wandb-fortran documentation

Contents:

• Installation
  • Prerequisites
  • Getting the source
  • Configuring the Python environment
  • Building
  • Running tests
  • Using as a dependency
• Usage
  • Quick start
  • API Reference
  • Hyperparameter sweeps
  • Offline mode
• Tutorials
  • athena Example
  • neural-fortran Example
  • fiats Example
  • Sweep Example
• API Reference
  • Module wf

Overview

wandb-fortran (wf) is a standalone Fortran interface to Weights & Biases experiment tracking.

It embeds the Python interpreter at runtime via the Python C API, imports the wandb Python package, and forwards all Fortran API calls to Python — giving you the full wandb feature set (dashboards, sweeps, artifacts) without any Python in your training loop.

Indices and tables

• Index

• Search Page