01 Package
mlbundle
Pickle-free, verifiable model bundles.
One file with the model in its native format, a manifest, its schema and its structure. Works with scikit-learn, XGBoost, CatBoost and ONNX.
pip install mlbundlePut machine-learning models in front of people.
Open-source tools that take a trained model from a file on disk to a form anyone can fill in, with every prediction traceable and every correction kept.
Open source
Each piece works on its own. Together they cover the loop from training to feedback.
02 Describe
mlschema
Field schemas inferred from your data.
Infers validated, frontend-ready field schemas from pandas DataFrames, so a model’s inputs are typed before anyone fills them in.
pip install mlschema03 Render
mlform
Forms from schemas, as Web Components.
A schema-driven form runtime and Web Component kit for machine-learning interfaces. Drop it into any framework, or none.
npm i mlform04 Run & review
MLSuite
Trained models, turned into reviewable apps.
Upload a bundle, run predictions through a generated form, collect human feedback and export it for evaluation or retraining. Self-host it with Docker Compose.
706downloads last month
7GitHub stars
4open-source projects