How naveen2002-ncg/ml-automation-pipeline works
This repository is an end-to-end classification pipeline. Its CLI entrypoint loads and splits Iris data, trains a Random Forest, evaluates metrics against an accuracy threshold, and exports a report and run log. Runtime configuration supplies environment-based parameters and output paths. The README documents the model, report, and log artifacts. Relationships are based on the sampled entrypoint and source modules; CI and tests are excluded as delivery infrastructure.
Main entry points: run_pipeline.py Pipeline entrypoint · config.py Runtime configuration · model_trainer.py Random Forest training · reporter.py Report generation
Components 9
Pipeline orchestration
- Pipeline entrypoint
run_pipeline.py
Machine learning stages
- Data loading and split
src/data_loader.py - Random Forest training
src/model_trainer.py - Metrics and threshold
src/evaluator.py
Configuration and outputs
- Report generation
src/reporter.py - Runtime configuration
src/config.py - Serialized model
- Report and run log
Other components
- Iris dataset
Connections 11
- Pipeline entrypoint → Runtime configuration: reads settings
run_pipeline.py - Pipeline entrypoint → Data loading and split: loads and splits
run_pipeline.py - Data loading and split → Iris dataset: loads dataset
src/data_loader.py - Pipeline entrypoint → Random Forest training: trains and saves
run_pipeline.py - Pipeline entrypoint → Metrics and threshold: scores model
run_pipeline.py - Pipeline entrypoint → Report generation: records and exports
run_pipeline.py - Random Forest training → Serialized model: serializes model
src/model_trainer.py - Report generation → Report and run log: writes report and log
src/reporter.py - Random Forest training → Runtime configuration: gets model path
src/model_trainer.py - Metrics and threshold → Runtime configuration: gets default threshold
src/evaluator.py - Report generation → Runtime configuration: gets output paths
src/reporter.py