A curated set of Weights & Biases (W&B) onboarding examples, organized by difficulty to help you progress from beginner to advanced workflows.
Whether you're brand new to W&B or looking to integrate complex automations, this repository has you covered.
This repository is divided into three main difficulty tiers, plus dedicated API example sections.
Introductory material to get started with W&B.
- Basic experiment tracking
- Logging metrics and artifacts
- Simple dataset registration
Build on the basics and incorporate more advanced features.
- Fine-tuning a Transformer with PyTorch Lightning – Integrate W&B into model fine-tuning workflows.
- Organizing Hyperparameter Sweeps in PyTorch – Efficiently manage and visualize sweeps.
- Log a Confusion Matrix with W&B – Visualize and interpret model performance.
- W&B End-to-End with PyTorch Lightning – A complete training-to-logging workflow example.
For experienced W&B users who want to automate and extend functionality.
- Using the Reports API to generate programmatic reports
- wandb-scim – SCIM (System for Cross-domain Identity Management) API examples
- Automating user and group provisioning in W&B