Ray Train API

The Ray Train API reference documents Ray Train’s public Python API. Each page below covers one group of APIs, such as a framework integration or the configuration classes.

Important

These API references are for the revamped Ray Train V2 implementation, which is the default starting in Ray 2.51. In Ray 2.43 through 2.50, enable it by setting the environment variable RAY_TRAIN_V2_ENABLED=1.

See Ray Train V1 API for the old API references and the Ray Train V2 Migration Guide.

• PyTorch
  • Trainer and configs
  • Training loop utilities
• PyTorch Lightning
• Hugging Face Transformers
• TensorFlow and Keras
• XGBoost
• LightGBM
• JAX
• Configuration
• Training function utilities
  • Classes
  • Functions
  • Collective operations
• Results
• Exceptions
• Ray Tune integration
• Developer APIs
  • Trainer base class
  • Backend base classes
  • Trainer callbacks
  • Exceptions