User Guides

If you’re new to Ray Data, start with the Ray Data Quickstart. This user guide helps you navigate the Ray Data project and show you how achieve several tasks.

• Loading Data
  • Reading files
  • Loading data from other libraries
  • Reading databases
  • Creating synthetic data
  • Loading other datasources
  • Performance considerations
• Inspecting Data
  • Describing datasets
  • Inspecting rows
  • Inspecting batches
  • Inspecting execution statistics
• Transforming Data
  • Transforming rows
  • Transforming batches
  • Ordering of rows
  • Stateful Transforms
  • Avoiding out-of-memory errors
  • Aggregations
  • Group-by and transforming groups
• Iterating over Data
  • Iterating over rows
  • Iterating over batches
  • Iterating over batches with shuffling
  • Splitting datasets for distributed parallel training
• Joining datasets
  • Configuring Joins
  • Configuring number of partitions
  • Configuring number of Aggregators
• Shuffling Data
  • Types of shuffling
  • Advanced: Optimizing shuffles
• Saving Data
  • Writing data to files
  • Converting Datasets to other Python libraries
• Working with Images
  • Reading images
  • Transforming images
  • Performing inference on images
  • Saving images
• Working with Text
  • Reading text files
  • Transforming text
  • Performing inference on text
  • Saving text
• Working with Tensors / NumPy
  • Tensor data representation
  • Transforming tensor data
  • Saving tensor data
• Working with PyTorch
  • Iterating over Torch tensors for training
  • Transformations with Torch tensors
  • Batch inference with PyTorch
  • Saving Datasets containing Torch tensors
  • Migrating from PyTorch Datasets and DataLoaders
• Working with LLMs
  • Perform batch inference with LLMs
  • Configure vLLM for LLM inference
  • Batch inference with vision-language-model (VLM)
  • Batch inference with an OpenAI-compatible endpoint
  • Usage Data Collection
  • Production guide
• Monitoring Your Workload
  • Ray Data progress bars
  • Ray Data dashboard
  • Ray Data logs
  • Ray Data stats
• Execution Configurations
  • Configuring ExecutionOptions
  • Configuring DataContext
• End-to-end: Offline Batch Inference
  • Quickstart
  • Configuration and troubleshooting
• Advanced: Performance Tips and Tuning
  • Optimizing transforms
  • Optimizing reads
  • Reducing memory usage
  • Configuring execution
  • Reproducibility
• Advanced: Read and Write Custom File Types
  • Read data from files
  • Write data to files