MODELS,
COMPARED.
Preprocessing, train-test split, linear models, model comparison, metrics and explainability. The uploaded project remains primary while patterns from scikit-learn/scikit-learn are adapted into a static, installable and offline-capable browser edition.
ModelForge
Preprocessing, train-test split, linear models, model comparison, metrics and explainability. Essential controls render immediately without an external origin.
Built to extend.
The first render is static and deterministic. Native runtimes and upstream services remain optional adapters.
scikit-learn/scikit-learn
scikit-learn-inspired pipelines, estimators, cross-validation and evaluation workflow.
Original project retained
The uploaded source is preserved in developer-source. Large archives are split below the Cloudflare per-file limit and include reconstruction instructions.
Upstream evidence
The real archive supplied through extra1, extra2, or Joplin was inspected. SHA-256, file inventory, README excerpt, license evidence and adapter blueprint are included under integration.
Direct upload.
Upload this individual ZIP as prebuilt assets. No build command, Worker, proxy or server origin is required.
01 Upload
index.html is directly at archive root.
02 Verify
Open /health.txt and then the main page.
03 Extend
Connect an optional backend only after the static shell works.