Expand description
Simple, intuitive pipelines for local LLM inference in Rust.
Powered by Candle, with an API inspired by Python’s Transformers. Includes pipelines for text generation, classification, and masked language modeling.
Modules§
- error
- Error types for this crate.
- fill_
mask - Masked language modeling pipeline.
- sentiment
- Sentiment analysis pipeline.
- text_
generation - Text generation pipeline for LLMs.
- zero_
shot - Zero-shot text classification pipeline.