |   </details>
18 |   
19 | ## Logistics
20 | 
21 | * Chair: Anssi
22 | * Scribe: ? ([howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md))
23 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
24 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
25 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
26 | * Minutes: https://www.w3.org/2022/03/09-webmachinelearning-minutes.html
27 |   
28 | ---
29 | 
30 | ## Agenda
31 | 
32 | ###  🔄 Model Loader API
33 | 
34 | #### ℹ️ Versioning
35 | 
36 | - How to version the API? (supported ops and formats)
37 | 
38 | #### ℹ️ Streaming inputs
39 | 
40 | - Dedicated interface for streaming inputs (e.g. buffered inputs for video, audio)?
41 | 
42 | #### ℹ️ Model format
43 | 
44 | - Prototype vs standard format requirements
45 | - "The Model Loader API needs a standard format supported across browsers and devices for broad interoperability." condition for standards adoption into WebML WG
46 | 
47 | #### ℹ️ Support for non-IEEE 754 float point types
48 | 
49 | - [ ] issue: Open question: support for non-IEEE 754 float point types #23 https://github.com/webmachinelearning/model-loader/issues/23
50 | 
51 | Some accelerators use non-standard float point types (e.g. bfloat16 and TF32) for performance, memory usage optimization. Discuss Model Loader experimentation with these types, open questions:
52 | 
53 | - Could WebNN and Model Loader API use a common approach for these types?
54 | - How to represent, manipulate these types in JavaScript? Or auto-convert at runtime with some heuristic?
55 | - How to have the two specs complement each other?
56 | - WebNN provides general purpose feature that runs everywhere (write once, run everywhere), Model Loader optionally makes use of accelerator specific features, and *could* run extremely efficiently based on the model?
57 | - How much information would this give away, privacy risk?
58 | 
59 | #### ℹ️ References
60 |   * [WebML CG Teleconference – 9 Feb 2022 minutes](https://www.w3.org/2022/02/09-webmachinelearning-minutes.html)
61 |   * [WebML CG Teleconference – 12 Jan 2022 minutes](https://www.w3.org/2022/01/12-webmachinelearning-minutes.html)
62 |   * [Model Loader API explainer](https://github.com/webmachinelearning/model-loader/blob/main/explainer.md) and [early spec draft](https://webmachinelearning.github.io/model-loader/)
63 |   * [Model Loader API Chromium prototype](https://chromium-review.googlesource.com/c/chromium/src/+/3341136)
64 |   
65 | 


--------------------------------------------------------------------------------
/telcons/2022-04-07-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 7 April 2022 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 7 April 2022 <td> 07:00 <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 7 April 2022 <td> 10:00 <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 7 April 2022 <td> 15:00 <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 7 April 2022 <td> 16:00 <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 7 April 2022 <td> 17:00 <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 7 April 2022 <td> 22:00 <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 7 April 2022 <td> 23:00 <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 7 April 2022 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20220407T14
15 | 
16 |   </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2022/04/07-webmachinelearning-minutes.html
26 | 
27 | #### ℹ️ Ethical Web Marchine Learning
28 | 
29 | - Early report from the review & brainstorm sessions held 4/5 Apr
30 | - Background video: https://www.w3.org/2022/03/ml-ethics/
31 |   
32 | #### ℹ️ Proposed new use cases
33 | 
34 | > Discuss and review the proposed new use cases.
35 | 
36 | - [ ] WebNN API in gaming scenarios (for discussion only)
37 |    - A guest speaker from Unity to share a perspective on leveraging WebNN API in a web backend of a game engine.
38 | - [ ] Performance Adaptation https://github.com/webmachinelearning/webnn/pull/207
39 | - [ ] Ethical Content Filtering https://github.com/webmachinelearning/webnn/pull/253
40 | 
41 | #### ℹ️ Context-based graph execution methods for different threading models
42 | 
43 | > Review the new proposal for context-based graph execution methods that is an improved design and considers feedback received from the group for the earlier proposal.
44 | 
45 | - [ ] issue: Should WebNN support async APIs? https://github.com/webmachinelearning/webnn/issues/230
46 | - [ ] PR: Context-based graph execution methods for different threading models https://github.com/webmachinelearning/webnn/pull/257
47 | 
48 | #### ℹ️ Candidate Recommendation readiness
49 | 
50 | >Discuss and review our progress toward WebNN API Candidate Recommendation.
51 | 
52 | - Use cases - scope, do we meet the requirements
53 | - CR technical scope - check agreement on [triaged CR issues](https://github.com/webmachinelearning/webnn/labels/cr) 
54 | - Ethical considerations - workshop organized, plan to integrate into spec
55 | - Implementation experience: w-p-t test suite, webnn-baseline, Chromium implementation
56 | - Wide review: :white_check_mark: TAG done, :white_check_mark: Security (HR in flight), :black_square_button: Privacy _refresh_, :black_square_button: a12y, :black_square_button: i18n, :black_square_button: WebGPU WG https://github.com/webmachinelearning/webnn/issues/239
57 | 
58 | #### ℹ️ Open issues and pull requests
59 | 
60 | > Discuss/resolve open issues and review open PRs pending review, time permitting.
61 | 
62 | - Open issues: https://github.com/webmachinelearning/webnn/issues
63 | - Open PRs: https://github.com/webmachinelearning/webnn/pulls
64 | 


--------------------------------------------------------------------------------
/telcons/2022-04-21-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 21 April 2022 - 14:00-15:00 UTC+0
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 21 April 2022 <td> 07:00 <td> UTC-7 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 21 April 2022 <td> 10:00 <td> UTC-4 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 21 April 2022 <td> 15:00 <td> UTC+1 hours
 7 | <tr><td> Berlin (Germany) <td> Thu, 21 April 2022 <td> 16:00 <td> UTC+2 hours
 8 | <tr><td> Helsinki (Finland) <td> Thu, 21 April 2022 <td> 17:00 <td> UTC+3 hours
 9 | <tr><td> Shanghai (China) <td> Thu, 21 April 2022 <td> 22:00 <td> UTC+8 hours
10 | <tr><td> Tokyo (Japan) <td> Thu, 21 April 2022 <td> 23:00 <td> UTC+9 hours
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 21 April 2022 <td colspan=2> 14:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20220421T14
15 | 
16 |   </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2022/04/21-webmachinelearning-minutes.html
26 | 
27 | #### ℹ️ Announcements
28 | 
29 | - TPAC 2022 Hybrid meeting announcement https://lists.w3.org/Archives/Public/public-webmachinelearning/2022Apr/0003.html
30 | - WebML and WebNN logos https://webmachinelearning.github.io/logos
31 | 
32 | #### ℹ️ Proposed new use cases
33 | 
34 | > Discuss and review the proposed new use cases.
35 | 
36 | - [ ] Performance Adaptation https://github.com/webmachinelearning/webnn/pull/207
37 | - [ ] Ethical Content Filtering https://github.com/webmachinelearning/webnn/pull/253
38 | 
39 | #### ℹ️ Context-based graph execution methods for different threading models
40 | 
41 | > Review the new proposal for context-based graph execution methods that is an improved design and considers feedback received from the group for the earlier proposal.
42 | 
43 | - [ ] issue: Should WebNN support async APIs? https://github.com/webmachinelearning/webnn/issues/230
44 | - [ ] PR: Context-based graph execution methods for different threading models https://github.com/webmachinelearning/webnn/pull/257
45 | 
46 | #### ℹ️ Candidate Recommendation maturity wide review expectations
47 | 
48 | >Discuss wide review expectations for the WebNN API Candidate Recommendation (CR). Work out a plan for Accessibility, Internationalization, WebGPU WG reviews.
49 | 
50 | - Accessibility checklist https://w3c.github.io/apa/fast/checklist.html
51 | - Internationalization checklist https://w3c.github.io/i18n-drafts/techniques/shortchecklist
52 | - WebGPU interop investigation https://github.com/gpuweb/gpuweb/issues/2500
53 |   
54 | See also:
55 | 
56 | - CR readiness tracker https://github.com/webmachinelearning/webnn/issues/240
57 | - Wide review tracker https://github.com/webmachinelearning/webnn/issues/239
58 | 
59 | #### ℹ️ Open issues and pull requests
60 | 
61 | > Discuss/resolve open issues and review open PRs pending review, time permitting.
62 | 
63 | - Open issues: https://github.com/webmachinelearning/webnn/issues
64 | - Open PRs: https://github.com/webmachinelearning/webnn/pulls
65 | 


--------------------------------------------------------------------------------
/telcons/2022-11-17-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 17 November 2022 - 15:00-16:00 UTC
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 17 November 2022 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 17 November 2022 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 17 November 2022 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 7 | <tr><td> Berlin (Germany) <td> Thu, 17 November 2022 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 8 | <tr><td> Helsinki (Finland) <td> Thu, 17 November 2022 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
 9 | <tr><td> Shanghai (China) <td> Thu, 17 November 2022 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
10 | <tr><td> Tokyo (Japan) <td> Fri, 17 November 2022 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 17 November 2022 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20221117T15
15 | 
16 |   </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2022/11/17-webmachinelearning-minutes.html
26 | 
27 | 
28 | #### ℹ️ XLA, OpenXLA Project, StableHLO (invited speaker Stella Laurenzo)
29 | 
30 | >An invited speaker Stella Laurenzo from Google's ML/Compilers Applied Research group gives a talk about the XLA compiler, OpenXLA Project and StableHLO op set. Presentation followed by a Q&A and discussion to identify any areas of joint interest that might warrant further work to unlock synergies.
31 | 
32 | - XLA compiler: https://www.tensorflow.org/xla
33 | - XLA Project: https://github.com/openxla
34 | - StableHLO op set: https://github.com/openxla/stablehlo
35 | 
36 | #### ℹ️ WebML WG Charter 2023-2025 under development
37 | 
38 | >Web Machine Learning Working Group Charter for 2023-2025 is now under development. Please review the draft PR and open issues, provide your comments and open new issues as appropriate to help shape the WG's technical scope. 
39 | 
40 | - Announcement https://lists.w3.org/Archives/Public/public-webmachinelearning-wg/2022Oct/0002.html
41 | - Charter PR https://github.com/w3c/machine-learning-charter/pull/19
42 | - Open issues https://github.com/w3c/machine-learning-charter/issues
43 |   - Features deferred to WebNN v2
44 |   - Dedicated ML hardware accelerators: NPU, VPU, xPU
45 |   - Set of ops supported must be more comprehensive
46 |   - Level of abstraction for neural net operations
47 |   - WebRTC coordination
48 |   - WebGPU interoperability
49 |   - On-device training
50 | 


--------------------------------------------------------------------------------
/telcons/2022-12-01-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 1 December 2022 - 15:00-16:00 UTC
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 1 December 2022 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 1 December 2022 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 1 December 2022 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 7 | <tr><td> Berlin (Germany) <td> Thu, 1 December 2022 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 8 | <tr><td> Helsinki (Finland) <td> Thu, 1 December 2022 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
 9 | <tr><td> Shanghai (China) <td> Thu, 1 December 2022 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
10 | <tr><td> Tokyo (Japan) <td> Fri, 1 December 2022 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 1 December 2022 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20221201T15
15 | 
16 |   </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2022/12/01-webmachinelearning-minutes.html
26 | 
27 | 
28 | #### ℹ️ WebNN API Candidate Recommendation open issues
29 | 
30 | >Review and discuss the current CR issues, work out a plan to address the issues prior to the expected CR publication.
31 | 
32 | - Current CR issues https://github.com/webmachinelearning/webnn/labels/cr
33 | 
34 | - Web platform tests
35 |   - [ ] tracker issue: https://github.com/webmachinelearning/webnn/issues/265
36 |   - Open issues: https://github.com/webmachinelearning/webnn-baseline/pulls
37 |   - Open PRs: https://github.com/webmachinelearning/webnn-baseline/pulls
38 |   - ➡️ Status: Discuss & resolve any blockers for w-p-t & webnn-baseline reference impl.
39 | 
40 | - Use modern WebIDL conventions
41 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/210
42 |   - ➡️ Status: Identify spec sections where editorial work is required to add algorithmic method steps, add internal slots, use getters, setters, define methods and constructors using the modern WebIDL conventions.
43 | 
44 | 
45 | #### ℹ️ WebML WG Charter 2023-2025 under development
46 | 
47 | >Web Machine Learning Working Group Charter for 2023-2025 is now under development. Please review the draft PR and open issues, provide your comments and open new issues as appropriate to help shape the WG's technical scope. 
48 | 
49 | - Announcement https://lists.w3.org/Archives/Public/public-webmachinelearning-wg/2022Oct/0002.html
50 | - Charter PR https://github.com/w3c/machine-learning-charter/pull/19
51 | - Open issues https://github.com/w3c/machine-learning-charter/issues
52 |   - OpenXLA coordination
53 |   - On-device training
54 |   - Features deferred to WebNN v2
55 |   - Dedicated ML hardware accelerators: NPU, VPU, xPU
56 |   - Set of ops supported must be more comprehensive
57 |   - Level of abstraction for neural net operations
58 |   - WebRTC coordination
59 |   - WebGPU interoperability
60 | 


--------------------------------------------------------------------------------
/telcons/2022-12-15-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 15 December 2022 - 15:00-16:00 UTC
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 15 December 2022 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 15 December 2022 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 15 December 2022 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 7 | <tr><td> Berlin (Germany) <td> Thu, 15 December 2022 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 8 | <tr><td> Helsinki (Finland) <td> Thu, 15 December 2022 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
 9 | <tr><td> Shanghai (China) <td> Thu, 15 December 2022 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
10 | <tr><td> Tokyo (Japan) <td> Fri, 15 December 2022 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 15 December 2022 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20221215T15
15 | 
16 |   </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2022/12/15-webmachinelearning-minutes.html
26 | 
27 | 
28 | #### ℹ️ WebNN API Candidate Recommendation open issues
29 | 
30 | >Review and discuss the current CR issues, work out a plan to address the key issues prior to the expected CR publication. For WPT, we need to show how adequate implementation experience will be demonstrated. For WebIDL conventions, we need to demonstrate the spec is considered complete and fit for purpose.
31 | 
32 | - Current CR issues https://github.com/webmachinelearning/webnn/labels/cr
33 | 
34 | - Web platform tests
35 |   - [ ] tracker issue: https://github.com/webmachinelearning/webnn/issues/265
36 |   - Open webnn-baseline PRs: https://github.com/webmachinelearning/webnn-baseline/pulls
37 |   - Open wpt PRs: https://github.com/web-platform-tests/wpt/labels/wg-s_webmachinelearning
38 |   - ➡️ Status: Discuss & resolve any blockers for webnn-baseline and WPT PRs
39 | 
40 | - Use modern WebIDL conventions
41 |   - [ ] tracker issue: https://github.com/webmachinelearning/webnn/issues/210
42 |   - ➡️ Status: Continue editorial work required to add algorithmic method steps, add internal slots, use getters, setters, define methods and constructors using the modern WebIDL conventions. Focus is on editorial enhancements. Normative changes and new feature work is deferred for later.
43 | 
44 | 
45 | #### ℹ️ WebML WG Charter 2023-2025 proposal ready for WG review
46 | 
47 | >Web Machine Learning Working Group Charter for 2023-2025 is now ready for WG review. Please review the draft PR and open issues, provide your comments and open new issues as appropriate to help shape the WG's technical scope. 
48 | 
49 | - Announcement https://lists.w3.org/Archives/Public/public-webmachinelearning-wg/2022Oct/0002.html
50 | - Charter PR https://github.com/w3c/machine-learning-charter/pull/19
51 | - Open issues https://github.com/w3c/machine-learning-charter/issues
52 |   - OpenXLA coordination
53 |   - On-device training
54 |   - Features deferred to WebNN v2
55 |   - Dedicated ML hardware accelerators: NPU, VPU, xPU
56 |   - Set of ops supported must be more comprehensive
57 |   - Level of abstraction for neural net operations
58 |   - WebRTC coordination
59 |   - WebGPU interoperability
60 | 


--------------------------------------------------------------------------------
/telcons/2023-01-12-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 12 January 2023 - 15:00-16:00 UTC
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 12 January 2023 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 12 January 2023 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 12 January 2023 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 7 | <tr><td> Berlin (Germany) <td> Thu, 12 January 2023 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 8 | <tr><td> Helsinki (Finland) <td> Thu, 12 January 2023 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
 9 | <tr><td> Shanghai (China) <td> Thu, 12 January 2023 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
10 | <tr><td> Tokyo (Japan) <td> Fri, 13 January 2023 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 12 January 2023 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230112T15
15 | 
16 |   </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2023/01/12-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | #### ℹ️ WebNN API open PRs and issues
30 |   
31 | >Review open PRs and discuss issues. Identify and fast track any priority changes that should get into the initial CR release train.
32 | 
33 | - Add `lstm` and `lstmCell` ops, rename `MLOperator` to `MLActivation`
34 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/321
35 | 
36 | - Simplify MLContext creation: remove `MLDeviceType`, remove `"high-performance"` from `MLPowerPreference`
37 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/322
38 | 
39 | - Transfer the input and output views for asynchronous execution
40 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/318
41 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/323
42 | 
43 | - Improve graph execution steps
44 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/316
45 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/319
46 |   
47 | - Simplify the operand layout support of conv2d and pooling 2d operations
48 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/324
49 | 
50 | #### ℹ️ WebNN API Candidate Recommendation readiness
51 | 
52 | >Confirm our CR readiness and discuss CR progressive enhancement. Work out a plan for our initial CR publication.
53 | - Current CR issues https://github.com/webmachinelearning/webnn/labels/cr
54 | 
55 | - Web platform tests
56 |   - [ ] tracker issue: https://github.com/webmachinelearning/webnn/issues/265
57 |   - Open webnn-baseline PRs: https://github.com/webmachinelearning/webnn-baseline/pulls
58 |   - Open wpt PRs: https://github.com/web-platform-tests/wpt/labels/wg-s_webmachinelearning
59 |   - ➡️ Status: Discuss & resolve any blockers for webnn-baseline and WPT PRs
60 | 
61 | - Use modern WebIDL and Infra standard conventions
62 |   - [ ] tracker issue: https://github.com/webmachinelearning/webnn/issues/210
63 |   - ➡️ Status: Continue editorial work required to add algorithmic method steps, add internal slots, use getters, setters, define methods and constructors using the modern WebIDL conventions. Focus is on editorial enhancements. Normative changes and new feature work is deferred for later.
64 | 
65 | 


--------------------------------------------------------------------------------
/telcons/2023-02-02-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 2 February 2023 - 15:00-16:00 UTC
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 2 February 2023 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 2 February 2023 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 2 February 2023 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 7 | <tr><td> Berlin (Germany) <td> Thu, 2 February 2023 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 8 | <tr><td> Helsinki (Finland) <td> Thu, 2 February 2023 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
 9 | <tr><td> Shanghai (China) <td> Thu, 2 February 2023 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
10 | <tr><td> Tokyo (Japan) <td> Fri, 3 February 2023 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 2 February 2023 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230202T15
15 | 
16 |   </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2023/02/02-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | #### ℹ️ WebNN API open PRs and issues
30 |   
31 | >Review open PRs and discuss issues. Identify and fast track any priority changes that should get into the initial CR release train.
32 | 
33 | - Simplify MLContext creation: remove `MLDeviceType`, remove `"high-performance"` from `MLPowerPreference`
34 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/322
35 |     - proposed as a v2 feature due to WebGPU dependency
36 |   
37 | - Rework the sync and async algorithms
38 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/316
39 |    - [ ] PR: https://github.com/webmachinelearning/webnn/pull/329
40 |    - Related PR: https://github.com/webmachinelearning/webnn/pull/323
41 | 
42 | - Improve batchNorm
43 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/334
44 |    - Note: We establish a blueprint here for formal op steps to be applied to all ops.
45 | 
46 | - Simplify the operand layout support of conv2d and pooling 2d operations
47 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/324
48 | 
49 | - Define the algorithm of calculating the effective padding for "same-upper" and "same-lower" option
50 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/326
51 | 
52 | - Clarify the usage of 32 bit floating point type and consider using double
53 |   - [ ] https://github.com/webmachinelearning/webnn/issues/325
54 | 
55 | 
56 | #### ℹ️ WebML WG Charter 2023-2025 Last Call
57 | 
58 | >Web Machine Learning Working Group Charter for 2023-2025 is now ready for W3C AC review. This is a Last Call for WG participants to provide comments before we merge the WebML WG Charter 2023-2025 PR and submit the updated charter for W3C Advisory Committee review. All issues have been triaged and those that received WG support have been addressed. We expect this charter to be operational starting Q2 2023.
59 | 
60 | - Announcement https://lists.w3.org/Archives/Public/public-webmachinelearning-wg/2022Oct/0002.html
61 | - Charter PR https://github.com/w3c/machine-learning-charter/pull/19
62 | - Issues https://github.com/w3c/machine-learning-charter/issues
63 | 


--------------------------------------------------------------------------------
/telcons/2023-02-16-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 16 February 2023 - 15:00-16:00 UTC
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 16 February 2023 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 16 February 2023 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 16 February 2023 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 7 | <tr><td> Berlin (Germany) <td> Thu, 16 February 2023 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 8 | <tr><td> Helsinki (Finland) <td> Thu, 16 February 2023 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
 9 | <tr><td> Shanghai (China) <td> Thu, 16 February 2023 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
10 | <tr><td> Tokyo (Japan) <td> Fri, 17 February 2023 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 16 February 2023 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230216T15
15 | 
16 |   </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2023/02/16-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | #### ℹ️ WebNN API open PRs and issues
30 |   
31 | >Review open PRs and discuss issues. Identify and fast track any priority changes that should get into the initial CR release train.
32 | 
33 | - Simplify MLContext creation: remove `MLDeviceType`, remove `"high-performance"` from `MLPowerPreference`
34 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/322
35 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/340
36 |   
37 | - Rework the sync and async algorithms
38 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/316
39 |    - [ ] PR: https://github.com/webmachinelearning/webnn/pull/329
40 |    - Related PR: https://github.com/webmachinelearning/webnn/pull/323
41 | 
42 | - Add internal slots to MLOperand and MLActivation
43 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/336
44 |    - [ ] PR: https://github.com/webmachinelearning/webnn/pull/337
45 | 
46 | - Improve batchNorm
47 |    - [ ] issue: https://github.com/webmachinelearning/webnn/issues/334
48 |    - [ ] PR: https://github.com/webmachinelearning/webnn/pull/339 (depends on PR #337)
49 |    - Note: We establish a blueprint here for formal op steps to be applied to all ops.
50 | 
51 | - Simplify the operand layout support of conv2d and pooling 2d operations
52 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/324
53 | 
54 | - Define the algorithm of calculating the effective padding for "same-upper" and "same-lower" option
55 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/326
56 | 
57 | - Clarify the usage of 32 bit floating point type and consider using double
58 |   - [ ] https://github.com/webmachinelearning/webnn/issues/325
59 | 


--------------------------------------------------------------------------------
/telcons/2023-03-02-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 2 March 2023 - 15:00-16:00 UTC
 2 | 
 3 | <table>
 4 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 2 March 2023 <td> 07:00 <td> UTC-8 hours
 5 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 2 March 2023 <td> 10:00 <td> UTC-5 hours
 6 | <tr><td> London (United Kingdom - England) <td> Thu, 2 March 2023 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 7 | <tr><td> Berlin (Germany) <td> Thu, 2 March 2023 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 8 | <tr><td> Helsinki (Finland) <td> Thu, 2 March 2023 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
 9 | <tr><td> Shanghai (China) <td> Thu, 2 March 2023 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
10 | <tr><td> Tokyo (Japan) <td> Fri, 3 March 2023 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
11 | <tr><td> Corresponding UTC (GMT) <td> Thu, 2 March 2023 <td colspan=2> 15:00 UTC
12 | </table>
13 | 
14 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230302T15
15 | 
16 |   </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2023/03/02-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | #### ℹ️ Call for Consensus: WebNN API Candidate Recommendation
30 | 
31 | >Call for Consensus to publish the Web Neural Network API as a Candidate Recommendation (CR) was issued.
32 | 
33 | Review by 2 March 2023: https://lists.w3.org/Archives/Public/public-webmachinelearning-wg/2023Feb/0005.html
34 | 
35 | #### ℹ️ WebNN API feedback at the GPU Web F2F
36 | 
37 | >WebNN API was on the agenda at the GPU Web F2F (2023-02-16/17).
38 | 
39 | Discuss feedback and next steps e.g.:
40 | - [ ] issue: https://github.com/webmachinelearning/webnn/issues/350
41 | - [ ] issue: https://github.com/webmachinelearning/webnn/issues/239
42 | 
43 | Agenda and minutes: https://lists.w3.org/Archives/Member/internal-gpu/2023Feb/0009.html (Member-only)
44 | 
45 | #### ℹ️ WebNN API open PRs and issues
46 |   
47 | >Review open PRs and discuss issues. Identify and fast track any priority changes that should get into the initial CR release train.
48 | 
49 | - Make axis definition of concat and split consistent 
50 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/345
51 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/352
52 | 
53 | - Use static padding values for pad operation
54 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/354
55 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/355
56 |   
57 | - Propose to add Parametric ReLU into the operation list
58 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/356
59 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/357
60 | 
61 | - Simplify the operand layout support of conv2d and pooling 2d operations
62 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/324
63 | 
64 | - Define the algorithm of calculating the effective padding for "same-upper" and "same-lower" option
65 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/326
66 | 
67 | - Clarify the usage of 32 bit floating point type and consider using double
68 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/325
69 | 
70 | - Subclass MLGraph based on the context that creates it
71 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/344
72 | 


--------------------------------------------------------------------------------
/telcons/2023-03-16-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | 
 2 | # WebML WG Teleconference – 16 March 2023 - 15:00-16:00 UTC
 3 | 
 4 | :bangbang: This meeting happens 1 hour later than usual in the US and Canada due to [Daylight Saving Time](https://www.timeanddate.com/time/dst/2023a.html) adjustment. Most of the US and Canada moved their clocks forward on Sunday, 12 March. This is two weeks earlier than Europe. Most European clocks will move 1 hour forward on Sunday, 26 March. 
 5 | 
 6 | <details open><summary>See the timezone table ...</summary>
 7 | <table>
 8 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 16 March 2023 <td> 08:00 <td> UTC-7 hours :bangbang:
 9 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 16 March 2023 <td> 11:00 <td> UTC-4 hours :bangbang:
10 | <tr><td> London (United Kingdom - England) <td> Thu, 16 March 2023 <td> 15:00 <td> UTC+0 hours
11 | <tr><td> Berlin (Germany) <td> Thu, 16 March 2023 <td> 16:00 <td> UTC+1 hours
12 | <tr><td> Helsinki (Finland) <td> Thu, 16 March 2023 <td> 17:00 <td> UTC+2 hours
13 | <tr><td> Shanghai (China) <td> Thu, 16 March 2023 <td> 23:00 <td> UTC+8 hours
14 | <tr><td> Tokyo (Japan) <td> Fri, 17 March 2023 <td> 00:00 <td> UTC+9 hours
15 | <tr><td> Corresponding UTC (GMT) <td> Thu, 16 March 2023 <td colspan=2> 15:00 UTC
16 | </table>
17 | 
18 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230316T15
19 |   </details>
20 | 
21 | ## Logistics
22 | 
23 | * Chair: Anssi
24 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
25 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
26 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
27 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
28 | * Minutes: https://www.w3.org/2023/03/16-webmachinelearning-minutes.html
29 | 
30 | ## Agenda
31 | 
32 | #### ℹ️ Proposed Charter 2023-2025 under W3C Advisory Committee review
33 |     
34 | Announcement: https://lists.w3.org/Archives/Public/public-webmachinelearning-wg/2023Mar/0002.html
35 | 
36 | Please make sure [your Advisory Committee representative](https://www.w3.org/Member/ACList) brings input (and ideally support) on the review.
37 | 
38 | #### ℹ️ WebNN - WebIDL and Infra standard conventions
39 | 
40 | - Sync and async algorithms
41 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/316
42 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/329
43 | 
44 | - MLOperand and MLActivation internal slots
45 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/336
46 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/337
47 |   
48 | #### ℹ️ WebNN - enhancements, editorials, questions
49 |   
50 | - Simplify the operand layout support of conv2d and pooling 2d operations
51 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/324
52 | 
53 | - Define the algorithm of calculating the effective padding for "same-upper" and "same-lower" option
54 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/326
55 | 
56 | - Clarify the usage of 32 bit floating point type and consider using double
57 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/325
58 | 
59 | - Subclass MLGraph based on the context that creates it
60 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/344
61 | 


--------------------------------------------------------------------------------
/telcons/2023-03-30-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | 
 2 | # WebML WG Teleconference – 30 March 2023 - 14:00-15:00 UTC+0
 3 | 
 4 | ⏰ We're back to the normal schedule. 
 5 | 
 6 | <details open><summary>See the timezone table ...</summary>
 7 | <table>
 8 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 30 March 2022 <td> 07:00 <td> UTC-7 hours
 9 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 30 March 2022 <td> 10:00 <td> UTC-4 hours
10 | <tr><td> London (United Kingdom - England) <td> Thu, 30 March 2022 <td> 15:00 <td> UTC+1 hours
11 | <tr><td> Berlin (Germany) <td> Thu, 30 March 2022 <td> 16:00 <td> UTC+2 hours
12 | <tr><td> Helsinki (Finland) <td> Thu, 30 March 2022 <td> 17:00 <td> UTC+3 hours
13 | <tr><td> Shanghai (China) <td> Thu, 30 March 2022 <td> 22:00 <td> UTC+8 hours
14 | <tr><td> Tokyo (Japan) <td> Thu, 30 March 2022 <td> 23:00 <td> UTC+9 hours
15 | <tr><td> Corresponding UTC (GMT) <td> Thu, 30 March 2022 <td colspan=2> 14:00 UTC
16 | </table>
17 | 
18 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230330T14
19 |   </details>
20 | 
21 | ## Logistics
22 | 
23 | * Chair: Anssi
24 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
25 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
26 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
27 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
28 | * Minutes: https://www.w3.org/2023/03/30-webmachinelearning-minutes.html
29 | 
30 | ## Agenda
31 | 
32 | #### ℹ️ Web Neural Network API W3C Candidate Recommendation published
33 | 
34 | Celebrate today's WebNN API CR publication and discuss next steps:
35 |   
36 | - 🚀 https://www.w3.org/TR/2023/CR-webnn-20230330/
37 | 
38 | #### ℹ️ WebNN - WebIDL and Infra standard conventions
39 | 
40 | - The constant() method steps
41 |   - [ ] Spec: https://www.w3.org/TR/webnn/#dom-mlgraphbuilder-constant
42 |   - [ ] Spec: https://www.w3.org/TR/webnn/#dom-mlgraphbuilder-constant-value-type
43 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/365
44 | 
45 | - MLOperand and MLActivation internal slots
46 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/336
47 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/337
48 | 
49 | - Sync and async algorithms
50 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/316
51 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/329
52 | 
53 | #### ℹ️ WebNN - enhancements, editorials, questions
54 | 
55 | - Define the algorithm of calculating the effective padding for "same-upper" and "same-lower" option
56 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/326
57 | 
58 | - Clarify the usage of 32 bit floating point type and consider using double
59 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/325
60 | 
61 | - Subclass MLGraph based on the context that creates it
62 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/344
63 | 


--------------------------------------------------------------------------------
/telcons/2023-04-27-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 27 April 2023 - 14:00-15:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 27 April 2022 <td> 07:00 <td> UTC-7 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 27 April 2022 <td> 10:00 <td> UTC-4 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 27 April 2022 <td> 15:00 <td> UTC+1 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 27 April 2022 <td> 16:00 <td> UTC+2 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 27 April 2022 <td> 17:00 <td> UTC+3 hours
10 | <tr><td> Shanghai (China) <td> Thu, 27 April 2022 <td> 22:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Thu, 27 April 2022 <td> 23:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 27 April 2022 <td colspan=2> 14:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230427T14
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2023/04/27-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | #### ℹ️ Contribution guidelines
30 | 
31 | Introduce and discuss improved contribution guidelines. Solicit feedback on further improvements and clarifications.
32 | 
33 | * webnn/CONTRIBUTING.md: https://github.com/webmachinelearning/webnn/blob/main/CONTRIBUTING.md
34 | 
35 | - [x] issue: https://github.com/webmachinelearning/webnn/issues/231
36 | - [x] PR: https://github.com/webmachinelearning/webnn/pull/381
37 | 
38 | #### ℹ️ WebIDL and Infra standard conventions
39 | 
40 | - The constant() method steps
41 |   - [ ] meta issue: https://github.com/webmachinelearning/webnn/issues/210
42 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/365
43 | 
44 | #### ℹ️ Enhancements, editorials, questions
45 | 
46 | - TC39 proposal to add Float16Array to JavaScript
47 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/373
48 |   - [ ] TC39 proposal: http://tc39.es/proposal-float16array/
49 | 
50 | - Define the algorithm of calculating the effective padding for "same-upper" and "same-lower" option
51 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/326
52 | 
53 | - Clarify the usage of 32 bit floating point type and consider using double
54 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/325
55 | 
56 | - Subclass MLGraph based on the context that creates it
57 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/344
58 | 
59 | #### ℹ️ Support for transformers
60 | 
61 | 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.
62 | 
63 | - [ ] issue: https://github.com/webmachinelearning/webnn/issues/375
64 | - Guidelines for adding new ops: https://github.com/webmachinelearning/webnn/blob/main/CONTRIBUTING.md#proposing-and-adding-a-new-operation
65 | 
66 | 
67 | 


--------------------------------------------------------------------------------
/telcons/2023-05-25-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 25 May 2023 - 14:00-15:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 25 May 2022 <td> 07:00 <td> UTC-7 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 25 May 2022 <td> 10:00 <td> UTC-4 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 25 May 2022 <td> 15:00 <td> UTC+1 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 25 May 2022 <td> 16:00 <td> UTC+2 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 25 May 2022 <td> 17:00 <td> UTC+3 hours
10 | <tr><td> Shanghai (China) <td> Thu, 25 May 2022 <td> 22:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Thu, 25 May 2022 <td> 23:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 25 May 2022 <td colspan=2> 14:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230525T14
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2023/05/25-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | #### ℹ️ WebIDL and Infra standard conventions
30 | 
31 | Review and land the open PRs to prep for an upcoming stylistic update:
32 | 
33 | - The computeSync and computeAsync algorithms
34 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/316
35 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/329
36 | 
37 | - The batchNormalization algorithm
38 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/334
39 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/339
40 | 
41 | - The clamp algorithm
42 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/347
43 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/348
44 | 
45 | - The concat algorithm
46 |   - [ ] meta issue: https://github.com/webmachinelearning/webnn/issues/210#issuecomment-1326361748
47 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/366
48 | 
49 | #### ℹ️ New features
50 | 
51 | - Float16Array (design finished, ready to integrate into WebNN)
52 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/373
53 |   - [ ] PR: https://github.com/webmachinelearning/webnn/pull/386
54 | 
55 | - 0D scalars
56 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/390
57 |   
58 | #### ℹ️ Enhancements, editorials, questions
59 |   
60 | - Error handling of MLNamedArrayBufferViews transfer algorithm
61 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/351
62 | 
63 | - Remove the support of null value in newShape parameter of reshape build method
64 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/388
65 | 
66 | #### ℹ️ Support for transformers
67 | 
68 | 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.
69 | 
70 | Status: The WG decided to start explore support for transformers in WebNN.
71 |   
72 | 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.
73 | 
74 | - [ ] issue: https://github.com/webmachinelearning/webnn/issues/375
75 | 


--------------------------------------------------------------------------------
/telcons/2023-06-08-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 8 June 2023 - 14:00-15:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 8 June 2022 <td> 07:00 <td> UTC-7 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 8 June 2022 <td> 10:00 <td> UTC-4 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 8 June 2022 <td> 15:00 <td> UTC+1 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 8 June 2022 <td> 16:00 <td> UTC+2 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 8 June 2022 <td> 17:00 <td> UTC+3 hours
10 | <tr><td> Shanghai (China) <td> Thu, 8 June 2022 <td> 22:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Thu, 8 June 2022 <td> 23:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 8 June 2022 <td colspan=2> 14:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230608T14
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
25 | * Minutes: https://www.w3.org/2023/06/08-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | #### ℹ️ Transformers.js with invited expert Joshua Lochner
30 | 
31 | - Presentation slides: https://lists.w3.org/Archives/Public/www-archive/2023Jun/att-0000/Transformers_js.pdf
32 | 
33 | Joshua Lochner (@xenova) will introduce Transformers.js and share his learnings from this project including practical use cases to help inform the WebNN v2 feature work. Our shared aspiration is to make WebNN the most performant and robust backend for a future version of Transformers.js.
34 |   
35 | - Transformers.js GH repo: https://github.com/xenova/transformers.js
36 | - Transformers.js demo: https://xenova.github.io/transformers.js/  
37 | - Transformer.js docs: https://huggingface.co/docs/transformers.js/index
38 | - Related WebNN issue: https://github.com/webmachinelearning/webnn/issues/375
39 | 
40 | Please use WebNN issue 375 for comments, feedback, brainstorming, additional information for this broader space of transformers.
41 | 
42 | #### ℹ️ Any other topics
43 |   
44 | Subject to time, we may discuss open issues and PRs for WebNN.
45 | 


--------------------------------------------------------------------------------
/telcons/2023-06-22-wg-agenda.md:
--------------------------------------------------------------------------------
1 | # WebML WG Teleconference – 22 June 2023 - 14:00-15:00 UTC
2 | 
3 | This meeting is rescheduled to [June 29, 2023](https://github.com/webmachinelearning/meetings/blob/main/telcons/2023-06-29-wg-agenda.md) due to overlap with the [Dragon Boat Festival](https://en.wikipedia.org/wiki/Dragon_Boat_Festival), a traditional Chinese holiday.
4 | 
5 | # 🐉🚣
6 | 


--------------------------------------------------------------------------------
/telcons/2023-06-29-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 29 June 2023 - 14:00-15:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 29 June 2022 <td> 07:00 <td> UTC-7 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 29 June 2022 <td> 10:00 <td> UTC-4 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 29 June 2022 <td> 15:00 <td> UTC+1 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 29 June 2022 <td> 16:00 <td> UTC+2 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 29 June 2022 <td> 17:00 <td> UTC+3 hours
10 | <tr><td> Shanghai (China) <td> Thu, 29 June 2022 <td> 22:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Thu, 29 June 2022 <td> 23:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 29 June 2022 <td colspan=2> 14:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230629T14
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2023/06/29-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | ### ℹ️ Announcements
30 | 
31 | Implementation Status of WebNN Operations https://webmachinelearning.github.io/webnn-status/
32 |  
33 | ### ℹ️ WebNN v2: text-to-image and text-to-text use cases and requirements
34 | 
35 | Discuss transformer-based generative AI use cases for text-to-image and text-to-text models. Solicit input on new ops and data types required to support selected models informed by prototyping efforts. Background material:
36 |  
37 | - Transformer-based models and use cases discussion: https://github.com/webmachinelearning/webnn/issues/375
38 | - WebNN DirectML prototype: https://github.com/fdwr/chromium-src-webnn-dml/pull/1
39 | - Transformers.js feasible tasks: https://lists.w3.org/Archives/Public/www-archive/2023Jun/att-0000/Transformers_js.pdf
40 |  
41 | ### ℹ️ WebIDL and Infra standard conventions
42 | 
43 | Per feedback we've shifted to use an integration branch for this conventions alignment task. We'll review the work mode to ensure everyone's up to date.
44 |  
45 | Task list: https://github.com/webmachinelearning/webnn/issues/210#issuecomment-1326361748
46 | 
47 | ### ℹ️ New features
48 | 
49 | - Allow 0 size dimensions
50 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/391
51 |  
52 |  ### ℹ️ Enhancements
53 |  
54 | - split() into sizes not widely supported
55 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/392
56 | - Clarify the restriction for minValue and maxValue of MLClampOptions
57 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/396
58 | 


--------------------------------------------------------------------------------
/telcons/2023-09-21-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 21 September 2023 - 14:00-15:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 21 September 2022 <td> 07:00 <td> UTC-7 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 21 September 2022 <td> 10:00 <td> UTC-4 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 21 September 2022 <td> 15:00 <td> UTC+1 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 21 September 2022 <td> 16:00 <td> UTC+2 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 21 September 2022 <td> 17:00 <td> UTC+3 hours
10 | <tr><td> Shanghai (China) <td> Thu, 21 September 2022 <td> 22:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Thu, 21 September 2022 <td> 23:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 21 September 2022 <td colspan=2> 14:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230921T14
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2023/09/21-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | ### ℹ️ Update from TPAC 2023
30 | 
31 | W3C's Web Machine Learning WG did not formally meet at this year's TPAC, but the WG's work received a lot of interest during the hallway track and in other meetings. Anssi will share an update.
32 | 
33 | ### ℹ️ WebNN v2: Proposed model targets
34 | 
35 | The WG has identified the following as its proposed v2 model targets:
36 | 
37 | - Text-to-image: Stable Diffusion unet/VAE/text encoder
38 | - Image segmentation: Segment Everything decoder
39 | - Speech-to-text: Whisper Tiny
40 | - Text-to-text generation (encoder-decoder): t5 and m2m100
41 | - Text-generation (decoder-only): llama
42 | 
43 | We will discuss any feedback and with no concerns we initiate work on an op breakdown to better understand what is common across these architectures to inform WebNN API v2 priorities.
44 | 
45 | Please provide your feedback via https://github.com/webmachinelearning/webnn/issues/375
46 | 
47 | ### ℹ️ Storage APIs for caching large models
48 | 
49 | Discussion on general issues related to caching large models, current Storage APIs, specific issues and workarounds for those APIs.
50 | 
51 | ### ℹ️ Security considerations
52 | 
53 | - Computation control-flow attack based on weights / constants change
54 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/443
55 | 
56 | ### ℹ️ New features
57 | 
58 | - Allow checking whether operators/types are supported for a backend before creating a graph
59 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/463
60 | 
61 | ### ℹ️ Enhancements
62 |  
63 | - split() into sizes not widely supported
64 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/392
65 | - Clarify the restriction for minValue and maxValue of MLClampOptions
66 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/396
67 | - Type of parameters should match the input data type
68 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/442
69 | 
70 | 


--------------------------------------------------------------------------------
/telcons/2023-10-05-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 5 October 2023 - 14:00-15:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 5 October 2022 <td> 07:00 <td> UTC-7 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 5 October 2022 <td> 10:00 <td> UTC-4 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 5 October 2022 <td> 15:00 <td> UTC+1 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 5 October 2022 <td> 16:00 <td> UTC+2 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 5 October 2022 <td> 17:00 <td> UTC+3 hours
10 | <tr><td> Shanghai (China) <td> Thu, 5 October 2022 <td> 22:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Thu, 5 October 2022 <td> 23:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 5 October 2022 <td colspan=2> 14:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20231005T14
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2023/10/05-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | ### ℹ️ WebNN v2: Review op breakdown for proposed model targets
30 | 
31 | The WG has identified the following as its v2 model targets:
32 | 
33 | - Text-to-image: Stable Diffusion unet/VAE/text encoder
34 | - Image segmentation: Segment Everything decoder
35 | - Speech-to-text: Whisper Tiny
36 | - Text-to-text generation (encoder-decoder): t5 and m2m100
37 | - Text-generation (decoder-only): llama
38 | 
39 | We will review the initial op breakdown to better understand what is common across these architectures to inform WebNN API v2 priorities.
40 | 
41 | Please provide your feedback via https://github.com/webmachinelearning/webnn/issues/375
42 | 
43 | ### ℹ️ Security considerations
44 | 
45 | - Computation control-flow attack based on weights / constants change
46 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/443
47 | 
48 | ### ℹ️ New features
49 | 
50 | - Allow checking whether operators/types are supported for a backend before creating a graph
51 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/463
52 | 
53 | ### ℹ️ Enhancements
54 |  
55 | - split() into sizes not widely supported
56 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/392
57 | - Clarify the restriction for minValue and maxValue of MLClampOptions
58 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/396
59 | - Type of parameters should match the input data type
60 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/442
61 | 


--------------------------------------------------------------------------------
/telcons/2023-10-19-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 19 October 2023 - 14:00-15:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 19 October 2022 <td> 07:00 <td> UTC-7 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 19 October 2022 <td> 10:00 <td> UTC-4 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 19 October 2022 <td> 15:00 <td> UTC+1 hours
 8 | <tr><td> Berlin (Germany) <td> Thu, 19 October 2022 <td> 16:00 <td> UTC+2 hours
 9 | <tr><td> Helsinki (Finland) <td> Thu, 19 October 2022 <td> 17:00 <td> UTC+3 hours
10 | <tr><td> Shanghai (China) <td> Thu, 19 October 2022 <td> 22:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Thu, 19 October 2022 <td> 23:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 19 October 2022 <td colspan=2> 14:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20231019T14
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2023/10/19-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | ### ℹ️ WebNN v2: Review transformer ops spec contributions
30 | 
31 | Discuss and review any spec PRs or contributions for the proposed transformer ops informed by the op breakdown exercise.
32 | 
33 | References:
34 | - Op breakdown (w/ TOSA, StableHLO mapping): https://docs.google.com/spreadsheets/d/1ELfHuv2UqP2LoXWLgqsC0L8T_qqfBx48KxzFighl8d8/edit#gid=86906292
35 | - Feedback via https://github.com/webmachinelearning/webnn/issues/375
36 | 
37 | ### ℹ️ WebGPU security technical report
38 | 
39 | Discuss the recently published WebGPU security technical report, hardening opportunities for the WebNN API based on these insights.
40 | 
41 | - WebGPU security technical report: https://chromium.googlesource.com/chromium/src/+/main/docs/security/research/graphics/webgpu_technical_report.md
42 | 
43 | ### ℹ️ Enhancements
44 | 
45 | - Softmax axis absent
46 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/466 
47 | - split() into sizes not widely supported
48 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/392
49 | - Clarify the restriction for minValue and maxValue of MLClampOptions
50 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/396
51 | 


--------------------------------------------------------------------------------
/telcons/2023-11-02-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 2 November 2023 - 14:00-15:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 2 November 2023 <td> 07:00 <td> UTC-7 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 2 November 2023 <td> 10:00 <td> UTC-4 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 2 November 2023 <td> 14:00 <td> UTC+0 hours (adjusted for DST)
 8 | <tr><td> Berlin (Germany) <td> Thu, 2 November 2023 <td> 15:00 <td> UTC+1 hours (adjusted for DST)
 9 | <tr><td> Helsinki (Finland) <td> Thu, 2 November 2023 <td> 16:00 <td> UTC+2 hours (adjusted for DST)
10 | <tr><td> Shanghai (China) <td> Thu, 2 November 2023 <td> 22:00 <td> UTC+8 hours
11 | <tr><td> Tokyo (Japan) <td> Thu, 2 November 2023 <td> 23:00 <td> UTC+9 hours
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 2 November 2023 <td colspan=2> 14:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20231102T14
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2023/11/02-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | ### ℹ️ WebNN v2: Review transformer ops spec contributions (continued)
30 | 
31 | Discuss and review any spec PRs or contributions for the proposed transformer ops informed by the op breakdown exercise.
32 | 
33 | References:
34 | - Op breakdown (w/ TOSA, StableHLO mapping): https://docs.google.com/spreadsheets/d/1ELfHuv2UqP2LoXWLgqsC0L8T_qqfBx48KxzFighl8d8/edit#gid=86906292
35 | - Feedback via https://github.com/webmachinelearning/webnn/issues/375
36 | 
37 | ### ℹ️ Enhancements
38 | 
39 | - 0D scalars
40 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/390
41 | - Softmax axis absent
42 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/466 
43 | - split() into sizes not widely supported
44 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/392
45 | - Clarify the restriction for minValue and maxValue of MLClampOptions
46 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/396
47 | - Simplify matmul op
48 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/470
49 | 


--------------------------------------------------------------------------------
/telcons/2023-11-16-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 16 November 2023 - 15:00-16:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 16 November 2023 <td> 07:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 16 November 2023 <td> 10:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 16 November 2023 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 8 | <tr><td> Berlin (Germany) <td> Thu, 16 November 2023 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 9 | <tr><td> Helsinki (Finland) <td> Thu, 16 November 2023 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
10 | <tr><td> Shanghai (China) <td> Thu, 16 November 2023 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
11 | <tr><td> Tokyo (Japan) <td> Fri, 17 November 2023 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 16 November 2023 <td colspan=2> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20231116T15
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2023/11/16-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | ### Announcements
30 | 
31 | - W3C Web & Network Interest Group rechartering, coordination opportunity with WebML WG
32 | 
33 | ### ℹ️ WebNN v2: Review transformer ops spec contributions (continued)
34 | 
35 | Discuss and review any spec PRs or contributions for the proposed transformer ops informed by the op breakdown exercise.
36 | 
37 | - [ ] issue https://github.com/webmachinelearning/webnn/issues/375
38 | - [ ] PR: https://github.com/webmachinelearning/webnn/pull/478
39 | 
40 | References:
41 | 
42 | - [Op breakdown (w/ TOSA, StableHLO mapping)](https://docs.google.com/spreadsheets/d/1ELfHuv2UqP2LoXWLgqsC0L8T_qqfBx48KxzFighl8d8/)
43 | 
44 | 
45 | ### ℹ️ Enhancements
46 | 
47 | - Simplify matmul op
48 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/470
49 | 
50 | - Define the algorithm of calculating the effective padding for "same-upper" and "same-lower" option
51 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/326
52 | 
53 | - API lacks handling for async ML device errors on the context
54 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/477
55 | 
56 | - Remove builder.constant(value, dataType) variant
57 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/475
58 | 
59 | - Need to restrict the value of alpha to be positive for elu operation
60 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/383
61 | 
62 | - Allow 0 size dimensions
63 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/391
64 | 


--------------------------------------------------------------------------------
/telcons/2023-11-30-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 30 November 2023 - 15:00-16:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 30 November 2023 <td> 07:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 30 November 2023 <td> 10:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 30 November 2023 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 8 | <tr><td> Berlin (Germany) <td> Thu, 30 November 2023 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 9 | <tr><td> Helsinki (Finland) <td> Thu, 30 November 2023 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
10 | <tr><td> Shanghai (China) <td> Thu, 30 November 2023 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
11 | <tr><td> Tokyo (Japan) <td> Fri, 31 November 2023 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 30 November 2023 <td colspan=2> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20231130T15
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2023/11/30-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | ### Announcements
30 | 
31 | - Implementation Status of WebNN Operations updated
32 |   - Table: https://webmachinelearning.github.io/webnn-status/ 
33 |   - Diff: https://github.com/webmachinelearning/webmachinelearning.github.io/pull/58/files
34 | 
35 | ### ℹ️ WebNN v2: Review transformer ops spec contributions (continued)
36 | 
37 | Discuss and review the PR for the proposed transformer ops informed by the op breakdown exercise.
38 | 
39 | - [ ] issue: https://github.com/webmachinelearning/webnn/issues/375
40 | - [ ] PR: https://github.com/webmachinelearning/webnn/pull/478
41 | 
42 | References:
43 | 
44 | - [Op breakdown (w/ TOSA, StableHLO mapping)](https://docs.google.com/spreadsheets/d/1ELfHuv2UqP2LoXWLgqsC0L8T_qqfBx48KxzFighl8d8/)
45 | 
46 | 
47 | ### ℹ️ Enhancements
48 | 
49 | - Remove builder.constant(value, dataType) variant
50 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/475
51 | 
52 | - Need to restrict the value of alpha to be positive for elu operation
53 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/383
54 | 
55 | - Allow 0 size dimensions
56 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/391
57 | 
58 | - Should scale and bias be required inputs for batchNormalization op?
59 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/481
60 | 
61 | ### ℹ️ New features
62 | 
63 | - Support for device-based tensor storage objects
64 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/482
65 | 


--------------------------------------------------------------------------------
/telcons/2023-12-14-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | # WebML WG Teleconference – 14 December 2023 - 15:00-16:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco (U.S.A. - California) <td> Thu, 14 December 2023 <td> 07:00 <td> UTC-8 hours
 6 | <tr><td> Boston (U.S.A. - Massachusetts) <td> Thu, 14 December 2023 <td> 10:00 <td> UTC-5 hours
 7 | <tr><td> London (United Kingdom - England) <td> Thu, 14 December 2023 <td> 15:00 <td> UTC+0 hours (adjusted for DST)
 8 | <tr><td> Berlin (Germany) <td> Thu, 14 December 2023 <td> 16:00 <td> UTC+1 hours (adjusted for DST)
 9 | <tr><td> Helsinki (Finland) <td> Thu, 14 December 2023 <td> 17:00 <td> UTC+2 hours (adjusted for DST)
10 | <tr><td> Shanghai (China) <td> Thu, 14 December 2023 <td> 23:00 <td> UTC+8 hours (adjusted for DST)
11 | <tr><td> Tokyo (Japan) <td> Fri, 15 December 2023 <td> 00:00 <td> UTC+9 hours (adjusted for DST)
12 | <tr><td> Corresponding UTC (GMT) <td> Thu, 14 December 2023 <td colspan=2> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20231214T15
16 | </details>
17 | 
18 | ## Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: ? (see [howto](https://github.com/webmachinelearning/meetings/blob/main/scribe-howto.md), volunteers welcome!)
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2023/12/14-webmachinelearning-minutes.html
26 | 
27 | ## Agenda
28 | 
29 | ### ℹ️ WebNN v2: Transformer ops spec contributions celebration
30 | 
31 | Topic reserved for the celebration of this major milestone. Please bring your own (virtual) beverages and snacks. 🥳🍿
32 | 
33 | - [x] issue: https://github.com/webmachinelearning/webnn/issues/375
34 | - [x] PR: https://github.com/webmachinelearning/webnn/pull/478
35 | 
36 | ### ℹ️ Enhancements
37 | 
38 | - Remove builder.constant(value, dataType) variant
39 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/475
40 | 
41 | - Need to restrict the value of alpha to be positive for elu operation
42 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/383
43 | 
44 | - Allow 0 size dimensions
45 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/391
46 | 
47 | - API lacks handling for async ML device errors on the context (revisit)
48 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/477
49 | 
50 | ### ℹ️ New features
51 | 
52 | - Support for device-based tensor storage objects
53 |   - [ ] issue: https://github.com/webmachinelearning/webnn/issues/482
54 | 


--------------------------------------------------------------------------------
/telcons/2025-02-26-cg-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 26 February 2025 - 00:00-01:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco <td> Tue, 25 February 2025 <td> 16:00
 6 | <tr><td> Boston <td> Tue, 25 February 2025 <td> 19:00  
 7 | <tr><td> London <td> Wed, 26 February 2025 <td> 00:00  
 8 | <tr><td> Berlin <td> Wed, 26 February 2025 <td> 01:00 
 9 | <tr><td> Helsinki <td> Wed, 26 February 2025 <td> 02:00 
10 | <tr><td> Shanghai <td> Wed, 26 February 2025 <td> 08:00
11 | <tr><td> Tokyo <td> Wed, 26 February 2025 <td> 09:00
12 | <tr><td> UTC <td> Wed, 26 February 2025 <td> 00:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20250226T00
16 | </details>
17 | 
18 | ### Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: Etienne
22 | * Zoom joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
23 | * Minutes: [2025-02-26-cg-minutes.md](/telcons/2025-02-26-cg-minutes.md)
24 | 
25 | ### Agenda
26 | 
27 | #### ⌛5:00 - Introductions
28 | 
29 | #### ⌛10:00 [Google: Domenic Denicola] - Summary of the current API proposals and describe status of each
30 | 
31 | - Prompt API
32 |    - Explainer: https://github.com/webmachinelearning/prompt-api
33 |    - Motivation
34 |    - Availability
35 |       - Origin Trial only for extension M131-136
36 |          - [Registration link](https://developer.chrome.com/origintrials/#/view_trial/320318523496726529)
37 |       - DevTrial drive-by web
38 | - Translator & Language Detector API
39 |    - Explainer:  https://github.com/webmachinelearning/translation-api
40 |    - Motivation
41 |    - Availability
42 |       - Translator
43 |          - Origin Trial M131-136
44 |             - [Registration link](https://developer.chrome.com/origintrials/#/view_trial/4445615782168100865)
45 |       - Language Detector
46 |          - Origin Trial M130-135
47 |             - [Registration link](https://developer.chrome.com/origintrials/#/view_trial/662592095176884225)
48 | - Writing Assistance APIs
49 |    - Explainer: https://github.com/webmachinelearning/writing-assistance-apis 
50 |    - Motivation
51 |    - Availability
52 |       - Summarizer
53 |          - Origin Trial M131-136
54 |             - [Registration link](https://developer.chrome.com/origintrials/#/view_trial/1923599990840623105)
55 |    - Writer
56 |       - Dev Trial
57 |       - Planned Origin Trial M137 or 138
58 |    - Rewriter
59 |       - Dev Trial
60 |       - Planned Origin Trial M137 or 138
61 | - General patterns, namespaces, etc.
62 | 
63 | #### ⌛10:00 [Browser vendors] - Roundtable with all the browser vendors
64 | 
65 | - Mozilla’s Firefox AI Runtime?
66 |    - @tomayac’s Experience with it: https://blog.tomayac.com/2025/02/07/playing-with-ai-inference-in-firefox-web-extensions/ 
67 |    - What interests or doesn’t interest you about the APIs?
68 |       - For the latter, we would like to understand concerns, reservations and what could make you change your mind?
69 | 
70 | #### ⌛20:00 [Developers]
71 | 
72 | - What do you like/not like about the current API proposals?
73 | - What are your most urgent needs in this field?
74 | 
75 | #### ⌛10:00 [Browser vendors] - What is your vision for AI in the browser?
76 | 
77 | - ⌛2:00 Etienne Noël Process discussion
78 |    - I want to provide you with the ability to suggest topics. I propose that we open a Github issue for each future iterations of this meeting: https://github.com/webmachinelearning/meetings/issues
79 | 


--------------------------------------------------------------------------------
/telcons/2025-02-27-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML WG Teleconference – 27 February 2025 - 15:00-16:00 UTC
 2 | 
 3 | <details><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco <td> Thu, 27 February 2025 <td> 07:00
 6 | <tr><td> Boston <td> Thu, 27 February 2025 <td> 10:00  
 7 | <tr><td> London <td> Thu, 27 February 2025 <td> 15:00  
 8 | <tr><td> Berlin <td> Thu, 27 February 2025 <td> 16:00 
 9 | <tr><td> Helsinki <td> Thu, 27 February 2025 <td> 17:00 
10 | <tr><td> Shanghai <td> Thu, 27 February 2025 <td> 23:00
11 | <tr><td> Tokyo <td> Fri, 28 February 2025 <td> 00:00
12 | <tr><td> UTC <td> Thu, 27 February 2025 <td> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20250227T15
16 | </details>
17 | 
18 | ### Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: Anssi
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Zoom joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2025/02/27-webmachinelearning-minutes.html
26 | 
27 | ### Agenda
28 | 
29 | #### 🧪 Incubations summary
30 | 
31 | A summary of the most recent WebML Community Group meeting will be shared to keep the Working Group abreast of incubations.
32 | 
33 | - [WebML CG Teleconference – 26 February 2025 - 00:00-01:00 UTC](https://github.com/webmachinelearning/meetings/blob/main/telcons/2025-02-26-cg-agenda.md)
34 | 
35 | #### ℹ️ Query mechanism for supported devices
36 | 
37 | Discuss, refine and solicit proposals and use cases for the query mechanism for supported devices. Consider implementability, privacy, security implications.
38 | 
39 | - ⨀ https://github.com/webmachinelearning/webnn/issues/815
40 | 
41 | #### ℹ️ Operator set Wave 3
42 | 
43 | Review draft PR and (many!) issues it closes. Discuss any opens as needed.
44 | 
45 | PR welcomes contributions to finish algorithm steps in gatherNd/scatterNd, update slice output shape.
46 | 
47 | - ⛙ https://github.com/webmachinelearning/webnn/pull/805
48 | 
49 | Closes: ⨀ [#93](https://github.com/webmachinelearning/webnn/issues/93), ⨀ [#467](https://github.com/webmachinelearning/webnn/issues/467), ⨀ [#772](https://github.com/webmachinelearning/webnn/issues/772), ⨀ [#767](https://github.com/webmachinelearning/webnn/issues/767), ⨀ [#773](https://github.com/webmachinelearning/webnn/issues/773), ⨀ [#779](https://github.com/webmachinelearning/webnn/issues/779)
50 | 
51 | #### 🏷️ [feature request](https://github.com/webmachinelearning/webnn/labels/feature%20request) issues
52 | 
53 | - Rounding operators
54 |   - ⨀ https://github.com/webmachinelearning/webnn/issues/817
55 |   - ☰ Discuss known gaps in rounding functions, review the initial proposal.
56 | 
57 | - Remove pool2d MLRoundingType
58 |   - ⨀ https://github.com/webmachinelearning/webnn/issues/324
59 |   - ⛙ https://github.com/webmachinelearning/webnn/pull/770
60 |   - ☰ Discuss remaining PR review comments.
61 | 
62 | #### 🏷️ [bug](https://github.com/webmachinelearning/webnn/labels/bug) issues
63 | 
64 | - Reject promises for in-progress operations on an MLTensor if it is destroyed
65 |   - ⨀ follow-up to https://github.com/webmachinelearning/webnn/issues/477
66 |   - ⛙ https://github.com/webmachinelearning/webnn/pull/799
67 |   - ☰ Review PR.
68 | 
69 | #### 🏷️ [interop](https://github.com/webmachinelearning/webnn/labels/interop) issues
70 | 
71 | - Behavior when there are NaNs in argmin/max inputs
72 |   - ⨀ https://github.com/webmachinelearning/webnn/issues/811
73 |   - ☰ Discuss how backends handle NaNs in the input and agree on next steps.
74 | 


--------------------------------------------------------------------------------
/telcons/2025-03-12-cg-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference – 12 March 2025 - 00:00-01:00 UTC
 2 | 
 3 | <details open><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco <td> Tue, 11 March 2025 <td> 17:00 adjusted for DST
 6 | <tr><td> Boston <td> Tue, 11 March 2025 <td> 20:00 adjusted for DST
 7 | <tr><td> London <td> Wed, 12 March 2025 <td> 00:00  
 8 | <tr><td> Berlin <td> Wed, 12 March 2025 <td> 01:00 
 9 | <tr><td> Helsinki <td> Wed, 12 March 2025 <td> 02:00 
10 | <tr><td> Shanghai <td> Wed, 12 March 2025 <td> 08:00
11 | <tr><td> Tokyo <td> Wed, 12 March 2025 <td> 09:00
12 | <tr><td> UTC <td> Wed, 12 March 2025 <td> 00:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20250312T00
16 | </details>
17 | 
18 | ### Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: Etienne
22 | * Zoom joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
23 | * Minutes: [2025-03-12-cg-minutes.md](/telcons/2025-03-12-cg-minutes.md)
24 | 
25 | ### Agenda
26 | 
27 | #### OT Feedback from partners
28 | 
29 | Presented by Etienne Noël, Kenji Baheux
30 | 
31 | #### Proofreader API
32 | 
33 | Presented by Etienne Noël
34 | 
35 | - [Proofreader API Explainer](https://github.com/explainers-by-googlers/proofreader-api)
36 | - A possible new CG deliverable? Feedback welcome!
37 | 
38 | #### Prompt API
39 | 
40 | - Appetite to standardize Prompt API and make it available on the Open Web?
41 | - Right now, in Chrome, it’s currently only for extensions
42 | 
43 | #### Questions to Developers
44 | 
45 | - What do you like/not like about the Built-In AI APIs proposal?
46 | - What are your most urgent needs not currently solved by the Built-In AI APIs?
47 | 
48 | #### Recent spec changes and discussions recap
49 | 
50 | By Domenic Denicola
51 | 
52 | - [Token accounting API renaming options](https://github.com/webmachinelearning/writing-assistance-apis/issues/5#issuecomment-2683970259)
53 | 


--------------------------------------------------------------------------------
/telcons/2025-04-10-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML WG Teleconference – 10 April 2025 - 14:00-15:00 UTC
 2 | 
 3 | > [!IMPORTANT]
 4 | >We're back to the usual US/Canada time and also start an hour earlier for China and Japan participants.
 5 | 
 6 | <details open><summary>See the timezone table ...</summary>
 7 | <table>
 8 | <tr><td> San Francisco <td> Thu, 10 April 2025 <td> 07:00
 9 | <tr><td> Boston <td> Thu, 10 April 2025 <td> 10:00
10 | <tr><td> London <td> Thu, 10 April 2025 <td> 15:00  
11 | <tr><td> Berlin <td> Thu, 10 April 2025 <td> 16:00 
12 | <tr><td> Helsinki <td> Thu, 10 April 2025 <td> 17:00 
13 | <tr><td> Shanghai <td> Thu, 10 April 2025 <td> 22:00
14 | <tr><td> Tokyo <td> Thu, 10 April 2025 <td> 23:00
15 | <tr><td> UTC <td> Thu, 10 April 2025 <td> 14:00 UTC
16 | </table>
17 | 
18 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20250410T14
19 | </details>
20 | 
21 | ### Logistics
22 | 
23 | * Chair: Anssi
24 | * Scribe: Anssi
25 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
26 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
27 | * Zoom joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
28 | * Minutes: https://www.w3.org/2025/04/10-webmachinelearning-minutes.html
29 | 
30 | ### Agenda
31 | 
32 | #### 🧪 Incubations summary
33 | 
34 | A summary of the most recent WebML Community Group meeting will be shared to keep the Working Group abreast of incubations.
35 | 
36 | #### 🤖 AI Agents
37 | 
38 | A debrief on the recent W3C breakout session on AI Agents.
39 | 
40 | #### ℹ️ Tensors for graph constants
41 | 
42 | Discuss PR review feedback and proposed API design alternatives on how to best represent constness.
43 | 
44 | - ⨀ https://github.com/webmachinelearning/webnn/issues/760
45 | - ⛙ https://github.com/webmachinelearning/webnn/pull/830
46 | 
47 | #### ℹ️ Caching mechanism for MLGraph
48 | 
49 | Discuss explainer updates & migration. Review requirements from web frameworks that would use this caching mechanism, related API proposals.
50 | 
51 | - ⨀ https://github.com/webmachinelearning/webnn/issues/807
52 | - ☰ Explainer: https://github.com/webmachinelearning/hybrid-ai/blob/main/proposals/cache.md
53 | - ☰ OnnxRuntime EP context cache: https://onnxruntime.ai/docs/execution-providers/EP-Context-Design.html
54 | - ☰ (related proposal) Cross-Origin Storage API: https://github.com/explainers-by-googlers/cross-origin-storage
55 | 
56 | #### ℹ️ Query mechanism for supported devices
57 | 
58 | Discuss new use case feedback, Apple's device privacy considerations. Address questions on the _capacity_ concept. Review the latest iteration of the API proposal.
59 | 
60 | - ⨀ https://github.com/webmachinelearning/webnn/issues/815
61 | 
62 | #### ℹ️ Core operator set
63 | 
64 | Revisit our core operator set effort that aims to identify current primitive gaps by mapping compositional fundamentals (e.g. PyTorch prims, TOSA, StableHLO) to WebNN operators.
65 | 
66 | - ⨀ https://github.com/webmachinelearning/webnn/issues/573
67 | - ☰ [Machine Learning Operator Mapping - All Raw Operators](https://onedrive.live.com/edit?id=EE82F5C6F06C7371!345450&resid=EE82F5C6F06C7371!345450&ithint=file%2Cxlsx&authkey=!AK8f-RDTleqlLXE&wdo=2&cid=ee82f5c6f06c7371)
68 | 
69 | 
70 | >[!NOTE]
71 | >We will discuss other [open issues](https://github.com/webmachinelearning/webnn/issues) based on participants' suggestions, time allowing.
72 | 


--------------------------------------------------------------------------------
/telcons/2025-05-08-wg-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML WG Teleconference – 8 May 2025 - 15:00-16:00 UTC
 2 | 
 3 | <details open><summary>See the timezone table ...</summary>
 4 | <table>
 5 | <tr><td> San Francisco <td> Thu, 8 May 2025 <td> 08:00
 6 | <tr><td> Boston <td> Thu, 8 May 2025 <td> 11:00
 7 | <tr><td> London <td> Thu, 8 May 2025 <td> 16:00  
 8 | <tr><td> Berlin <td> Thu, 8 May 2025 <td> 17:00 
 9 | <tr><td> Helsinki <td> Thu, 8 May 2025 <td> 18:00 
10 | <tr><td> Shanghai <td> Thu, 8 May 2025 <td> 23:00
11 | <tr><td> Tokyo <td> Fri, 9 May 2025 <td> 00:00
12 | <tr><td> UTC <td> Thu, 8 May 2025 <td> 15:00 UTC
13 | </table>
14 | 
15 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20250508T15
16 | </details>
17 | 
18 | ### Logistics
19 | 
20 | * Chair: Anssi
21 | * Scribe: Anssi
22 | * IRC: irc://irc.w3.org:6667/#webmachinelearning
23 | * IRC web client: https://irc.w3.org/?channels=#webmachinelearning
24 | * Zoom joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2023Jun/0000.html
25 | * Minutes: https://www.w3.org/2025/05/08-webmachinelearning-minutes.html
26 | 
27 | ### Agenda
28 | 
29 | #### 🧪 Incubations
30 | 
31 | Discuss recent WebML Community Group development to keep the Working Group abreast of incubation progress.
32 | 
33 | New proposal: ✨ Local Inference Web extension https://github.com/webmachinelearning/proposals/issues/9
34 | 
35 | 
36 | #### 🏷️ Operator specific issues
37 | 
38 | Review and discuss operator specific issues that reduce code complexity and improve maintainability, e.g.:
39 | 
40 | - `layerNormalization`
41 |   - ⨀ https://github.com/webmachinelearning/webnn/issues/748
42 | 
43 | - `triangular`
44 |   - ⨀ https://github.com/webmachinelearning/webnn/issues/768
45 | 
46 | - `sign`
47 |   - ⨀ https://github.com/webmachinelearning/webnn/issues/845
48 | 
49 | 
50 | #### ℹ️ WebNN wide review
51 | 
52 | Review and discuss interim wide review feedback and the group's proposed response.
53 | 
54 | - ☰ https://github.com/webmachinelearning/webnn/issues/239#issuecomment-2740740891
55 | 
56 | 
57 | #### ℹ️ Explainer updates: WebNN, MLTensor, MLGraph caching
58 | 
59 | Discuss recent explainer updates and suggested improvements.
60 | 
61 | - ☰ WebNN explainer https://github.com/webmachinelearning/webnn/blob/main/explainer.md
62 |    - ⨀ https://github.com/webmachinelearning/webnn/issues/840
63 | - ☰ MLTensor explainer https://github.com/webmachinelearning/webnn/blob/main/mltensor-explainer.md
64 |    - ⛙ https://github.com/webmachinelearning/webnn/pull/844
65 | - ☰ Caching mechanism for MLGraph explainer
66 | 
67 | 
68 | #### ℹ️ Query supported devices
69 | 
70 | Discuss the query supported devices feature, now split in two:
71 | 
72 | - ⨀ Before graph compilation https://github.com/webmachinelearning/webnn/issues/815
73 | - ⨀ After graph compilation https://github.com/webmachinelearning/webnn/issues/836
74 | 
75 | #### ℹ️ Core operator set
76 | 
77 | Revisit our core operator set effort that aims to identify current primitive gaps by mapping compositional fundamentals to WebNN operators.
78 | 
79 | Discuss any new information on rounding behavior across backends to understand feasibility for inclusion into the core operator set to help with e.g. quantization decomposition.
80 | 
81 | - ⨀ https://github.com/webmachinelearning/webnn/issues/573
82 | - ☰ [Machine Learning Operator Mapping - All Raw Operators](https://onedrive.live.com/edit?id=EE82F5C6F06C7371!345450&resid=EE82F5C6F06C7371!345450&ithint=file%2Cxlsx&authkey=!AK8f-RDTleqlLXE&wdo=2&cid=ee82f5c6f06c7371)
83 | 


--------------------------------------------------------------------------------
/telcons/2025-05-26-cg-agenda.md:
--------------------------------------------------------------------------------
 1 | ## WebML CG Teleconference [EU-APAC] – 26 May 2025 - 07:00-08:00 UTC
 2 | 
 3 | > [!IMPORTANT]
 4 | >This is the EU and APAC timezone friendly WebML CG Teleconference option.
 5 | >
 6 | >We alternate between the AMER-APAC (Tue/Wed 00:00-01:00 UTC) and this EU-APAC (Mon 07:00-08:00 UTC) option to cater to our geographically diversified participants. Feel free to join the option that fits your schedule better.
 7 | 
 8 | <details open><summary>See the timezone table ...</summary>
 9 | <table>
10 | <tr><td> San Francisco <td> Mon, 26 May 2025 <td> 00:00 💤
11 | <tr><td> Boston <td> Mon, 26 May 2025 <td> 03:00 💤
12 | <tr><td> London <td> Mon, 26 May 2025 <td> 08:00  
13 | <tr><td> Berlin <td> Mon, 26 May 2025 <td> 09:00 
14 | <tr><td> Helsinki <td> Mon, 26 May 2025 <td> 10:00 
15 | <tr><td> Shanghai <td> Mon, 26 May 2025 <td> 15:00
16 | <tr><td> Tokyo <td> Mon, 26 May 2025 <td> 16:00
17 | <tr><td> UTC <td> Mon, 26 May 2025 <td> 07:00 UTC
18 | </table>
19 | 
20 | Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20250526T07
21 | </details>
22 | 
23 | ### Logistics
24 | 
25 | * Chair: Anssi
26 | * Scribe: Anssi
27 | * Zoom joining instructions: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html
28 | * Minutes: [2025-05-26-cg-minutes.md](/telcons/2025-05-26-cg-minutes.md)
29 | 
30 | ### Agenda
31 | 
32 | #### Prompt API implementation experience from AiBrow
33 | 
34 | 🤝 Guest speakers: Rhys & Tom of the AiBrow project
35 | 
36 | An intro & learnings from AiBrow, a Chrome extension that implements the Prompt API. AiBrow uses llama.cpp web extension, WebGPU API, and native Chrome AI engine, supports custom models.
37 | 
38 | - https://github.com/axonzeta/aibrow
39 | 
40 | #### New proposals
41 | 
42 | Introduce new and review the existing proposals pipeline:
43 | 
44 | - ✨ Web AI for Time Series https://github.com/webmachinelearning/proposals/issues/10
45 | - ✨ Local Inference Web extension https://github.com/webmachinelearning/proposals/issues/9
46 | 
47 | Discuss other possible future work candidates: toxicity detection, sentiment analysis, PII detection ...
48 | 
49 | #### Proofreader API kick off
50 | 
51 | Kick off work on the Proofreader API deliverable recently adopted by the group. Recap the scope of work, discuss priority areas for community contributions.
52 | 
53 | - https://github.com/webmachinelearning/proofreader-api
54 | 
55 | #### Prompt API security and privacy
56 | 
57 | Review the initial responses to the Security and Privacy Self-Review.
58 | - https://github.com/webmachinelearning/prompt-api/blob/main/security-privacy-questionnaire.md
59 | 
60 | #### AOB
61 | 
62 | Time permitting, we'll discuss additional topics proposed by the participants.
63 | 


--------------------------------------------------------------------------------