Web Neural Network API 2024-01-29 > 2024-02-29

[anssiko]

Name of your specification

Web Neural Network API

URL of your specification

https://www.w3.org/TR/webnn/

Does your specification include an Internationalization Considerations section?

• My specification includes an Internationalization Considerations section

When does the review need to be finished?

2024-02-29

What has changed since any previous review?

Since the initial Candidate Recommendation Snapshot the Working Group has gathered further implementation experience and added new operations and data types needed for well-known transformers to support generative AI use cases. In addition, the group has removed select features informed by this implementation experience: higher-level operations that can be expressed in terms of lower-level primitives in a performant manner, and support for synchronous execution. The group has also updated the specification to use modern authoring conventions to improve interoperability and precision of normative definitions and is developing a new feature, a backend-agnostic storage type, to improve performance and interoperability between the WebNN, WebGPU APIs and purpose-built hardware for ML.

Please point to the results of your own self-review

webmachinelearning/webnn#262

Where and how should the i18n WG raise issues?

https://github.com/webmachinelearning/webnn/issues

Pointer to any explainer for the spec

https://github.com/webmachinelearning/webnn/blob/main/explainer.md

Other comments

We're looking to publish a new CR Snapshot of the Web Neural Network API in Q1'24.

We have considered your earlier review feedback from #179 and continue to evaluate our work using your Internationalization Checklist, with our latest status tracked in webmachinelearning/webnn#262

[bert-github]

I reviewed the spec and did not find any issues.

About the strings mentioned in the self-review: Some data sets are given names (arbitrary, non-empty DOMStrings, chosen by the programmer) under which they are stored in a program's memory and with which they can be retrieved again. These strings do not appear in a UI, do not need to be localized and are not compared, other than for equality (presumably using the equivalent of JavaScript's ‘==’ or ‘===’ operator). This seems harmless.

[anssiko]

Thank you @bert-github and i18n folks for your review. Feel free to proceed to close this issue at your convenience.

We will get back to you should the WG make future changes that would benefit from your review.