How-to guides

How-to guides#

Explore our how-to guides for optimizing, debugging, and managing your MaxText workloads.

⚡ Optimization

Techniques for maximizing performance, including sharding strategies, Pallas kernels, and benchmarking.

Optimization
💾 Data Pipelines

Configure input pipelines using Grain (recommended for determinism), HuggingFace, or TFDS.

Data pipelines
🔄 Checkpointing

Manage GCS checkpoints, handle preemption with emergency checkpointing, and configure multi-tier storage.

Checkpointing
🔍 Monitoring & Debugging

Tools for observability: goodput monitoring, hung job debugging, and Vertex AI TensorBoard integration.

Monitoring and debugging
🐍 Python Notebooks

Interactive development guides for running MaxText on Google Colab or local JupyterLab environments.

Run MaxText Python Notebooks on TPUs
🌱 Model Bringup

A step-by-step guide for the community to help expand MaxText’s model library.

MaxText Model Bringup: Community Contributor Guide
🎓 Distillation

How online distillation works in MaxText: loss anatomy, α / β / temperature schedule tuning, layer indices, monitoring metrics, and troubleshooting.

Distillation