.rst .pdf

repository open issue suggest edit

Tune API Reference

Tip

We’d love to hear your feedback on using Tune - get in touch!

This section contains a reference for the Tune API. If there is anything missing, please open an issue on Github.

• Execution (tune.run, tune.Experiment)
  • tune.run
  • tune.run_experiments
  • tune.Experiment
  • tune.SyncConfig
• Training (tune.Trainable, tune.report)
  • Function API
  • Trainable Class API
  • Advanced Resource Allocation
  • tune.report / tune.checkpoint (Function API)
  • tune.Trainable (Class API)
  • Utilities
  • Distributed Torch
  • Distributed TensorFlow
  • tune.with_parameters
  • StatusReporter
• Console Output (Reporters)
  • CLIReporter
  • JupyterNotebookReporter
  • ProgressReporter
• Analysis (tune.analysis)
  • ExperimentAnalysis (tune.ExperimentAnalysis)
  • TrialCheckpoint (tune.cloud.TrialCheckpoint)
• Search Space API
  • Overview
  • Custom/Conditional Search Spaces
  • Random Distributions API
  • Grid Search API
  • References
• Search Algorithms (tune.suggest)
  • Summary
  • Saving and Restoring
  • Random search and grid search (tune.suggest.basic_variant.BasicVariantGenerator)
  • Ax (tune.suggest.ax.AxSearch)
  • Bayesian Optimization (tune.suggest.bayesopt.BayesOptSearch)
  • BOHB (tune.suggest.bohb.TuneBOHB)
  • BlendSearch (tune.suggest.flaml.BlendSearch)
  • CFO (tune.suggest.flaml.CFO)
  • Dragonfly (tune.suggest.dragonfly.DragonflySearch)
  • HEBO (tune.suggest.hebo.HEBOSearch)
  • HyperOpt (tune.suggest.hyperopt.HyperOptSearch)
  • Nevergrad (tune.suggest.nevergrad.NevergradSearch)
  • Optuna (tune.suggest.optuna.OptunaSearch)
  • SigOpt (tune.suggest.sigopt.SigOptSearch)
  • Scikit-Optimize (tune.suggest.skopt.SkOptSearch)
  • ZOOpt (tune.suggest.zoopt.ZOOptSearch)
  • Repeated Evaluations (tune.suggest.Repeater)
  • ConcurrencyLimiter (tune.suggest.ConcurrencyLimiter)
  • Custom Search Algorithms (tune.suggest.Searcher)
  • Shim Instantiation (tune.create_searcher)
• Trial Schedulers (tune.schedulers)
  • Summary
  • ASHA (tune.schedulers.ASHAScheduler)
  • HyperBand (tune.schedulers.HyperBandScheduler)
  • Median Stopping Rule (tune.schedulers.MedianStoppingRule)
  • Population Based Training (tune.schedulers.PopulationBasedTraining)
  • Population Based Training Replay (tune.schedulers.PopulationBasedTrainingReplay)
  • Population Based Bandits (PB2) (tune.schedulers.pb2.PB2)
  • BOHB (tune.schedulers.HyperBandForBOHB)
  • ResourceChangingScheduler
  • FIFOScheduler
  • TrialScheduler
  • Shim Instantiation (tune.create_scheduler)
• Scikit-Learn API (tune.sklearn)
  • TuneGridSearchCV
  • TuneSearchCV
• Stopping mechanisms (tune.stopper)
  • Stopper (tune.Stopper)
  • MaximumIterationStopper (tune.stopper.MaximumIterationStopper)
  • ExperimentPlateauStopper (tune.stopper.ExperimentPlateauStopper)
  • TrialPlateauStopper (tune.stopper.TrialPlateauStopper)
  • TimeoutStopper (tune.stopper.TimeoutStopper)
• Loggers (tune.logger)
  • Custom Loggers
  • Trainable Logging
  • Viskit
  • TBXLogger
  • JsonLogger
  • CSVLogger
  • MLFlowLogger
  • WandbLogger
  • LoggerCallback
• External library integrations (tune.integration)
  • Docker (tune.integration.docker)
  • Keras (tune.integration.keras)
  • Kubernetes (tune.integration.kubernetes)
  • MLflow (tune.integration.mlflow)
  • MXNet (tune.integration.mxnet)
  • PyTorch Lightning (tune.integration.pytorch_lightning)
  • Torch (tune.integration.torch)
  • Horovod (tune.integration.horovod)
  • Weights and Biases (tune.integration.wandb)
  • XGBoost (tune.integration.xgboost)
  • LightGBM (tune.integration.lightgbm)
• Tune Internals
  • Main Components
  • RayTrialExecutor
  • TrialExecutor
  • TrialRunner
  • Trial
  • Callbacks
  • PlacementGroupFactory
  • Registry
• Tune Client API
• Tune CLI (Experimental)
• Scalability and overhead benchmarks
  • Result throughput
  • Network overhead

Examples Execution (tune.run, tune.Experiment)