codecov-commenter
commented
5 months ago Codecov Report
Attention: 2 lines in your changes are missing coverage. Please review.
Comparison is base (
72f713d) 75.58% compared to head (fb0224d) 76.08%.
| Files | Patch % | Lines |
|---|---|---|
| garden_ai/entrypoints.py | 83.33% | 2 Missing :warning: |
:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.
Additional details and impacted files
```diff @@ Coverage Diff @@ ## main #410 +/- ## ========================================== + Coverage 75.58% 76.08% +0.49% ========================================== Files 45 46 +1 Lines 2757 2789 +32 ========================================== + Hits 2084 2122 +38 + Misses 673 667 -6 ```:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
This came up in the course of #402
Overview
DLHub wraps user input to models in a dictionary before sending it along to Globus Compute. Similarly, when receiving output from a successful run, DLHub unwraps the output from a nested tuple and returns the inner output to the user.
This PR makes it so that Garden will do the same for the legacy DLHub models we're migrating.
Discussion
This is a little kludgey but good enough for now. We might be able to do something a little fancier where we add this code to the DLHub functions registered with Globus Compute if/when we want to get rid of this little hack.
Testing
I tested this against sample input and output I got from a DLHub model that I migrated to Garden dev. I also added new unit tests. We will still need to test this end to end in the context of the full migration dry run on dev.
Documentation
None
📚 Documentation preview 📚: https://garden-ai--410.org.readthedocs.build/en/410/