datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
populace-us
populace-us
The populace-built US population: a calibrated synthetic microdataset for
PolicyEngine-US, built by the
populace stack entirely from
primary sources — the enhanced CPS appears only as the benchmark this file is
scored against, never as a build input. It loads anywhere the enhanced CPS
loads (an API-compatible alternative population), with its own calibrated
weights — and its own strengths and gaps, both documented below.
Load it
pip install… See the full description on the dataset page: https://huggingface.co/datasets/policyengine/populace-us.microcosm-evaluation
Microcosm cross-dataset evaluation bundles
Immutable frontend bundles (cross_dataset.frontend_bundle.v1) consumed by the calibration
diagnostics dashboard's Cross-dataset page: every source (a model × dataset pair, or a standalone
dataset) scored against the same Chronicle facts for a jurisdiction. Produced by
PolicyEngine/calibration-diagnostics scripts/run_full_chronicle_evaluation.py +
publish_cross_dataset_frontend_bundle.py; every partition carries the run and snapshot ids… See the full description on the dataset page: https://huggingface.co/datasets/policyengine/microcosm-evaluation.microcosm-us-sources
Microcosm US build sources
Processed Current Population Survey Annual Social and Economic Supplement (CPS ASEC) files that the PolicyEngine Microcosm US build reads as inputs to its base stage (tools/build_us_puf_support_base.py --asec-h5 YEAR=PATH in PolicyEngine/microcosm).
They are mirrored here unchanged so that the build's inputs do not exist on one machine only. Microcosm pins each file by repository revision, SHA-256 and size, and refuses a file whose digest differs.… See the full description on the dataset page: https://huggingface.co/datasets/policyengine/microcosm-us-sources.microplex-us-deployed-datasets
Microplex US Deployed Datasets
Stable deployed PolicyEngine-US H5 artifacts produced by Microplex-US.
Layout
staging/<run_id>/policyengine_us.h5
staging/<run_id>/manifest.json
policyengine_us.h5
manifest.json
The staging/<run_id>/... paths are immutable once published. The root policyengine_us.h5 and manifest.json point to the promoted current dataset.
microplex-us-diagnostics
Microplex US Diagnostics
Stable diagnostics registry for Microplex-US artifact bundles.
Layout
runs/<run_id>/manifest.json
runs/<run_id>/policyengine_native_scores.json
runs/<run_id>/pe_us_data_rebuild_native_audit.json
runs/<run_id>/pe_native_target_diagnostics.json
latest.json
run_registry.jsonl
The runs/<run_id>/... paths are immutable once published. latest.json and run_registry.jsonl are mutable discovery files.