Wrap server generated yaml files inside et_operator_library #5777
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Summary: This diff is to make the model config codegen flow work on Edge MLHub by enabling the `et_operator_library` API to take a .yaml file. - The code changes in `fbcode/executorch/codegen/codegen.bzl` was copied from a stale diff D61824614 made by Mengwei. That one didn't landed because the owner of SceneX/v5002 model didn't have time to test it. - The code changes in `fbcode/executorch/build/fb/clients.bzl` is to make auto-generated BUCKs under the directory `fbsource//xplat/executorch_model/...` have visibility to the `et_operator_library` API. By design, this is the codegen flow: 1. Use click the "Generate ET model config" button on Edge ML Hub; 2. Two files will be generated, one is a .yaml with a list of kernels, the other is a BUCK file under path `xplat/executorch_model/build/...` that calls `et_operator_library` and pass in the .yaml as a parameter; 3. A bot would accept and land the diff if CI passes; 4. User can run inference. Next I will modify the codegen script so that it creates the .yaml and BUCK in the correct format for this flow to work E2E. Differential Revision: D63305843
facebook-github-bot
added
the
CLA Signed
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/5777
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 1ba447d with merge base b60fa71 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D63305843 |
Olivia-liu
added a commit
to Olivia-liu/executorch-1
that referenced
this pull request
Oct 1, 2024
…5777) Summary: Pull Request resolved: pytorch#5777 ## This diff This diff is to make the model config codegen flow work on Edge MLHub by enabling the `et_operator_library` API to take a .yaml file. - The code changes in `fbcode/executorch/codegen/codegen.bzl` was copied from a stale diff D61824614 made by Mengwei. That one didn't landed because the owner of SceneX/v5002 model didn't have time to test it. ## Context By design, this is the codegen flow: 1. Use click the "Generate ET model config" button on Edge ML Hub; 2. Two files will be generated, one is a .yaml with a list of kernels, the other is a BUCK file under path `xplat/executorch_model/build/...` that calls `et_operator_library` and pass in the .yaml as a parameter. See D63667456 for an example; 3. A bot would accept and land the diff if CI passes; 4. User can run inference. Differential Revision: D63305843
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
This diff is to make the model config codegen flow work on Edge MLHub by enabling the
et_operator_libraryAPI to take a .yaml file.The code changes in
fbcode/executorch/codegen/codegen.bzlwas copied from a stale diff D61824614 made by Mengwei. That one didn't landed because the owner of SceneX/v5002 model didn't have time to test it.The code changes in
fbcode/executorch/build/fb/clients.bzlis to make auto-generated BUCKs under the directoryfbsource//xplat/executorch_model/...have visibility to theet_operator_libraryAPI.By design, this is the codegen flow:
xplat/executorch_model/build/...that callset_operator_libraryand pass in the .yaml as a parameter;Next I will modify the codegen script so that it creates the .yaml and BUCK in the correct format for this flow to work E2E.
Differential Revision: D63305843