Reduce memory requirement on export_llama tests with no params#11175
Closed
jackzhxng wants to merge 5 commits into
Closed
Reduce memory requirement on export_llama tests with no params#11175jackzhxng wants to merge 5 commits into
jackzhxng wants to merge 5 commits into
Conversation
[ghstack-poisoned]
This was referenced May 28, 2025
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/11175
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New FailureAs of commit 385c5e1 with merge base d5c4ba7 ( NEW FAILURE - The following job has failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This was referenced May 28, 2025
facebook-github-bot
added
the
CLA Signed
…rams" [ghstack-poisoned]
Contributor
Author
|
@jackzhxng has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
larryliu0820
approved these changes
May 28, 2025
…rams" For some reason, after the previous PR in the stack, test_export_llama_lib was ooming on gh actions CI and I couldn't really figure out why, since I profiled the running memory on the test both before and after they were the same. This addresses fixes the ci oom, and I've been meaning to do this anyway since if we are loading a transformer without params specified, we likely just want to test some basic functionality, so a 1 layer makes more sense than a 8 layer default. Have made sure code elsewhere is not relying on this 8 layer default atm. Differential Revision: [D75498713](https://our.internmc.facebook.com/intern/diff/D75498713) [ghstack-poisoned]
jackzhxng
added a commit
that referenced
this pull request
May 29, 2025
Pull Request resolved: #11175 @imported-using-ghimport Differential Revision: [D75498713](https://our.internmc.facebook.com/intern/diff/D75498713/) ghstack-source-id: 286576048
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D75498713 |
…rams" For some reason, after the previous PR in the stack, test_export_llama_lib was ooming on gh actions CI and I couldn't really figure out why, since I profiled the running memory on the test both before and after they were the same. This addresses fixes the ci oom, and I've been meaning to do this anyway since if we are loading a transformer without params specified, we likely just want to test some basic functionality, so a 1 layer makes more sense than a 8 layer default. Have made sure code elsewhere is not relying on this 8 layer default atm. Differential Revision: [D75498713](https://our.internmc.facebook.com/intern/diff/D75498713) [ghstack-poisoned]
jackzhxng
added a commit
that referenced
this pull request
Jun 2, 2025
Pull Request resolved: #11175 @imported-using-ghimport Differential Revision: [D75498713](https://our.internmc.facebook.com/intern/diff/D75498713/) ghstack-source-id: 287631950
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D75498713 |
…rams" For some reason, after the previous PR in the stack, test_export_llama_lib was ooming on gh actions CI and I couldn't really figure out why, since I profiled the running memory on the test both before and after they were the same. This addresses fixes the ci oom, and I've been meaning to do this anyway since if we are loading a transformer without params specified, we likely just want to test some basic functionality, so a 1 layer makes more sense than a 8 layer default. Have made sure code elsewhere is not relying on this 8 layer default atm. Differential Revision: [D75498713](https://our.internmc.facebook.com/intern/diff/D75498713) [ghstack-poisoned]
jackzhxng
added a commit
that referenced
this pull request
Jun 3, 2025
Pull Request resolved: #11175 @imported-using-ghimport Differential Revision: [D75498713](https://our.internmc.facebook.com/intern/diff/D75498713/) ghstack-source-id: 287798908
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D75498713 |
jackzhxng
had a problem deploying
to
cherry-pick-bot
GitHub Actions
Failure
jackzhxng
pushed a commit
to jackzhxng/executorch
that referenced
this pull request
Jun 3, 2025
…ch#11175) Summary: Pull Request resolved: pytorch#11175 Reviewed By: larryliu0820 Differential Revision: D75498713
facebook-github-bot
pushed a commit
that referenced
this pull request
Jun 5, 2025
Differential Revision: D75498713 Pull Request resolved: #11339
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
Stack from ghstack (oldest at bottom):
For some reason, after the previous PR in the stack, test_export_llama_lib was ooming on gh actions CI and I couldn't really figure out why, since I profiled the running memory on the test both before and after they were the same. This addresses fixes the ci oom, and I've been meaning to do this anyway since if we are loading a transformer without params specified, we likely just want to test some basic functionality, so a 1 layer makes more sense than a 8 layer default. Have made sure code elsewhere is not relying on this 8 layer default atm.
Differential Revision: D75498713