-
Notifications
You must be signed in to change notification settings - Fork 607
Update MemoryPlanning Verifier to only assume model has user input if it has at least one tensor input #10617
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update MemoryPlanning Verifier to only assume model has user input if it has at least one tensor input #10617
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/10617
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ❌ 1 New FailureAs of commit 8a87967 with merge base 9d899c9 ( NEW FAILURE - The following job has failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
Didn't find following labels among repository labels: Release notes: exir |
@pytorchbot label "release notes: exir" |
Started CI (Need to do for new, external contributions). Rebase please. Also marked it as Ready to signal reviewers that it's, well, ready for a review. Thanks for the contribution. Welcome to Executorch 😊 |
Thank you, I just rebased. I also fixed the linting issues which the CI tests identified. [Sorry, I messed this up, hence the subsequent force-push.] |
… it has at least one tensor input MemoryPlanning verifier currently blows up if all the user inputs are prims. This change means its helper function _do_user_inputs_exist returns false if all its inputs are prims. It only returns true if at least one input is a tensor.
fbb360a
to
19371e3
Compare
cc @JacobSzwejbka can you take a look? 🙏 |
@JacobSzwejbka gentle ping |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I dont see any obvious issue with the code, the all red ci though made me think it was still under iteration.
Many thanks for looking at this @JacobSzwejbka. Is there anything I can do now to help get this PR closed? |
Kicked off CI jobs. |
I don't think there's any other comment so approving |
I'm so sorry @larryliu0820, I accidentally merged the branch which seems to have stopped the running workflows. Could you please kick off the CI jobs again? |
Please correct me if I'm wrong, but I believe the failing check (pull / android / run-emulator (pull_request)) is unrelated to my change. It seems to be failing in other PRs too. For example: has the same failed check here, with the same failure output:
What is the best way to proceed to get this merged? |
Yeah unrelated. Merging. Thanks for the PR! |
Fantastic, thank you so much for your help and patience throughout! This was my first code contribution to a major open-source project. 😄 |
Fixes #10522
Summary
MemoryPlanning verifier currently blows up if all the user inputs are prims. #10522 suggested:
This PR implements this suggestion, with accompanying unit tests.
_do_user_inputs_exist
now returns True if it has at least one tensor input, and False otherwise.Test plan
Added unit tests to
test_memory_planning.py
and ran them withPlease note you must comment out the line 69 from
/pytest.ini
for this to work:On my machine, I also had to comment out line 60 from
test_memory_planning.py
for the test to run without errors.[The tests I wrote aren't dependent on this library.]