ml-wandb-tracking

What problem does it solve?

Weights & Biases (W&B) helps you stop guessing which training runs work by centralizing metrics, hyperparameters, and artifacts so experiments stay comparable and reproducible.

Core Features & Use Cases

• Experiment tracking: Log metrics, configuration values, and system stats during training to keep results tied to the exact run.
• Hyperparameter sweeps: Automate search over learning rates, batch sizes, and model options using W&B Sweeps for systematic tuning.
• Artifacts & model lifecycle: Version datasets and model checkpoints, then use the W&B Model Registry to promote models across staging and production.

Quick Start

Use the ml-wandb-tracking skill to set up W&B tracking for a PyTorch Lightning Trainer and log your train/validation metrics plus model checkpoints for later comparison.