foundry-tabular-adult

Binary classifier trained on Adult census income (UCI). Source rows 48842.

Data

• Dataset license: CC BY 4.0, verified on the UCI Machine Learning Repository page on 2026-10-06. Weights are a derivative of that data and are released under the same license; credit the UCI Machine Learning Repository and the dataset's authors. Code files (treepack.py, trees.mojo) are MIT.
• Split: 60/20/20 stratified, seed 42. Selection on validation only.
• Sensitive attributes present in features: sex, race, age, native-country. No fairness evaluation was performed.

Metrics (ROC AUC)

• Validation: 0.929
• Held-out test (sklearn original): 0.9282
• Held-out test (exported model, this file): 0.9282
• Max |probability difference| to sklearn on test rows: 5.25e-09

Use and limits

• Research baseline on one public dataset. Not for credit, hiring, insurance or medical decisions.
• Distribution shift, calibration and subgroup behaviour are untested.
• model.npz contains arrays only (no pickle). SHA-256 5981baa5834d6cd458042d08799f30f9332f55529f560257b76575d6e439e813, 445953 bytes.
• Inference: from treepack import Forest; Forest("model.npz").predict_proba(df). Uses a Mojo kernel when libfoundry_trees.so is built, else a numpy fallback.

CPU throughput (2 cores, Mojo kernel, includes preprocessing)

batch	sklearn rows/s	exported rows/s	ms/call
1	19	227	4.396
64	944	8218	7.788
1024	7058	18183	56.315
8192	9326	23502	348.563