RESCHEDULED: WebML CG Teleconference – 26 January 2022 - 05:00-06:00 UTC+0

Hi All,

We're rescheduling this meeting from 26 January 2022 to 9 February 2022 to allow our Australian friends enjoy Australia Day! đŸ‡ĻđŸ‡ē

I've redirected the agenda to the new date. Updated invitations have been sent.

Sometimes calendar apps get confused with updates, so please manually check that the invite moves to 9 February 2022. If you still run into issues, you can manually export the updated invite from


-Anssi (WebML WG/CG chair)

On 19. Jan 2022, at 11.24, Kostiainen, Anssi <<>> wrote:

Latest version:

WebML CG Teleconference – 26 January 2022 - 05:00-06:00 UTC+0
See the timezone table ...

  *   Chair: Anssi
  *   Scribe: ? (howto<>)
  *   IRC: irc://

  *   IRC web client:

  *   Call-in details:

  *   Minutes:

<>🔄 Model Loader API
<>ℹī¸ Benchmark discussion

  *   Continued from 12 Jan meeting: open questions, proposed next steps

<>ℹī¸ Hardware acceleration

  *   What will the interface look like when hardware acceleration is supported?
     *   What if the data of a tensor is on GPU/TPU?
     *   Shall we let users know which ML accelerator is available on the platform? (Performance v.s. Privacy)

<>ℹī¸ Versioning

  *   How to version the API? (supported ops and formats)

<>ℹī¸ Streaming inputs

  *   Dedicated interface for streaming inputs (e.g. buffered inputs for video, audio)?

<>ℹī¸ Model format

  *   Prototype vs standard format requirements
  *   "The Model Loader API needs a standard format supported across browsers and devices for broad interoperability." condition for standards adoption into WebML WG

<>ℹī¸ References

  *   WebML CG Teleconference – 12 Jan 2022 minutes<>
  *   Updated explainer<> and early spec draft<>
  *   Chromium prototype<>

Received on Thursday, 20 January 2022 09:36:25 UTC