Expand description
Fast, robust text splitting with bindings for Python, Rust and Javascript. This crate contains the core splitting logic which is shared between Javascript, Python and Rust. Each binding then implements a backend separately.
See tract_backend::NNSplit for information for using NNSplit from Rust.
Re-exports§
pub use tract_backend::NNSplit;
Modules§
- model_
loader - Caching and downloading of models.
- tract_
backend - Backend to run models using tch-rs.
Structs§
- Level
- A Split level, used to describe what this split corresponds to (e. g. a sentence).
- NNSplit
Logic - The logic by which texts are split.
- NNSplit
Options - Options for splitting text.
- Split
Sequence - Instructions for how to convert neural network outputs and a text to
Splitobjects.
Enums§
- Split
- A splitted text.
- Split
Instruction - Instruction to split text.