Expand description
trueno-rag has moved to aprender-rag.
This crate re-exports aprender-rag for backward compatibility.
New code should depend on aprender-rag directly.
Modules§
- chunk
- Document chunking strategies for RAG pipelines
- embed
- Embedding generation for RAG pipelines
- error
- Error types for Trueno-RAG
- fusion
- Score fusion strategies for hybrid retrieval
- index
- Indexing for RAG pipelines (BM25 sparse index and vector store)
- loader
- Document loading abstraction for pluggable file format support.
- media
- Media types for subtitle and transcript support.
- metrics
- Retrieval evaluation metrics
- pipeline
- RAG Pipeline implementation with context assembly
- preprocess
- Query preprocessing strategies for improved retrieval.
- rerank
- Reranking module for RAG pipelines
- retrieve
- Retrieval module for RAG pipelines
- sqlite
- SQLite+FTS5 persistent storage backend for RAG indices.
Structs§
- Aggregated
Metrics - Aggregated metrics across multiple queries
- BM25
Index - BM25 index implementation
- Chunk
- A chunk of text from a document
- ChunkId
- Unique chunk identifier
- Chunk
Metadata - Metadata associated with a chunk
- Context
Assembler - Assembles retrieved chunks into a coherent context
- Document
- A document to be indexed
- Document
Id - Document identifier
- Embedding
Config - Configuration for embedding generation
- Fixed
Size Chunker - Fixed-size chunker implementation
- Hybrid
Retriever - Hybrid retriever combining dense and sparse retrieval
- Loader
Registry - Registry that dispatches file loading to the appropriate
DocumentLoader. - Paragraph
Chunker - Paragraph-based chunker
- RagPipeline
- Complete RAG pipeline
- Recursive
Chunker - Recursive chunker implementation
- Retrieval
Metrics - Retrieval metrics for evaluation
- Retrieval
Result - Result of a retrieval operation
- Semantic
Chunker - Semantic chunker that groups sentences by embedding similarity
- Sentence
Chunker - Sentence-based chunker
- Sqlite
Index - SQLite-backed sparse index using FTS5 for BM25 search.
- Sqlite
Store - Combined document store + BM25 index backed by SQLite.
- Structural
Chunker - Structural chunker that respects document structure (headers, sections)
- Subtitle
Cue - A single timed text cue from a subtitle file.
- Subtitle
Loader - Loads subtitle files and produces Documents with timestamp metadata.
- Subtitle
Track - Parsed subtitle file.
- Text
Loader - Loads plain text and Markdown files.
- Timestamp
Chunker - Timestamp-aware chunker for subtitle/transcript content.
- Vector
Store - Vector store for dense retrieval
Enums§
- Chunking
Strategy - Chunking strategy configuration
- Error
- Error types for RAG pipeline operations
- Fusion
Strategy - Strategy for fusing dense and sparse retrieval results
- Pooling
Strategy - Pooling strategy for token embeddings
- Subtitle
Format - Subtitle file format.
Traits§
- Chunker
- Trait for document chunkers
- Document
Loader - Abstraction for loading files of any format into Documents.
- Embedder
- Trait for embedding generation
- Reranker
- Trait for reranking retrieved results
- Sparse
Index - Sparse index trait for lexical retrieval
Functions§
- parse_
subtitles - Parse SRT or VTT from a string, auto-detecting format.
Type Aliases§
- Result
- Result type for Trueno-RAG operations