Get started

This section is intentionally small. Its job is to get NexuML installed, show the scenario → config → pipeline flow, and then point you to the right learning path.

1. Install NexuML

Follow Installation. For the built-in example scenarios, install nexuml[library].

2. Inspect your first scenario

Follow Your first scenario to discover cifar-resnet, resolve it to YAML, and build the pipeline before spending any GPU time.

3. Learn by building your own library

After the first run, move to the NexuML Tutorials. The tutorials are the primary progressive learning material: they start with a small external MNIST library and then add native DALI audio loading and model composition.

Where to go after that

• I want to create my own scenario → Define a scenario
• I want to understand the architecture → Mental model
• I want to add my own dataset/model/evaluation code → Build a custom library
• I want to train, evaluate, tune, or export → Guides
• I need the exact API or CLI syntax → Reference

You do not need to read every concept page before using NexuML. Learn the framework through the tutorials and use Concepts/Reference when a question comes up.