Datasets
This chapter covers all built-in dataset types and how to implement your own.
The Dataset trait is introduced in A minimal pipeline — Datasets. Here we go deeper into each concrete implementation.
- Custom Datasets — implementing the
Datasettrait for your own types - Param — read-only parameter values
- Memory Dataset — thread-safe in-memory storage
- Cell Dataset — stack-friendly
no_stdstorage - Partitioned Dataset — directory of files loaded as a map
- Cache Dataset — caching wrapper for any dataset
- List of Datasets — quick reference for all built-in types