autoforge-engine 0.1.0__tar.gz

autoforge_engine-0.1.0/PKG-INFO

@@ -0,0 +1,105 @@
+Metadata-Version: 2.4
+Name: autoforge-engine
+Version: 0.1.0
+Summary: A transparent, local-first AutoML framework for automated ML pipeline discovery.
+Author: Aditya Kumar Singh
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy>=1.26
+Requires-Dist: pandas>=2.1
+Requires-Dist: scikit-learn>=1.4
+Requires-Dist: rich>=13.7
+Requires-Dist: typer>=0.12
+Requires-Dist: pyyaml>=6.0
+Provides-Extra: boosting
+Requires-Dist: xgboost>=2.0; extra == "boosting"
+Requires-Dist: lightgbm>=4.0; extra == "boosting"
+Requires-Dist: catboost>=1.2; extra == "boosting"
+Provides-Extra: optimization
+Requires-Dist: optuna>=3.6; extra == "optimization"
+Provides-Extra: dev
+Requires-Dist: pytest>=8.0; extra == "dev"
+Requires-Dist: ruff>=0.5; extra == "dev"
+Requires-Dist: black>=24.0; extra == "dev"
+Requires-Dist: mypy>=1.10; extra == "dev"
+Dynamic: license-file
+
+ # ModelForge
+
+ModelForge is a transparent, local-first AutoML framework for discovering,
+evaluating, explaining, and persisting scikit-learn pipelines. It keeps the
+workflow inspectable: data audits, preprocessing, model screening,
+cross-validation, ranking, reproducibility metadata, and artifacts are all
+available as Python objects and local files.
+
+## Quick start
+
+```bash
+python -m pip install -e ".[dev]"
+modelforge train --data examples/house_prices.csv --target price \
+	--task-type regression --output model.joblib
+modelforge predict --model model.joblib --data examples/house_prices_new.csv
+```
+
+The same workflow is available from Python:
+
+```python
+from modelforge import AutoML
+
+automl = AutoML(cv=5, random_state=42)
+result = automl.fit(data, target="price", task_type="regression")
+automl.save("artifacts/house-prices.joblib", overwrite=True)
+print(result["best_model"])
+```
+
+## What is included
+
+ModelForge currently provides ingestion, target detection, profiling, column
+intelligence, data-quality and leakage auditing, preprocessing, feature
+engineering and selection, a model registry, screening, cross-validation,
+ranking, explainability, persistence, prediction validation, optimization,
+experiment tracking, run management, configuration, CLI workflows, failure
+isolation, robustness checks, and reproducibility snapshots.
+
+Optional model and optimization integrations are declared in `pyproject.toml`:
+
+```bash
+python -m pip install -e ".[boosting,optimization]"
+```
+
+## CLI
+
+```bash
+modelforge --help
+modelforge models --task-type classification
+modelforge experiments list --directory .modelforge/experiments
+modelforge experiments get EXPERIMENT_ID
+modelforge experiments compare ID_ONE ID_TWO
+```
+
+The CLI uses Rich tables and keeps generated models, experiment JSON, and
+metadata local by default.
+
+## Examples and benchmarks
+
+- `examples/end_to_end_regression.py` demonstrates a complete Python workflow.
+- `examples/end_to_end_classification.py` demonstrates classification and
+	probability prediction.
+- `benchmarks/benchmark_baselines.py` compares ModelForge with simple baseline
+	estimators and records runtime and quality metrics.
+
+Run an example with `python examples/end_to_end_regression.py` or run the
+benchmark with `python benchmarks/benchmark_baselines.py`.
+
+## Development
+
+```bash
+.venv\\Scripts\\python.exe -m pytest -q
+ruff check .
+```
+
+GitHub Actions runs the test suite on supported Python versions. The roadmap
+for the remaining product work lives in `docs/ROADMAP.md`; it separates
+implemented foundations from integrations that require optional dependencies
+or additional design work.

autoforge_engine-0.1.0/README.md

@@ -0,0 +1,78 @@
+ # ModelForge
+
+ModelForge is a transparent, local-first AutoML framework for discovering,
+evaluating, explaining, and persisting scikit-learn pipelines. It keeps the
+workflow inspectable: data audits, preprocessing, model screening,
+cross-validation, ranking, reproducibility metadata, and artifacts are all
+available as Python objects and local files.
+
+## Quick start
+
+```bash
+python -m pip install -e ".[dev]"
+modelforge train --data examples/house_prices.csv --target price \
+	--task-type regression --output model.joblib
+modelforge predict --model model.joblib --data examples/house_prices_new.csv
+```
+
+The same workflow is available from Python:
+
+```python
+from modelforge import AutoML
+
+automl = AutoML(cv=5, random_state=42)
+result = automl.fit(data, target="price", task_type="regression")
+automl.save("artifacts/house-prices.joblib", overwrite=True)
+print(result["best_model"])
+```
+
+## What is included
+
+ModelForge currently provides ingestion, target detection, profiling, column
+intelligence, data-quality and leakage auditing, preprocessing, feature
+engineering and selection, a model registry, screening, cross-validation,
+ranking, explainability, persistence, prediction validation, optimization,
+experiment tracking, run management, configuration, CLI workflows, failure
+isolation, robustness checks, and reproducibility snapshots.
+
+Optional model and optimization integrations are declared in `pyproject.toml`:
+
+```bash
+python -m pip install -e ".[boosting,optimization]"
+```
+
+## CLI
+
+```bash
+modelforge --help
+modelforge models --task-type classification
+modelforge experiments list --directory .modelforge/experiments
+modelforge experiments get EXPERIMENT_ID
+modelforge experiments compare ID_ONE ID_TWO
+```
+
+The CLI uses Rich tables and keeps generated models, experiment JSON, and
+metadata local by default.
+
+## Examples and benchmarks
+
+- `examples/end_to_end_regression.py` demonstrates a complete Python workflow.
+- `examples/end_to_end_classification.py` demonstrates classification and
+	probability prediction.
+- `benchmarks/benchmark_baselines.py` compares ModelForge with simple baseline
+	estimators and records runtime and quality metrics.
+
+Run an example with `python examples/end_to_end_regression.py` or run the
+benchmark with `python benchmarks/benchmark_baselines.py`.
+
+## Development
+
+```bash
+.venv\\Scripts\\python.exe -m pytest -q
+ruff check .
+```
+
+GitHub Actions runs the test suite on supported Python versions. The roadmap
+for the remaining product work lives in `docs/ROADMAP.md`; it separates
+implemented foundations from integrations that require optional dependencies
+or additional design work.

autoforge_engine-0.1.0/autoforge_engine.egg-info/PKG-INFO

@@ -0,0 +1,105 @@
+Metadata-Version: 2.4
+Name: autoforge-engine
+Version: 0.1.0
+Summary: A transparent, local-first AutoML framework for automated ML pipeline discovery.
+Author: Aditya Kumar Singh
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy>=1.26
+Requires-Dist: pandas>=2.1
+Requires-Dist: scikit-learn>=1.4
+Requires-Dist: rich>=13.7
+Requires-Dist: typer>=0.12
+Requires-Dist: pyyaml>=6.0
+Provides-Extra: boosting
+Requires-Dist: xgboost>=2.0; extra == "boosting"
+Requires-Dist: lightgbm>=4.0; extra == "boosting"
+Requires-Dist: catboost>=1.2; extra == "boosting"
+Provides-Extra: optimization
+Requires-Dist: optuna>=3.6; extra == "optimization"
+Provides-Extra: dev
+Requires-Dist: pytest>=8.0; extra == "dev"
+Requires-Dist: ruff>=0.5; extra == "dev"
+Requires-Dist: black>=24.0; extra == "dev"
+Requires-Dist: mypy>=1.10; extra == "dev"
+Dynamic: license-file
+
+ # ModelForge
+
+ModelForge is a transparent, local-first AutoML framework for discovering,
+evaluating, explaining, and persisting scikit-learn pipelines. It keeps the
+workflow inspectable: data audits, preprocessing, model screening,
+cross-validation, ranking, reproducibility metadata, and artifacts are all
+available as Python objects and local files.
+
+## Quick start
+
+```bash
+python -m pip install -e ".[dev]"
+modelforge train --data examples/house_prices.csv --target price \
+	--task-type regression --output model.joblib
+modelforge predict --model model.joblib --data examples/house_prices_new.csv
+```
+
+The same workflow is available from Python:
+
+```python
+from modelforge import AutoML
+
+automl = AutoML(cv=5, random_state=42)
+result = automl.fit(data, target="price", task_type="regression")
+automl.save("artifacts/house-prices.joblib", overwrite=True)
+print(result["best_model"])
+```
+
+## What is included
+
+ModelForge currently provides ingestion, target detection, profiling, column
+intelligence, data-quality and leakage auditing, preprocessing, feature
+engineering and selection, a model registry, screening, cross-validation,
+ranking, explainability, persistence, prediction validation, optimization,
+experiment tracking, run management, configuration, CLI workflows, failure
+isolation, robustness checks, and reproducibility snapshots.
+
+Optional model and optimization integrations are declared in `pyproject.toml`:
+
+```bash
+python -m pip install -e ".[boosting,optimization]"
+```
+
+## CLI
+
+```bash
+modelforge --help
+modelforge models --task-type classification
+modelforge experiments list --directory .modelforge/experiments
+modelforge experiments get EXPERIMENT_ID
+modelforge experiments compare ID_ONE ID_TWO
+```
+
+The CLI uses Rich tables and keeps generated models, experiment JSON, and
+metadata local by default.
+
+## Examples and benchmarks
+
+- `examples/end_to_end_regression.py` demonstrates a complete Python workflow.
+- `examples/end_to_end_classification.py` demonstrates classification and
+	probability prediction.
+- `benchmarks/benchmark_baselines.py` compares ModelForge with simple baseline
+	estimators and records runtime and quality metrics.
+
+Run an example with `python examples/end_to_end_regression.py` or run the
+benchmark with `python benchmarks/benchmark_baselines.py`.
+
+## Development
+
+```bash
+.venv\\Scripts\\python.exe -m pytest -q
+ruff check .
+```
+
+GitHub Actions runs the test suite on supported Python versions. The roadmap
+for the remaining product work lives in `docs/ROADMAP.md`; it separates
+implemented foundations from integrations that require optional dependencies
+or additional design work.

autoforge_engine-0.1.0/autoforge_engine.egg-info/SOURCES.txt

@@ -0,0 +1,72 @@
+LICENSE
+README.md
+pyproject.toml
+autoforge_engine.egg-info/PKG-INFO
+autoforge_engine.egg-info/SOURCES.txt
+autoforge_engine.egg-info/dependency_links.txt
+autoforge_engine.egg-info/entry_points.txt
+autoforge_engine.egg-info/requires.txt
+autoforge_engine.egg-info/top_level.txt
+modelforge/artifact_manager.py
+modelforge/automl.py
+modelforge/cli.py
+modelforge/column_intelligence.py
+modelforge/config.py
+modelforge/cross_validation.py
+modelforge/data_audit.py
+modelforge/data_loader.py
+modelforge/evaluation.py
+modelforge/experiment_tracker.py
+modelforge/explainability.py
+modelforge/feature_engineering.py
+modelforge/feature_selection.py
+modelforge/hyperparameter_optimization.py
+modelforge/model_registry.py
+modelforge/model_screening.py
+modelforge/persistence.py
+modelforge/pipeline_generator.py
+modelforge/prediction_validator.py
+modelforge/preprocessing.py
+modelforge/profiler.py
+modelforge/ranking.py
+modelforge/reproducibility.py
+modelforge/reproducibility_integration.py
+modelforge/run_manager.py
+modelforge/target_selector.py
+tests/test_artifact_manager.py
+tests/test_automl.py
+tests/test_automl_reproducibility.py
+tests/test_automl_robustness.py
+tests/test_cli.py
+tests/test_cli_workflow.py
+tests/test_column_intelligence.py
+tests/test_config.py
+tests/test_config_automl_integration.py
+tests/test_cross_validation.py
+tests/test_data_audit.py
+tests/test_data_loader.py
+tests/test_end_to_end.py
+tests/test_evaluation.py
+tests/test_experiment_artifacts.py
+tests/test_experiment_tracker.py
+tests/test_experiment_tracker_reproducibility.py
+tests/test_explainability.py
+tests/test_failure_isolation.py
+tests/test_feature_engineering.py
+tests/test_feature_selection.py
+tests/test_full_system.py
+tests/test_hyperparameter_optimization.py
+tests/test_model_registry.py
+tests/test_model_screening.py
+tests/test_persistence.py
+tests/test_pipeline_generator.py
+tests/test_prediction_robustness.py
+tests/test_prediction_validator.py
+tests/test_preprocessing.py
+tests/test_profiler.py
+tests/test_ranking.py
+tests/test_reproducibility.py
+tests/test_reproducibility_integration.py
+tests/test_robustness.py
+tests/test_run_manager.py
+tests/test_target_selector.py

autoforge_engine-0.1.0/autoforge_engine.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

autoforge_engine-0.1.0/autoforge_engine.egg-info/entry_points.txt

@@ -0,0 +1,2 @@
+[console_scripts]
+modelforge = modelforge.cli:app

autoforge_engine-0.1.0/autoforge_engine.egg-info/requires.txt

@@ -0,0 +1,20 @@
+numpy>=1.26
+pandas>=2.1
+scikit-learn>=1.4
+rich>=13.7
+typer>=0.12
+pyyaml>=6.0
+
+[boosting]
+xgboost>=2.0
+lightgbm>=4.0
+catboost>=1.2
+
+[dev]
+pytest>=8.0
+ruff>=0.5
+black>=24.0
+mypy>=1.10
+
+[optimization]
+optuna>=3.6

autoforge_engine-0.1.0/autoforge_engine.egg-info/top_level.txt

@@ -0,0 +1 @@
+modelforge