Expose negate_key_metric in FedAvg recipes

[chesterxgchen]

Summary

• Add negate_key_metric pass-through to the base FedAvg recipe and framework-specific FedAvg recipe wrappers.
• Wire the flag to IntimeModelSelector so recipe users can select lower-is-better metrics such as losses without reporting a separate negated metric from clients.
• Add focused recipe tests for PyTorch and NumPy recipe pass-through.

Testing

• python -m pytest tests/unit_test/recipe/fedavg_recipe_test.py -q
• git diff --check

[chesterxgchen]

/build

[greptile-apps]

Greptile Summary

This PR exposes negate_key_metric through the entire FedAvg recipe stack so users can train on lower-is-better metrics (e.g., losses) without manually negating them on the client side.

• Base layer: negate_key_metric: bool = False is added to _FedAvgValidator, FedAvgRecipe.__init__, and BaseFedJob.__init__, where it is wired into the auto-created IntimeModelSelector(key_metric=key_metric, negate_key_metric=negate_key_metric).
• Framework wrappers: All six recipe variants (PyTorch, NumPy, sklearn, KMeans, SVM, TensorFlow) receive and forward the flag consistently to the base recipe.
• Tests: New parametric tests cover all six framework paths, including a TF stub that bypasses the TensorFlow import requirement at test time.

Confidence Score: 5/5

Safe to merge — the change is a straightforward additive parameter thread-through with no logic changes to existing code paths.

The flag is forwarded correctly at every layer (validator → recipe → BaseFedJob → IntimeModelSelector), IntimeModelSelector already supported the parameter, and all six framework recipe variants are covered by new tests. No existing behavior is altered when negate_key_metric is left at its default value of False.

No files require special attention.

Important Files Changed

Filename	Overview
nvflare/recipe/fedavg.py	Adds negate_key_metric to _FedAvgValidator, FedAvgRecipe.init, and the BaseFedJob instantiation; correctly stored as self.negate_key_metric and forwarded.
nvflare/job_config/base_fed_job.py	Adds negate_key_metric to BaseFedJob and correctly passes it to IntimeModelSelector when auto-created; docstring correctly notes the flag is ignored when a custom model_selector is supplied.
nvflare/app_opt/pt/recipes/fedavg.py	Clean pass-through of negate_key_metric to UnifiedFedAvgRecipe with matching docstring update.
nvflare/app_common/np/recipes/fedavg.py	Clean pass-through of negate_key_metric to UnifiedFedAvgRecipe with matching docstring update.
nvflare/app_opt/sklearn/recipes/fedavg.py	Adds negate_key_metric after key_metric in signature and forwards it to super().init(); consistent with other wrappers.
nvflare/app_opt/sklearn/recipes/kmeans.py	Adds negate_key_metric at the end of the constructor signature and forwards it to super().init(); docstring updated.
nvflare/app_opt/sklearn/recipes/svm.py	Adds negate_key_metric at the end of the constructor signature and forwards it to super().init(); docstring updated.
nvflare/app_opt/tf/recipes/fedavg.py	Adds negate_key_metric after key_metric and forwards it to super().init(); docstring updated correctly.
tests/unit_test/recipe/fedavg_recipe_test.py	Adds six new tests covering negate_key_metric pass-through for all framework variants (PT, NumPy, sklearn, KMeans, SVM, TF); TF test uses a careful monkeypatch to avoid requiring TensorFlow at test time.

Sequence Diagram

sequenceDiagram
    participant U as User Code
    participant R as FedAvgRecipe (framework-specific)
    participant BR as FedAvgRecipe (base / unified)
    participant BFJ as BaseFedJob
    participant IMS as IntimeModelSelector

    U->>R: "FedAvgRecipe(key_metric="val_loss", negate_key_metric=True)"
    R->>BR: "super().__init__(key_metric=key_metric, negate_key_metric=negate_key_metric)"
    BR->>BR: _FedAvgValidator stores negate_key_metric
    BR->>BR: "self.negate_key_metric = v.negate_key_metric"
    BR->>BFJ: "BaseFedJob(key_metric=self.key_metric, negate_key_metric=self.negate_key_metric)"
    BFJ->>IMS: "IntimeModelSelector(key_metric="val_loss", negate_key_metric=True)"
    IMS-->>BFJ: selector registered on server
    BFJ-->>BR: job configured
    BR-->>R: recipe ready
    R-->>U: "recipe instance (recipe.negate_key_metric == True)"

Loading

_{Reviews (4): Last reviewed commit: "Address negate metric recipe review comm..." | Re-trigger Greptile}

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 62.85%. Comparing base (d714126) to head (98a994f).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #4953      +/-   ##
==========================================
+ Coverage   62.83%   62.85%   +0.02%     
==========================================
  Files         995      995              
  Lines       97868    97870       +2     
==========================================
+ Hits        61496    61520      +24     
+ Misses      36372    36350      -22     

Flag	Coverage Δ
unit-tests	62.85% <100.00%> (+0.02%)	⬆️

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

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[chesterxgchen]

/build

[chesterxgchen]

/build

[chesterxgchen]

/build