Skip to content
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

Split out model vs microbatch execution #10737

Merged
merged 3 commits into from
Sep 19, 2024

Conversation

QMalcolm
Copy link
Contributor

Resolves #N/A

Description

In #10677 we added microbatch model execution logic to the ModelRunner.execute. In doing so we changed the order of operations of some normal model execution steps. This PR splits the logic so that we can return the order of operations for normal model execution back to what it was previously while still supporting microbatch model execution. We don't know if the change of order operations that happened in #10677 was problematic, but moving it back is being done out caution.

Checklist

  • I have read the contributing guide and understand what's expected of me.
  • I have run this code in development, and it appears to resolve the stated issue.
  • This PR includes tests, or tests are not required or relevant for this PR.
  • This PR has no interface changes (e.g., macros, CLI, logs, JSON artifacts, config files, adapter interface, etc.) or this PR has already received feedback and approval from Product or DX.
  • This PR includes type annotations for new and modified functions.

@QMalcolm QMalcolm added the Skip Changelog Skips GHA to check for changelog file label Sep 18, 2024
@QMalcolm QMalcolm requested a review from a team as a code owner September 18, 2024 20:54
@cla-bot cla-bot bot added the cla:yes label Sep 18, 2024
…l execution

In commit 8fe5ea1 the adding of the relationship
to the adapter cache got moved to _before_ the post hook during model execution. This
was done because of some wonkiness of having the microbatch model and model execution
logic in the same place. Now that we've split the relevant logic, we can move the caching
back to it's original spot in the order of operations. We don't think this actually changes
anything, but we're moving it back juuuust in case because this is a very hot code path.
@QMalcolm QMalcolm force-pushed the qmalcolm--split-out-model-vs-microbatch-execution branch from dfb5076 to 89f1c67 Compare September 18, 2024 20:56
Copy link

codecov bot commented Sep 18, 2024

Codecov Report

Attention: Patch coverage is 94.44444% with 1 line in your changes missing coverage. Please review.

Project coverage is 88.95%. Comparing base (f8bfd32) to head (5251a88).
Report is 1 commits behind head on main.

Additional details and impacted files
@@            Coverage Diff             @@
##             main   #10737      +/-   ##
==========================================
- Coverage   89.00%   88.95%   -0.05%     
==========================================
  Files         181      181              
  Lines       22964    22971       +7     
==========================================
- Hits        20440    20435       -5     
- Misses       2524     2536      +12     
Flag Coverage Δ
integration 86.13% <94.44%> (-0.13%) ⬇️
unit 62.39% <11.11%> (-0.02%) ⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Components Coverage Δ
Unit Tests 62.39% <11.11%> (-0.02%) ⬇️
Integration Tests 86.13% <94.44%> (-0.13%) ⬇️

@QMalcolm
Copy link
Contributor Author

The failing unit test, test_false_evaluation_triggers_snowplow_tracking, is also failing on main right now. Looks like this was a recently added test via #10721. Unfortunately, the failing unit test is blocking merges.

@QMalcolm QMalcolm merged commit 3308a43 into main Sep 19, 2024
60 of 62 checks passed
@QMalcolm QMalcolm deleted the qmalcolm--split-out-model-vs-microbatch-execution branch September 19, 2024 03:13
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
cla:yes Skip Changelog Skips GHA to check for changelog file
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants