dataspoof-mlpilot 2.1.0__tar.gz → 2.3.0__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/.github/workflows/CI.yml +44 -1
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/CHANGELOG.md +326 -204
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/CONTRIBUTING.md +2 -2
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/Cargo.lock +1 -1
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/Cargo.toml +1 -1
- dataspoof_mlpilot-2.3.0/LICENSE +21 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/PKG-INFO +1877 -1715
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/README.md +1861 -1707
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/SECURITY.md +2 -2
- dataspoof_mlpilot-2.3.0/docs/api/README.md +36 -0
- dataspoof_mlpilot-2.3.0/docs/api/anomaly.md +69 -0
- dataspoof_mlpilot-2.3.0/docs/api/automl.md +437 -0
- dataspoof_mlpilot-2.3.0/docs/api/causal.md +112 -0
- dataspoof_mlpilot-2.3.0/docs/api/cleaning.md +911 -0
- dataspoof_mlpilot-2.3.0/docs/api/compression.md +125 -0
- dataspoof_mlpilot-2.3.0/docs/api/core.md +459 -0
- dataspoof_mlpilot-2.3.0/docs/api/deploy.md +642 -0
- dataspoof_mlpilot-2.3.0/docs/api/environments.md +223 -0
- dataspoof_mlpilot-2.3.0/docs/api/evaluation.md +625 -0
- dataspoof_mlpilot-2.3.0/docs/api/feature_store.md +142 -0
- dataspoof_mlpilot-2.3.0/docs/api/features.md +389 -0
- dataspoof_mlpilot-2.3.0/docs/api/forecast.md +126 -0
- dataspoof_mlpilot-2.3.0/docs/api/hub.md +621 -0
- dataspoof_mlpilot-2.3.0/docs/api/ingestion.md +589 -0
- dataspoof_mlpilot-2.3.0/docs/api/insights.md +161 -0
- dataspoof_mlpilot-2.3.0/docs/api/integrations.md +855 -0
- dataspoof_mlpilot-2.3.0/docs/api/lineage.md +186 -0
- dataspoof_mlpilot-2.3.0/docs/api/media.md +344 -0
- dataspoof_mlpilot-2.3.0/docs/api/model_security.md +232 -0
- dataspoof_mlpilot-2.3.0/docs/api/monitoring.md +746 -0
- dataspoof_mlpilot-2.3.0/docs/api/operations.md +334 -0
- dataspoof_mlpilot-2.3.0/docs/api/orchestration.md +420 -0
- dataspoof_mlpilot-2.3.0/docs/api/quality.md +781 -0
- dataspoof_mlpilot-2.3.0/docs/api/registry.md +471 -0
- dataspoof_mlpilot-2.3.0/docs/api/scale.md +83 -0
- dataspoof_mlpilot-2.3.0/docs/api/secrets_security.md +171 -0
- dataspoof_mlpilot-2.3.0/docs/api/splitting.md +120 -0
- dataspoof_mlpilot-2.3.0/docs/api/stats.md +195 -0
- dataspoof_mlpilot-2.3.0/docs/api/versioning.md +180 -0
- dataspoof_mlpilot-2.3.0/docs/api/vision.md +491 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/architecture.md +32 -32
- dataspoof_mlpilot-2.3.0/docs/enterprise-gap-audit.md +233 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/deployment.md +72 -64
- dataspoof_mlpilot-2.3.0/docs/guide/enterprise-operations.md +246 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/media.md +48 -32
- dataspoof_mlpilot-2.3.0/docs/guide/quantization.md +143 -0
- dataspoof_mlpilot-2.3.0/docs/guide/vision.md +155 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/index.md +105 -105
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/installation.md +90 -89
- dataspoof_mlpilot-2.3.0/docs/reference/compression.md +5 -0
- dataspoof_mlpilot-2.3.0/docs/reference/model_privacy.md +3 -0
- dataspoof_mlpilot-2.3.0/docs/reference/operations.md +11 -0
- dataspoof_mlpilot-2.3.0/docs/reference/orchestration.md +9 -0
- dataspoof_mlpilot-2.3.0/examples/enterprise_walkthrough.ipynb +1325 -0
- dataspoof_mlpilot-2.3.0/examples/titanic_walkthrough.ipynb +1313 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mkdocs.yml +149 -138
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/__init__.py +66 -60
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/automl/automl.py +630 -630
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/automl/models.py +16 -1
- dataspoof_mlpilot-2.3.0/mlpilot/backup.py +217 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/cicd.py +437 -396
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/cleaners.py +8 -2
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/cleaning.py +1367 -1368
- dataspoof_mlpilot-2.3.0/mlpilot/compression.py +221 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/__init__.py +21 -18
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/cli.py +419 -284
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/export.py +317 -285
- dataspoof_mlpilot-2.3.0/mlpilot/deploy/quantize.py +341 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/server.py +100 -6
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/wsgi.py +59 -53
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/features/text.py +156 -133
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/hub/app.py +854 -700
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/hub/db.py +49 -1
- dataspoof_mlpilot-2.3.0/mlpilot/labeling.py +229 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/lineage.py +1 -1
- dataspoof_mlpilot-2.3.0/mlpilot/media/__init__.py +45 -0
- dataspoof_mlpilot-2.3.0/mlpilot/media/benchmark.py +172 -0
- dataspoof_mlpilot-2.3.0/mlpilot/media/conditions.py +414 -0
- dataspoof_mlpilot-2.3.0/mlpilot/media/faces.py +343 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/media/image.py +365 -353
- dataspoof_mlpilot-2.3.0/mlpilot/media/plates.py +292 -0
- dataspoof_mlpilot-2.3.0/mlpilot/model_privacy.py +132 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/monitoring/__init__.py +6 -1
- dataspoof_mlpilot-2.3.0/mlpilot/monitoring/incidents.py +265 -0
- dataspoof_mlpilot-2.3.0/mlpilot/orchestration/__init__.py +30 -0
- dataspoof_mlpilot-2.3.0/mlpilot/orchestration/engine.py +477 -0
- dataspoof_mlpilot-2.3.0/mlpilot/orchestration/goal.py +240 -0
- dataspoof_mlpilot-2.3.0/mlpilot/orchestration/tools.py +413 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/quality/__init__.py +29 -25
- dataspoof_mlpilot-2.3.0/mlpilot/quality/observability.py +303 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/registry/__init__.py +16 -15
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/registry/_store.py +3 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/registry/models.py +356 -280
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/steps.py +930 -926
- dataspoof_mlpilot-2.3.0/mlpilot/supply_chain.py +260 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/pyproject.toml +235 -222
- dataspoof_mlpilot-2.3.0/scripts/gen_api_ref.py +283 -0
- dataspoof_mlpilot-2.3.0/tests/assets/astronaut.jpg +0 -0
- dataspoof_mlpilot-2.3.0/tests/assets/coffee.jpg +0 -0
- dataspoof_mlpilot-2.3.0/tests/test_backup.py +90 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_cleaners.py +9 -0
- dataspoof_mlpilot-2.3.0/tests/test_fault_injection.py +124 -0
- dataspoof_mlpilot-2.3.0/tests/test_hub_tenancy.py +111 -0
- dataspoof_mlpilot-2.3.0/tests/test_incidents.py +85 -0
- dataspoof_mlpilot-2.3.0/tests/test_labeling.py +67 -0
- dataspoof_mlpilot-2.3.0/tests/test_model_governance.py +59 -0
- dataspoof_mlpilot-2.3.0/tests/test_model_privacy.py +49 -0
- dataspoof_mlpilot-2.3.0/tests/test_observability.py +80 -0
- dataspoof_mlpilot-2.3.0/tests/test_orchestration.py +205 -0
- dataspoof_mlpilot-2.3.0/tests/test_quantization.py +244 -0
- dataspoof_mlpilot-2.3.0/tests/test_serving_hardening.py +78 -0
- dataspoof_mlpilot-2.3.0/tests/test_supply_chain.py +77 -0
- dataspoof_mlpilot-2.3.0/tests/test_vision.py +273 -0
- dataspoof_mlpilot-2.1.0/mlpilot/media/__init__.py +0 -19
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/.dockerignore +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/.github/workflows/docs.yml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/.gitignore +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/BUILD.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/Chart.yaml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/templates/_helpers.tpl +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/templates/deployment.yaml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/templates/ingress.yaml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/templates/pdb.yaml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/templates/pvc.yaml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/templates/secret.yaml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/templates/service.yaml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/templates/serviceaccount.yaml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/helm/mlpilot-hub/values.yaml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/hub/Dockerfile +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/hub/docker-compose.yml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/monitoring/docker-compose.yml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/monitoring/grafana/dashboards/mlpilot-model.json +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/monitoring/grafana/provisioning/dashboards/mlpilot.yml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/monitoring/grafana/provisioning/datasources/prometheus.yml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/monitoring/prometheus/alerts.yml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/monitoring/prometheus/alerts_test.yml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/monitoring/prometheus/prometheus.yml +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/opa/policies/mlpilot/authz.rego +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/opa/policies/mlpilot/config/data.json +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/opa/policies/mlpilot/policy_test.rego +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/opa/policies/mlpilot/promotion.rego +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/terraform/aws/main.tf +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/terraform/aws/outputs.tf +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/terraform/aws/variables.tf +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/terraform/aws-ecs/main.tf +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/terraform/aws-ecs/outputs.tf +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/terraform/aws-ecs/variables.tf +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/terraform/aws-sagemaker/main.tf +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/terraform/aws-sagemaker/outputs.tf +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/deploy/terraform/aws-sagemaker/variables.tf +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/assets/workflow.svg +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/changelog.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/contributing.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/anomaly.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/automl.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/aws.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/causal.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/cleaning.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/environments-cicd.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/evaluation.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/feature-store.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/features.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/forecasting.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/hub.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/ingestion.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/insights.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/integrations.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/lineage.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/model-security.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/monitoring.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/observability.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/quality.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/registry.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/secrets-policy.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/splitting.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/stats.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/guide/versioning.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/policies.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/anomaly.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/automl.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/causal.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/cicd.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/cleaners.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/core.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/deploy.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/environments.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/evaluation.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/feature_store.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/features.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/forecast.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/hub.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/ingestion.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/insights.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/integrations.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/lineage.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/media.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/monitoring.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/poisoning.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/quality.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/registry.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/robustness.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/scale.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/secretstore.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/security.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/splitting.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/stats.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/reference/versioning.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/security.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/docs/workflow.md +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/_compat.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/_detect.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/_rustcore.pyi +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/anomaly.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/automl/__init__.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/automl/_persist.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/automl/cluster.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/automl/frame.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/automl/tasks.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/causal.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/core/__init__.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/core/pipeline.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/core/registry.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/core/step.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/__main__.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/_names.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/asgi.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/aws.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/deploy/scoring.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/environments.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/evaluation/__init__.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/evaluation/_slicing.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/evaluation/adapter.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/evaluation/card.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/evaluation/errors.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/evaluation/explain.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/evaluation/fairness.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/evaluation/metrics.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/feature_store.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/features/__init__.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/features/dates.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/features/defaults.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/features/encoding.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/features/generators.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/features/leakage.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/features/selection.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/forecast.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/hub/__init__.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/hub/client.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/hub/policy.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/hub/security.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/hub/storage.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/hub/ui.html +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/hub/ui.js +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/ingestion/__init__.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/ingestion/_util.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/ingestion/api.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/ingestion/core.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/ingestion/files.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/ingestion/incremental.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/ingestion/schema.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/ingestion/sql.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/ingestion/warehouse.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/insights/__init__.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/insights/finder.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/insights/narrate.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/insights/report.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/__init__.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/airflow.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/aws_glue.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/azure_data_factory.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/dagster.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/dvc.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/evidently.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/feast.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/kafka.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/mlflow.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/prefect.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/integrations/tasks.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/media/audio.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/monitoring/drift.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/monitoring/estimate.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/monitoring/monitor.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/monitoring/ops.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/poisoning.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/py.typed +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/quality/checks.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/quality/contracts.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/quality/pipeline_steps.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/quality/profile.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/quality/schema.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/registry/tracking.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/robustness.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/scale.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/secretstore.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/security.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/splitting.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/stats.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/mlpilot/versioning.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/src/clean.rs +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/src/lib.rs +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/src/stats.rs +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/src/text.rs +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_advanced.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_automl.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_aws_deploy.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_cicd.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_core.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_data_versions_and_feature_store.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_deploy.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_docker_stacks.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_environments.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_evaluation.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_features.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_hub.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_ingestion.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_insights.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_label_free_monitoring.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_lineage_and_catalogs.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_media.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_model_security.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_monitoring.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_ops.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_quality.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_real_integrations.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_registry.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_rename_compat.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_secretstore.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_serving_feedback_reload.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_splitting.py +0 -0
- {dataspoof_mlpilot-2.1.0 → dataspoof_mlpilot-2.3.0}/tests/test_stats.py +0 -0
|
@@ -153,6 +153,49 @@ jobs:
|
|
|
153
153
|
cargo generate-lockfile
|
|
154
154
|
cargo audit
|
|
155
155
|
|
|
156
|
+
certify:
|
|
157
|
+
name: release certification
|
|
158
|
+
# Evidence that the production guarantees hold on the built package: the autonomous engine end
|
|
159
|
+
# to end, model-file scanning, disaster-recovery drill, fault injection, tenancy isolation, SBOM.
|
|
160
|
+
runs-on: ubuntu-latest
|
|
161
|
+
env:
|
|
162
|
+
AC_OFFLINE: "1"
|
|
163
|
+
steps:
|
|
164
|
+
- uses: actions/checkout@v4
|
|
165
|
+
- uses: actions/setup-python@v5
|
|
166
|
+
with:
|
|
167
|
+
python-version: "3.12"
|
|
168
|
+
cache: pip
|
|
169
|
+
- uses: dtolnay/rust-toolchain@stable
|
|
170
|
+
- uses: Swatinem/rust-cache@v2
|
|
171
|
+
- name: Install the package
|
|
172
|
+
run: python -m pip install ".[test,server,signing]"
|
|
173
|
+
- name: Production-guarantee test suites
|
|
174
|
+
run: |
|
|
175
|
+
mkdir _cert && cp -r tests _cert/ && cd _cert
|
|
176
|
+
python -m pytest -q -p no:cacheprovider -o pythonpath= tests/test_orchestration.py \
|
|
177
|
+
tests/test_fault_injection.py tests/test_backup.py tests/test_supply_chain.py \
|
|
178
|
+
tests/test_serving_hardening.py tests/test_hub_tenancy.py tests/test_incidents.py \
|
|
179
|
+
tests/test_model_governance.py tests/test_model_privacy.py tests/test_observability.py
|
|
180
|
+
- name: Live certification run
|
|
181
|
+
run: |
|
|
182
|
+
mkdir -p evidence && cd evidence
|
|
183
|
+
python -c "import numpy as np, polars as pl; r=np.random.default_rng(0); x=r.normal(size=(800,3)); pl.DataFrame({'a':x[:,0],'b':x[:,1],'c':x[:,2],'y':(x[:,0]+r.normal(0,.7,800)>0).astype(int)}).write_csv('data.csv')"
|
|
184
|
+
mlpilot run --text "predict y from data.csv with AUC above 0.7, latency below 500 ms; deploy to staging" \
|
|
185
|
+
--no-llm --name cert --out runs/cert --by ci
|
|
186
|
+
mlpilot scan-model runs/cert/artifacts/model
|
|
187
|
+
mlpilot dr-drill --source runs=runs/cert --source registry=registry --workdir dr --rto-limit 120
|
|
188
|
+
mlpilot sbom --out sbom.mlpilot.cdx.json
|
|
189
|
+
cp runs/cert/report.md certification-report.md
|
|
190
|
+
- uses: actions/upload-artifact@v4
|
|
191
|
+
with:
|
|
192
|
+
name: certification-evidence
|
|
193
|
+
path: |
|
|
194
|
+
evidence/certification-report.md
|
|
195
|
+
evidence/sbom.mlpilot.cdx.json
|
|
196
|
+
evidence/runs/cert/manifest.json
|
|
197
|
+
evidence/runs/cert/events.jsonl
|
|
198
|
+
|
|
156
199
|
linux:
|
|
157
200
|
runs-on: ubuntu-latest
|
|
158
201
|
strategy:
|
|
@@ -236,7 +279,7 @@ jobs:
|
|
|
236
279
|
name: Release to PyPI
|
|
237
280
|
runs-on: ubuntu-latest
|
|
238
281
|
if: startsWith(github.ref, 'refs/tags/v')
|
|
239
|
-
needs: [test, core-only, typecheck, security, linux, windows, macos, sdist]
|
|
282
|
+
needs: [test, core-only, typecheck, security, certify, linux, windows, macos, sdist]
|
|
240
283
|
environment:
|
|
241
284
|
name: pypi
|
|
242
285
|
url: https://pypi.org/project/dataspoof-mlpilot/
|
|
@@ -1,204 +1,326 @@
|
|
|
1
|
-
# Changelog
|
|
2
|
-
|
|
3
|
-
All notable changes are listed here. The project follows [Semantic Versioning](https://semver.org/)
|
|
4
|
-
(see the [versioning policy](https://dataspoof.github.io/dataspoof-
|
|
5
|
-
|
|
6
|
-
## [2.
|
|
7
|
-
|
|
8
|
-
###
|
|
9
|
-
-
|
|
10
|
-
|
|
11
|
-
-
|
|
12
|
-
|
|
13
|
-
-
|
|
14
|
-
|
|
15
|
-
|
|
16
|
-
|
|
17
|
-
-
|
|
18
|
-
-
|
|
19
|
-
|
|
20
|
-
-
|
|
21
|
-
|
|
22
|
-
|
|
23
|
-
|
|
24
|
-
|
|
25
|
-
|
|
26
|
-
|
|
27
|
-
|
|
28
|
-
-
|
|
29
|
-
|
|
30
|
-
-
|
|
31
|
-
-
|
|
32
|
-
|
|
33
|
-
-
|
|
34
|
-
|
|
35
|
-
|
|
36
|
-
|
|
37
|
-
-
|
|
38
|
-
|
|
39
|
-
|
|
40
|
-
-
|
|
41
|
-
|
|
42
|
-
|
|
43
|
-
|
|
44
|
-
|
|
45
|
-
-
|
|
46
|
-
|
|
47
|
-
-
|
|
48
|
-
|
|
49
|
-
|
|
50
|
-
|
|
51
|
-
-
|
|
52
|
-
|
|
53
|
-
|
|
54
|
-
|
|
55
|
-
-
|
|
56
|
-
|
|
57
|
-
|
|
58
|
-
- **
|
|
59
|
-
|
|
60
|
-
|
|
61
|
-
|
|
62
|
-
-
|
|
63
|
-
-
|
|
64
|
-
-
|
|
65
|
-
|
|
66
|
-
|
|
67
|
-
-
|
|
68
|
-
|
|
69
|
-
|
|
70
|
-
|
|
71
|
-
|
|
72
|
-
|
|
73
|
-
- **
|
|
74
|
-
|
|
75
|
-
|
|
76
|
-
|
|
77
|
-
-
|
|
78
|
-
-
|
|
79
|
-
|
|
80
|
-
-
|
|
81
|
-
-
|
|
82
|
-
|
|
83
|
-
-
|
|
84
|
-
-
|
|
85
|
-
|
|
86
|
-
-
|
|
87
|
-
-
|
|
88
|
-
-
|
|
89
|
-
- `
|
|
90
|
-
|
|
91
|
-
-
|
|
92
|
-
|
|
93
|
-
- `
|
|
94
|
-
-
|
|
95
|
-
|
|
96
|
-
|
|
97
|
-
|
|
98
|
-
|
|
99
|
-
|
|
100
|
-
- **
|
|
101
|
-
|
|
102
|
-
|
|
103
|
-
|
|
104
|
-
-
|
|
105
|
-
|
|
106
|
-
-
|
|
107
|
-
|
|
108
|
-
|
|
109
|
-
|
|
110
|
-
|
|
111
|
-
|
|
112
|
-
|
|
113
|
-
-
|
|
114
|
-
|
|
115
|
-
|
|
116
|
-
- **
|
|
117
|
-
`
|
|
118
|
-
|
|
119
|
-
|
|
120
|
-
|
|
121
|
-
|
|
122
|
-
|
|
123
|
-
-
|
|
124
|
-
|
|
125
|
-
-
|
|
126
|
-
|
|
127
|
-
|
|
128
|
-
|
|
129
|
-
|
|
130
|
-
|
|
131
|
-
- **
|
|
132
|
-
|
|
133
|
-
-
|
|
134
|
-
|
|
135
|
-
- **
|
|
136
|
-
|
|
137
|
-
|
|
138
|
-
|
|
139
|
-
-
|
|
140
|
-
- **
|
|
141
|
-
(`
|
|
142
|
-
-
|
|
143
|
-
|
|
144
|
-
-
|
|
145
|
-
-
|
|
146
|
-
|
|
147
|
-
|
|
148
|
-
|
|
149
|
-
|
|
150
|
-
- **
|
|
151
|
-
|
|
152
|
-
-
|
|
153
|
-
-
|
|
154
|
-
-
|
|
155
|
-
|
|
156
|
-
|
|
157
|
-
|
|
158
|
-
|
|
159
|
-
|
|
160
|
-
|
|
161
|
-
###
|
|
162
|
-
-
|
|
163
|
-
|
|
164
|
-
|
|
165
|
-
-
|
|
166
|
-
|
|
167
|
-
|
|
168
|
-
|
|
169
|
-
|
|
170
|
-
-
|
|
171
|
-
-
|
|
172
|
-
|
|
173
|
-
-
|
|
174
|
-
|
|
175
|
-
|
|
176
|
-
|
|
177
|
-
-
|
|
178
|
-
|
|
179
|
-
|
|
180
|
-
|
|
181
|
-
|
|
182
|
-
|
|
183
|
-
-
|
|
184
|
-
-
|
|
185
|
-
-
|
|
186
|
-
-
|
|
187
|
-
-
|
|
188
|
-
|
|
189
|
-
|
|
190
|
-
|
|
191
|
-
-
|
|
192
|
-
|
|
193
|
-
|
|
194
|
-
|
|
195
|
-
-
|
|
196
|
-
|
|
197
|
-
|
|
198
|
-
|
|
199
|
-
-
|
|
200
|
-
-
|
|
201
|
-
-
|
|
202
|
-
-
|
|
203
|
-
-
|
|
204
|
-
-
|
|
1
|
+
# Changelog
|
|
2
|
+
|
|
3
|
+
All notable changes are listed here. The project follows [Semantic Versioning](https://semver.org/)
|
|
4
|
+
(see the [versioning policy](https://dataspoof.github.io/dataspoof-mlpilot/policies/)).
|
|
5
|
+
|
|
6
|
+
## [2.3.0] — 2026-11-22
|
|
7
|
+
|
|
8
|
+
### Added: the autonomous execution engine
|
|
9
|
+
- **`mlpilot.orchestration`:** a goal is a YAML file or a sentence, drafted by rules plus
|
|
10
|
+
optional Claude and always validated.
|
|
11
|
+
- The engine plans it into a typed workflow of registered tools only. Steps: ingest →
|
|
12
|
+
quality / contract → leak-checked split → poisoning check → AutoML → evaluation against the
|
|
13
|
+
goal → latency and size limits → SHAP → fairness → robustness → privacy → model-file scan →
|
|
14
|
+
model card → register → promote → monitor baseline.
|
|
15
|
+
- Runs can be resumed: each step's inputs are hashed, so only changed steps re-run.
|
|
16
|
+
- Promotion waits for four-eyes approvals tied to the exact inputs.
|
|
17
|
+
- Gates stop a run and are never retried away; unexpected errors are retried.
|
|
18
|
+
- Each run records cost per step with a budget stop, an append-only event log, a
|
|
19
|
+
reproducibility manifest and a report.
|
|
20
|
+
- A killed run is taken over and resumed.
|
|
21
|
+
- **Cost-aware AutoML:** with latency or size limits, the engine searches for the best model
|
|
22
|
+
that meets them.
|
|
23
|
+
- **CLI:** `mlpilot plan`, `run`, `approve`, `status`.
|
|
24
|
+
|
|
25
|
+
### Added: enterprise operations
|
|
26
|
+
- **Data observability and reconciliation** (`quality.DataMonitor`, `quality.reconcile`):
|
|
27
|
+
- freshness, volume, schema, nulls and distribution over time, with SLAs;
|
|
28
|
+
- source-to-destination checks of keys, changed rows, sums and schema.
|
|
29
|
+
- **Incident management** (`monitoring.diagnose`, `IncidentLog`):
|
|
30
|
+
- root causes: upstream data, infrastructure, bad release, concept drift, covariate shift;
|
|
31
|
+
- playbooks where low-risk actions run automatically and high-risk actions (rollback) need
|
|
32
|
+
approval;
|
|
33
|
+
- an incident closes only after a healthy verification.
|
|
34
|
+
- **Backups and disaster recovery** (`mlpilot.backup`, `mlpilot backup | restore | dr-drill`):
|
|
35
|
+
- checksummed archives, consistent snapshots of live SQLite databases;
|
|
36
|
+
- restores that are verified first and refuse path traversal;
|
|
37
|
+
- drills that measure RTO and RPO, plus retention.
|
|
38
|
+
- **Supply chain** (`mlpilot.supply_chain`, `mlpilot scan-model | sbom`):
|
|
39
|
+
- malicious pickle and joblib detection without executing anything, and ONNX checks;
|
|
40
|
+
- CycloneDX 1.5 SBOM.
|
|
41
|
+
- **Serving:** per-client rate limits (`rate_limit=`, `AC_RATE_LIMIT`; `429` with
|
|
42
|
+
`Retry-After`), and request validation against the model's input schema (`validate=`,
|
|
43
|
+
`AC_VALIDATE`).
|
|
44
|
+
- **Model governance:**
|
|
45
|
+
- `ModelRegistry.set_risk()`: unacceptable models are blocked; high-risk ones need two
|
|
46
|
+
approvers and a current review;
|
|
47
|
+
- `retire()`: archives every version and points to the replacement.
|
|
48
|
+
- **Active learning** (`mlpilot.labeling`): uncertainty and diversity selection, a labeling
|
|
49
|
+
queue with votes and agreement, and a simulation against random labeling.
|
|
50
|
+
- **Privacy attacks** (`mlpilot.model_privacy`):
|
|
51
|
+
- membership inference and model extraction tests;
|
|
52
|
+
- an engine gate, `max_membership_auc`.
|
|
53
|
+
- **Hub multi-tenancy:**
|
|
54
|
+
- organisations with quotas (workspaces, model versions, storage, runs per day);
|
|
55
|
+
- cross-tenant membership refused;
|
|
56
|
+
- usage metered per tenant and day;
|
|
57
|
+
- existing databases migrated in place.
|
|
58
|
+
- **CI:** a `certify` job that every release requires. It runs the production-guarantee
|
|
59
|
+
suites and a live sequence (autonomous run, model scan, DR drill, SBOM) and uploads the
|
|
60
|
+
evidence.
|
|
61
|
+
- **Docs:**
|
|
62
|
+
- [Autonomous engine & operations guide](docs/guide/enterprise-operations.md);
|
|
63
|
+
- [enterprise gap audit](docs/enterprise-gap-audit.md);
|
|
64
|
+
- the GitHub-readable API reference is now generated by `scripts/gen_api_ref.py`.
|
|
65
|
+
|
|
66
|
+
### Fixed
|
|
67
|
+
- `AutoML` on a pure free-text dataset: the cleaning step dropped the text column as an ID
|
|
68
|
+
(every review is unique). Text columns are now kept and embedded.
|
|
69
|
+
|
|
70
|
+
## [2.2.0] — 2026-11-12
|
|
71
|
+
|
|
72
|
+
### Added: images taken in hard conditions
|
|
73
|
+
- **Weather and lighting** (`media.conditions`):
|
|
74
|
+
- detects fog, rain, snow, low light, glare, motion blur, sensor noise and JPEG damage
|
|
75
|
+
(each scored 0-1 and tagged at 0.5);
|
|
76
|
+
- `condition_report` shows which conditions a dataset covers and flags the rare or missing ones;
|
|
77
|
+
- `apply_condition` simulates all of them, plus shadow and cutout, at severity 1-5;
|
|
78
|
+
- `enhance` restores images: dehaze, low-light brightening, denoise, sharpen.
|
|
79
|
+
- **Faces** (`media.faces`, `[vision]` extra):
|
|
80
|
+
- detection;
|
|
81
|
+
- how much of a face is covered, and by what (hair, mask, sunglasses or object), in a per-image
|
|
82
|
+
`occlusion_report`;
|
|
83
|
+
- simulated coverings for training;
|
|
84
|
+
- blurring or pixelating for privacy (`anonymize`, `blur_regions`).
|
|
85
|
+
- **Licence plates** (`media.plates`, `[ocr]` extra):
|
|
86
|
+
- detection and OCR, with plate patterns (`IN`, `IN_BH`, `UK`, `EU`, `US`) that fix OCR
|
|
87
|
+
look-alikes;
|
|
88
|
+
- optional restoration before reading;
|
|
89
|
+
- `plate_report`, which pairs each read with the conditions in the image.
|
|
90
|
+
- **Robustness benchmark** (`media.benchmark`):
|
|
91
|
+
- `condition_benchmark` re-scores any image model per condition and severity, with face
|
|
92
|
+
coverings included;
|
|
93
|
+
- `condition_augment` builds training sets from the weakest conditions;
|
|
94
|
+
- ready-made predictors for AutoML image classifiers, plate reading and face counting.
|
|
95
|
+
- New extras: `vision` (OpenCV 4: OpenCV 5 no longer ships the detector models) and `ocr`
|
|
96
|
+
(RapidOCR; its models are bundled, so nothing is downloaded at run time).
|
|
97
|
+
- Image functions also accept numpy arrays, alongside PIL images and paths.
|
|
98
|
+
|
|
99
|
+
### Added: quantization
|
|
100
|
+
- **Models** (`mlpilot.deploy.quantize`, `[onnx]` extra):
|
|
101
|
+
- `quantize_model` writes int8 and fp16 ONNX variants of a model and measures each on held-out
|
|
102
|
+
data (file size, speed, metric change and agreement with the original), then recommends
|
|
103
|
+
one;
|
|
104
|
+
- `export_model(..., format="onnx", quantize="int8" | "uint8" | "fp16")`, `quantize_onnx`
|
|
105
|
+
for any ONNX model, and a `mlpilot quantize` CLI command;
|
|
106
|
+
- models with nothing to quantize (tree ensembles, linear models, k-NN) are reported as
|
|
107
|
+
such rather than exported unchanged.
|
|
108
|
+
- **AutoML:** a new opt-in `mlp` model family (scikit-learn neural network), used only when
|
|
109
|
+
named in `include=`. It is the AutoML model family that quantization shrinks.
|
|
110
|
+
- **CI/CD:** a `quantization` gate in `gates.yml` (`mode`, `max_drop`, `min_agreement`,
|
|
111
|
+
`required`).
|
|
112
|
+
- **Embeddings:**
|
|
113
|
+
- `TextEmbeddings(dtype="float32" | "float16" | "int8")`;
|
|
114
|
+
- `quantize_vectors` / `dequantize_vectors` with a per-dimension int8 scale learned on
|
|
115
|
+
training data.
|
|
116
|
+
- **Tables:**
|
|
117
|
+
- `optimize_dtypes` and the `OptimizeDtypes` pipeline step shrink column types without
|
|
118
|
+
changing values (integers by range, floats only when they survive the round trip, text
|
|
119
|
+
to Categorical);
|
|
120
|
+
- the step learns the types on training data and never overflows on new data.
|
|
121
|
+
|
|
122
|
+
### Fixed
|
|
123
|
+
- `save_cleaned_data(quantize=True)` no longer fails on integers above 2.1 billion (it cast
|
|
124
|
+
every `Int64` column to `Int32`).
|
|
125
|
+
- `ClipOutliers` no longer fails on small integer columns (`Int8`, `Int16`, `UInt8`, …)
|
|
126
|
+
whose learned fences lie outside the type's range.
|
|
127
|
+
|
|
128
|
+
## [2.1.0] — 2026-11-09
|
|
129
|
+
|
|
130
|
+
### Renamed: AutomatedCleaning is now MLPilot
|
|
131
|
+
- **Names:**
|
|
132
|
+
- the distribution is `dataspoof-mlpilot` (PyPI refused `mlpilot` as too similar to an existing project);
|
|
133
|
+
- the import is `import mlpilot`, and the CLI is `mlpilot`;
|
|
134
|
+
- `MLPilotWarning` replaces `AutomatedCleaningWarning`, which remains as an alias.
|
|
135
|
+
- **Compatibility:**
|
|
136
|
+
- `import automatedcleaning` and pickles that refer to `automatedcleaning.*` classes resolve to
|
|
137
|
+
the same `mlpilot` modules, so models saved by AutomatedCleaning keep loading with the same
|
|
138
|
+
classes;
|
|
139
|
+
- the `AutomatedCleaning` project on PyPI gets a final 2.0.1 that installs `dataspoof-mlpilot`.
|
|
140
|
+
- **Unchanged:** the `AC_*` environment variables, the `ac_*` Prometheus metrics and the saved-file
|
|
141
|
+
formats (`dataspoof.*`).
|
|
142
|
+
- **Renamed with the library:**
|
|
143
|
+
- the Helm chart (`mlpilot-hub`), the OPA policy package (`data.mlpilot`) and the Prometheus job;
|
|
144
|
+
- the Grafana dashboard file;
|
|
145
|
+
- the MLflow tag prefix and the default OpenLineage namespace.
|
|
146
|
+
|
|
147
|
+
## [2.0.0] — 2026-11-09
|
|
148
|
+
|
|
149
|
+
### Breaking
|
|
150
|
+
- **Lean core install.** `pip install AutomatedCleaning` now installs only the core
|
|
151
|
+
dependencies. The heavy ones moved to extras:
|
|
152
|
+
- `[nlp]`: nltk
|
|
153
|
+
- `[dashboard]`: plotly, matplotlib, missingno, pyfiglet
|
|
154
|
+
- `[pii]`: presidio
|
|
155
|
+
- `[llm]`: anthropic, langchain-anthropic
|
|
156
|
+
|
|
157
|
+
To get what 1.x installed, use `pip install "AutomatedCleaning[all]"`. A feature whose extra
|
|
158
|
+
is missing raises an `ImportError` naming the extra to install.
|
|
159
|
+
- `setup.py` was removed. The build is defined only in `pyproject.toml` (maturin).
|
|
160
|
+
|
|
161
|
+
### Changed
|
|
162
|
+
- **No side effects at import.**
|
|
163
|
+
- Importing the library no longer downloads NLTK data, changes the application's warning
|
|
164
|
+
filters, or reconfigures logging.
|
|
165
|
+
- NLTK corpora are downloaded on first use, and never when `AC_OFFLINE` is set. Text cleaning
|
|
166
|
+
falls back to a built-in stopword list when NLTK isn't installed.
|
|
167
|
+
- **Typed package.** It ships a `py.typed` marker and stubs for the Rust extension, and passes
|
|
168
|
+
`mypy`.
|
|
169
|
+
- **CI** now runs:
|
|
170
|
+
- tests across Linux, macOS and Windows × Python 3.11–3.13 × Polars 1/2;
|
|
171
|
+
- a core-only install test;
|
|
172
|
+
- mypy;
|
|
173
|
+
- pip-audit and cargo-audit.
|
|
174
|
+
|
|
175
|
+
Releases publish with PyPI Trusted Publishing, signed build-provenance attestations and a
|
|
176
|
+
CycloneDX SBOM.
|
|
177
|
+
- Interactive AI corrections use `ANTHROPIC_API_KEY` when it is set, and the current Claude model.
|
|
178
|
+
|
|
179
|
+
### Added
|
|
180
|
+
- **Hub, the multi-user platform server** (`automatedcleaning.hub`, `[server]` extra):
|
|
181
|
+
- workspaces, with SSO through OIDC and identity-provider group mapping;
|
|
182
|
+
- API tokens and service accounts;
|
|
183
|
+
- four roles (viewer, data scientist, approver, admin);
|
|
184
|
+
- runs and a model registry with immutable, checksummed artifacts on disk, S3, GCS or Azure;
|
|
185
|
+
- two-person production approvals with policy gates;
|
|
186
|
+
- monitoring results;
|
|
187
|
+
- a hash-chained, verifiable and exportable audit log;
|
|
188
|
+
- a web UI, the Python client `HubClient`, and `automatedcleaning hub serve | create-admin`.
|
|
189
|
+
- **Hub deployment:** a Docker image with compose (plus PostgreSQL), a Helm chart, and a
|
|
190
|
+
Terraform module for AWS (S3, RDS, Secrets Manager, IAM role, Helm release).
|
|
191
|
+
- **Forecasting** (`Forecaster`):
|
|
192
|
+
- multiple series and multi-step horizons;
|
|
193
|
+
- rolling-origin backtests against seasonal-naive;
|
|
194
|
+
- conformal intervals and hierarchical reconciliation.
|
|
195
|
+
- **Causal inference** (`estimate_effect`): doubly-robust AIPW with overlap and balance
|
|
196
|
+
diagnostics. **Uplift:** `UpliftModel` and `qini`.
|
|
197
|
+
- **Anomaly detection** (`AutoAnomaly`): an ensemble of three detectors with plain-language
|
|
198
|
+
explanations.
|
|
199
|
+
- **AutoML:**
|
|
200
|
+
- conformal `predict_interval` and `predict_set`;
|
|
201
|
+
- `search="optuna"`, `imbalance="auto"`, `n_jobs`;
|
|
202
|
+
- `predict_with_proba` and `set_inference_threads`.
|
|
203
|
+
- **Serving:**
|
|
204
|
+
- Prometheus `/metrics` and OpenTelemetry tracing;
|
|
205
|
+
- canary and shadow rollouts, and `app_from_registry`;
|
|
206
|
+
- SageMaker, Vertex AI and Azure ML routes and Docker targets;
|
|
207
|
+
- an ASGI app and the `container` entry point;
|
|
208
|
+
- single-row latency down from about 500 ms to about 50 ms.
|
|
209
|
+
- **Security:**
|
|
210
|
+
- Ed25519 model signing, verified before unpickling whenever `AC_MODEL_PUBLIC_KEY` is set;
|
|
211
|
+
- `skops` export;
|
|
212
|
+
- signing on registration with `register(..., sign_with=...)`.
|
|
213
|
+
- **Monitoring:**
|
|
214
|
+
- PagerDuty and Opsgenie alert handlers;
|
|
215
|
+
- `join_labels` and `Monitor.check_labels` for labels that arrive later;
|
|
216
|
+
- `sql_sink` and `Monitor.on_check`;
|
|
217
|
+
- `SQLSource.insert` and `execute`.
|
|
218
|
+
- **Scale:**
|
|
219
|
+
- `transform_file` for files of any size;
|
|
220
|
+
- Spark, Ray and Dask scoring;
|
|
221
|
+
- `to_sql` SQL pushdown of fitted cleaning pipelines.
|
|
222
|
+
- **Integrations:** Airflow operators, Prefect tasks and flow, Dagster assets, Kafka stream
|
|
223
|
+
scoring, and plain `integrations.tasks`.
|
|
224
|
+
- **MLflow** (`integrations.mlflow`, `[mlflow]` extra):
|
|
225
|
+
- batched, idempotent export of runs and whole trackers;
|
|
226
|
+
- AutoML models logged as pyfunc models with a signature, so they serve with `mlflow models serve`;
|
|
227
|
+
- `export_registry` (stages become aliases) and `load_automl`.
|
|
228
|
+
- **DVC** (`integrations.dvc`, `[dvc]` extra): `DVCRepo` (init, remotes, track, push and pull,
|
|
229
|
+
read any revision, checkout, `log_to_run`) and `export_repository` for `DataRepository`
|
|
230
|
+
history.
|
|
231
|
+
- **Feast** (`integrations.feast`, `[feast]` extra): `export_to_feast` writes, applies and
|
|
232
|
+
materializes a Feast repository.
|
|
233
|
+
- **Data splitting** (`split_data`, `cv_folds`, `check_split`): stratified, grouped and time-ordered
|
|
234
|
+
splits with gaps; every split is leak-checked (shared groups, time overlap, duplicate rows, unseen classes).
|
|
235
|
+
- **Model security:** `robustness_report` (noise stability, a minimal adversarial attack,
|
|
236
|
+
feature sensitivity, extreme inputs) and `poisoning_report` (confident-learning label errors,
|
|
237
|
+
conflicting duplicates, backdoor triggers, outliers).
|
|
238
|
+
- **Environments** (`Environments`, `environments.yml`): isolated dev / QA / prod registries;
|
|
239
|
+
`release` enforces order, source stage, metric gates, four-eyes approvals and signatures, and
|
|
240
|
+
records provenance on both sides. The registry and Hub gain a `qa` stage.
|
|
241
|
+
- **Model CI/CD:** the CLI commands `train`, `gate`, `register`, `promote`, `release`, `latest`,
|
|
242
|
+
`deploy-sagemaker` and `cicd init` (GitHub Actions / GitLab CI pipelines, checked with actionlint).
|
|
243
|
+
- **Secrets** (`secretstore`): `vault://` (token, AppRole, Kubernetes; KV v1 and v2), `aws-sm://`,
|
|
244
|
+
`env://` and `file://` references, resolved wherever credentials are read.
|
|
245
|
+
- **Open Policy Agent for the Hub** (`hub.policy`): deny-overrides access rules, promotion rules
|
|
246
|
+
and fail-closed behaviour, plus Rego policies with unit tests in `deploy/opa`.
|
|
247
|
+
- **Serving:**
|
|
248
|
+
- `POST /feedback` with prediction IDs, live-accuracy metrics and `join_feedback`;
|
|
249
|
+
- hot reload (`watch=`, `POST /admin/reload`, registry-following);
|
|
250
|
+
- live drift, estimated-performance and alert gauges (`monitor=`).
|
|
251
|
+
- **Monitoring before labels:** `PerformanceEstimator` (CBPE, direct loss estimation),
|
|
252
|
+
`domain_classifier_drift`, and `estimated_performance` alerts in `Monitor.check`.
|
|
253
|
+
- **Grafana + Prometheus** (`deploy/monitoring`): a dashboard, 14 alert rules with `promtool`
|
|
254
|
+
unit tests, and a compose stack.
|
|
255
|
+
- **Lineage** (`lineage`): OpenLineage run events with schema, quality and version facets; HTTP
|
|
256
|
+
and file transports; `emit_tracked_run`, `emit_dataset_version`, `LineageGraph`.
|
|
257
|
+
- **Catalogs and orchestration:** AWS Glue (catalog tables, partitions, reads, job runs) and Azure
|
|
258
|
+
Data Factory pipeline runs.
|
|
259
|
+
- **AWS deployment** (`deploy.aws`): ECR push and blue/green SageMaker endpoints (serverless or
|
|
260
|
+
instances, autoscaling, data capture, rollback alarms); Terraform for SageMaker and ECS Fargate.
|
|
261
|
+
- **Evidently** (`integrations.evidently`): reports plus agreement with the built-in drift report.
|
|
262
|
+
- **Tests against real infrastructure:** Prometheus, Grafana, Vault, OPA and Marquez in Docker
|
|
263
|
+
(`AC_DOCKER_TESTS=1`); moto for AWS; dedicated CI jobs.
|
|
264
|
+
- These three are tested against the real tools (MLflow server and model server, git + DVC
|
|
265
|
+
with a remote, Feast), including in a dedicated CI job.
|
|
266
|
+
- **Image and audio data** (`media`):
|
|
267
|
+
- quality audits (corrupt files, exact and near duplicates, blur, exposure, clipping, silence,
|
|
268
|
+
mixed formats);
|
|
269
|
+
- preprocessors (EXIF-aware image resize and normalise; audio resample, trim and level);
|
|
270
|
+
- augmentation;
|
|
271
|
+
- features for AutoML (colour and edge-orientation histograms; MFCC and spectral features).
|
|
272
|
+
- **Statistical testing** (`stats`):
|
|
273
|
+
- automatic test selection with assumption checks, effect sizes and confidence intervals;
|
|
274
|
+
- chi-square and Fisher's exact test, correlations;
|
|
275
|
+
- A/B tests, power analysis, multiple-testing correction, `test_all`.
|
|
276
|
+
- **Data version control** (`DataRepository`): content-addressed commits, tags, row-level
|
|
277
|
+
diffs, checksum-verified checkout, links to experiment runs, garbage collection.
|
|
278
|
+
- **Feature store** (`FeatureStore`, `Entity`, `FeatureView`): point-in-time-correct
|
|
279
|
+
historical joins with TTL, an SQLite online store, `materialize`, schema guards.
|
|
280
|
+
- **Documentation site** (MkDocs Material with an API reference), plus `SECURITY.md`,
|
|
281
|
+
`CONTRIBUTING.md` and a versioning policy.
|
|
282
|
+
|
|
283
|
+
### Removed
|
|
284
|
+
- `automatedcleaning/cleaning_update.py`: dead code that nothing imported.
|
|
285
|
+
|
|
286
|
+
## [1.14.0] — 2026-11-08
|
|
287
|
+
- **Ingestion & connectors:**
|
|
288
|
+
- files, globs, HTTP, and S3 / GCS / Azure;
|
|
289
|
+
- any SQL database (bound parameters);
|
|
290
|
+
- Snowflake and BigQuery;
|
|
291
|
+
- REST APIs with five pagination styles;
|
|
292
|
+
- reusable schema inference and incremental loads.
|
|
293
|
+
- **Experiment tracking & model registry:** runs, dataset fingerprints, dev → staging →
|
|
294
|
+
production promotion with metric gates, audit history and rollback.
|
|
295
|
+
- **Deployment:**
|
|
296
|
+
- batch scoring and a REST API (API keys, prediction log);
|
|
297
|
+
- bundle, pickle and ONNX export;
|
|
298
|
+
- Docker build contexts and a CLI.
|
|
299
|
+
- **Monitoring:** data and prediction drift (PSI), accuracy decay, alerts and webhooks, and
|
|
300
|
+
champion-vs-challenger retraining.
|
|
301
|
+
- **Insights & reporting:** significance-tested findings, executive reports, and optional
|
|
302
|
+
narratives written by Claude.
|
|
303
|
+
|
|
304
|
+
## [1.13.0] — 2026-11-06
|
|
305
|
+
- **Feature engineering:**
|
|
306
|
+
- dates, out-of-fold target encoding, interactions, aggregations, lags, text embeddings;
|
|
307
|
+
- feature selection and leakage detection.
|
|
308
|
+
- **AutoML:** task detection, time-budgeted model search, an ensemble, and AutoCluster.
|
|
309
|
+
- **Evaluation & explainability:** metrics with confidence intervals, built-in SHAP, error
|
|
310
|
+
analysis, fairness checks and model cards.
|
|
311
|
+
|
|
312
|
+
## [1.12.0] — 2026-11-06
|
|
313
|
+
- Profiling & data quality: automated EDA, the 0–100 quality score, YAML data contracts, schema
|
|
314
|
+
diffs and quality gates.
|
|
315
|
+
|
|
316
|
+
## [1.11.0] — 2026-11-05
|
|
317
|
+
- Shared `Step` / `Pipeline` core and the stateful cleaning pipeline (JSON save/load, audit
|
|
318
|
+
reports, schema checks).
|
|
319
|
+
|
|
320
|
+
## [1.10.0] and earlier — 2026-07
|
|
321
|
+
- 1.10.0: 8 more data-quality checks in the dashboard.
|
|
322
|
+
- 1.9.0: column pickers in the dashboard.
|
|
323
|
+
- 1.8.x: the dashboard data-quality panel.
|
|
324
|
+
- 1.7.0: the enterprise dashboard redesign.
|
|
325
|
+
- 1.6.x: 43 additional cleaning steps and a dependency cleanup.
|
|
326
|
+
- 1.5.x: the Rust compute backend.
|
|
@@ -6,8 +6,8 @@ Thanks for helping! The basics:
|
|
|
6
6
|
|
|
7
7
|
```bash
|
|
8
8
|
# needs the Rust toolchain: https://rustup.rs
|
|
9
|
-
git clone https://github.com/DataSpoof/dataspoof-
|
|
10
|
-
cd dataspoof-
|
|
9
|
+
git clone https://github.com/DataSpoof/dataspoof-mlpilot.git
|
|
10
|
+
cd dataspoof-mlpilot
|
|
11
11
|
python -m venv .venv && source .venv/bin/activate # Windows: .venv\Scripts\activate
|
|
12
12
|
pip install maturin
|
|
13
13
|
maturin develop --release --extras dev # builds the Rust extension in place
|