Refresh backend README with progress timeline (#20115)#20115
Conversation
|
@JulianCloudNTH has exported this pull request. If you are a Meta employee, you can view the originating Diff in D107742574. |
This PR needs a
|
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/20115
Note: Links to docs will display an error until the docs builds have been completed. This comment was automatically generated by Dr. CI and updates every 15 minutes. |
51dd128 to
79dca63
Compare
Summary: Update the WebGPU backend README to reflect the current state of the backend: - Add a Progress section listing milestones landed on `main` (#18808, #19963, #19964, #19981, #20036) and work in review (#20079, #20080), each linking its pull request. - Update the operator support table to include `rms_norm` and refresh the planned/roadmap list toward end-to-end LLM inference. - Update the directory structure to match the current layout. Docs-only change; no code or build impact. Reviewed By: psiddh Differential Revision: D107742574
Summary:
Update the WebGPU backend README to reflect the current state of the
backend:
main([ETVK] WebGPU runtime #18808, [ExecuTorch][WebGPU] Add rms_norm op #19963, [ExecuTorch][WebGPU] Enable backend test suite + x86 CI #19964, Generate *_wgsl.h embedded shaders from *.wgsl (#19981) #19981, [ExecuTorch][WebGPU] Fix CI: link shared wgpu-native to avoid LTO bitcode clash #20036) and work in review
([ExecuTorch][WebGPU] Switch native backend from wgpu-native to Dawn (Tint) + SwiftShader #20079, [ExecuTorch][WebGPU] Add per-pass dispatch ordering + scratch buffer tests #20080), each linking its pull request.
rms_normand refreshthe planned/roadmap list toward end-to-end LLM inference.
Docs-only change; no code or build impact.
Reviewed By: psiddh
Differential Revision: D107742574