Intent to Prototype: Web Machine Learning: Model Loader API
Jonathan Bingham
Contact emails
bing...@google.com, hong...@google.com, nap...@google.com
Explainer
https://github.com/webmachinelearning/model-loader/blob/master/explainer.md
Specification
https://webmachinelearning.github.io/model-loader/#api
Summary
Run machine learning models natively, taking advantage of hardware acceleration. Currently, javascript libraries must parse a model and call WebGL, WebGPU and WASM for the individual compute operations. Providing a model loader API gains performance by optimizing in the browser and potentially taking advantage of new ML processors like TPUs.
Blink component
Motivation
Model Loader is a proposed web API to load a custom, pre-trained machine learning model in a standard format, compile it for the available hardware, and apply it to example data in JavaScript in order to perform inference, like classification, regression, or ranking. The idea is to make it as easy as possible for web developers to use a custom, pre-built machine learning model in their web app, across devices and browsers.
Performing inference locally can:
- Preserve privacy, by not shipping user data across the network
- Improve performance, by eliminating network latency, running models natively and taking advantage of hardware acceleration, including specialized hardware not available with WebGL, WebGPU or WASM.
- Provide a fallback if network access is unavailable, possibly using a smaller and lower quality model
The graph-based Web NN API and the Model Loader API are complementary approaches. We'll need to do some benchmarking to understand if there are performance differences, and get feedback from developers to see if it's valuable to offer both types of API.
Initial public proposal
https://webmachinelearning.github.io/model-loader-intro/
TAG review
TAG review status
Pending
Risks
The Model Loader API is currently included as a Tentative Specification in the Web Machine Learning Working Group Charter.
Interoperability and Compatibility
Gecko: No signals
WebKit: No signals
Chrome/TensorFlow: Positive
Intel : Collaborating in Working Group
Microsoft/ONNX: Collaborating in Working Group
PyTorch: No signals
Salesforce: Collaborating in Working Group
Debuggability
Is this feature fully tested by web-platform-tests?
No
Flag name
Requires code in //chrome?
False
Tracking bug
https://bugs.chromium.org/p/chromium/issues/detail?id=1263240
Estimated milestones
No milestones specified
Link to entry on the Chrome Platform Status
https://www.chromestatus.com/feature/6505078216196096
This intent message was generated by Chrome Platform Status.
Victor Miura
--
You received this message because you are subscribed to the Google Groups "blink-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAEK6eFxSmK6ajiQ6_7rhPspiaDXO3xM3uNJnbSgCobQ_nk%2Bxpg%40mail.gmail.com.