Add comprehensive ML Model Reproducibility documentation

[Copilot]

This PR adds a comprehensive guide on machine learning model reproducibility to address the need for documented best practices in ensuring consistent and reliable ML experiments.

What's Added

New Documentation: ml-model-reproducibility.md

A complete guide covering all aspects of ML reproducibility including:

• Random Seed Management: Comprehensive examples for setting seeds across all major ML libraries (PyTorch, TensorFlow, NumPy, scikit-learn)
• Environment Management: Best practices for dependency versioning, virtual environments, and Docker containerization
• Data Version Control: Techniques for tracking dataset changes, data hashing, and ensuring consistent data splits
• Model Configuration Management: Structured approaches to storing and versioning hyperparameters and model configurations
• Complete Pipeline Examples: Working code for reproducible ML pipelines with proper logging and experiment tracking

Key Features

Practical Code Examples:

def set_reproducible_seeds(seed=42):
    """Set seeds for all random number generators"""
    random.seed(seed)
    np.random.seed(seed)
    torch.manual_seed(seed)
    torch.cuda.manual_seed_all(seed)
    tf.random.set_seed(seed)
    torch.backends.cudnn.deterministic = True

Tools Integration: Examples with MLflow, Weights & Biases, DVC for experiment tracking and version control

Testing Framework: Unit tests for validating reproducibility across different runs

Deployment Considerations: Docker configurations and environment reproducibility strategies

Repository Updates

• Added the new guide to the Machine Learning & Data Science section in README.md
• Follows the same documentation style and format as existing guides in the repository
• Includes practical examples, best practices, and references for further learning

This documentation provides developers and data scientists with actionable guidance for building reproducible ML systems, addressing common challenges like hardware differences, dependency conflicts, and non-deterministic data loading.

Fixes #68.

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.