Statistics and experimental design · MLD-1024
Reproducibility
The ability to obtain consistent results using documented data, code, settings, and conditions.
Original reference page · detailed expansion pending. Original definitions, examples, notes and citations are preserved below. New examples are illustrative, not benchmark results.
Original example
Rerun a training experiment from a saved configuration.
Original caveat
Definitions of reproducibility and replicability vary by field.
Context and aliases
Level: Intermediate. Stages: Theory · Evaluation · Research.
Tags: ML · DL · SLM · LLM.
Aliases: None recorded.
Related concepts
Concepts linking here
Technology connections
- Apptainer — Container system designed for scientific/HPC workloads.
- BoTorch — Bayesian optimization toolkit built on PyTorch.
- ClearML — Experiment tracking and ML workflow platform.
- Conda — Cross-language package and environment manager.
- Cython — Compiler for Python-like code with optional static types.
- DVC — Data and pipeline versioning linked to Git workflows.
- Git — Distributed source-code version control.
- Git LFS — Git extension for externally stored large-file content.
- Hydra — Configuration composition and experiment-launch framework.
- Hypothesis — Property-based testing library for Python.
- lakeFS — Version-control-like operations for object storage.
- MLflow — Experiment, artifact, model and AI lifecycle tooling.
- Optuna — Hyperparameter optimization library.
- Ray Tune — Distributed experiment and hyperparameter tuning library.
- Ruff — Python linter and formatter.
- TensorBoard — Visualization tools for model training logs and artifacts.
- uv — Python package and project-management tool.
- Weights & Biases — Experiment tracking and model-development platform.
Original references
- scikit-learn: cross-validation ↗ · Reference link
- Mathematics for Machine Learning ↗ · Reference link
- scikit-learn: evaluation metrics ↗ · Reference link
- Original reference URL ↗