Image Classification & Experiment Tracking

This project demonstrates image classification on the FashionMNIST dataset using a simple CNN model, with integrated experiment tracking via:

• MLflow – tracks metrics, parameters, and artifacts
• TensorBoard – visualizes loss/accuracy logs and images
• DVC – manages dataset versioning, model outputs, and pipeline stages

---

Getting Started

1. Install dependencies

pip install -r requirements.txt

2. Run training (standalone)

python train.py

3. Run with DVC (reproducible pipeline)

dvc repro