weights-and-biases

Call wandb.init with your project name and config, then call wandb.log with metrics like loss and accuracy inside your training loop. Use wandb.watch to automatically log gradients and model parameters, and wandb.finish when training completes.

Define a sweep config with a search method (grid, random, or bayes), a target metric, and parameter distributions, then create it with wandb.sweep. Launch agents with wandb.agent passing your training function, which reads hyperparameters from wandb.config.

Yes, set report_to="wandb" in TrainingArguments and the HuggingFace Trainer automatically logs metrics, losses, and evaluation results to W&B. You can also add custom WandbCallback subclasses for additional logging.

Yes, set the WANDB_MODE environment variable to offline before initializing your run. Metrics are stored locally and can be uploaded later using the wandb sync command on the run directory.

Artifacts version any file type (datasets, models, predictions) with automatic lineage tracking between runs. The Model Registry is a curated collection where model artifacts are linked and promoted through stages like staging and production using aliases.

Bayesian optimization is recommended for expensive training runs since it learns from previous trials and is most sample-efficient. Use grid search for few discrete parameters and random search for quick exploration across many parameters.