Getting Started Overview

This section provides lightweight companions to the unified workflow. Follow the workflow for the full narrative; dip into these pages for quick reminders and platform-specific tips.

---

Quick Navigation

Guide	Purpose	When to read
Installation	Hardware/software requirements, concise install recipe, platform notes, troubleshooting highlights	Setting up a new machine or debugging installs
Data Preparation	Required raw/ + splits/ layout, ml-split customization, validation utilities	Organising any dataset before training
Quick Start	5-minute sample run, sanity checks, pointers back to the workflow	First-look experience or smoke test

For the sequential end-to-end workflow (install → prep → tune → train → infer → export), start with Workflow Guide.

---

Suggested First Steps

1. Install the package (Workflow Step 1) and confirm ml-train --help works.
2. Organise your dataset using the layout from the Data Preparation guide (Workflow Step 2).
3. Run the Quick Start commands to train the bundled ants vs bees example.

After that, continue through the workflow for cross-validation, tuning, inference, and export.