Expand description
SetConfig configuration support (TOML, JSON5, YAML, JSON) TOML Configuration System for GraphRAG Complete configuration management with extensive TOML support
Structsยง
- Algorithmic
Embeddings Config - Algorithmic embeddings configuration (hash-based, TF-IDF)
- Algorithmic
Entity Config - Algorithmic entity extraction configuration (pattern-based)
- Algorithmic
Graph Config - Algorithmic graph construction configuration (token overlap)
- Algorithmic
Pipeline Config - Algorithmic/Classic NLP pipeline configuration Uses pattern matching, TF-IDF, and keyword-based methods
- Algorithmic
Retrieval Config - Algorithmic retrieval configuration (BM25 keyword search)
- Auto
Save SetConfig - Auto-save configuration
- Cleaning
Config - Text cleaning options configuration
- Community
Detection Config - Community detection configuration
- E2GraphRAG
Config - E2GraphRAG configuration Efficient entity extraction using SpaCy-like approach without LLM Achieves 10x faster indexing and 100x faster retrieval
- Entity
Extraction Config - Entity extraction configuration
- Entity
Extraction TopLevel Config - Top-level entity extraction configuration (gleaning settings)
- Entity
Filters Config - Entity filtering configuration
- Experimental
Config - Experimental features configuration
- General
Config - General system configuration settings
- Graph
Building Config - Graph building configuration
- Hybrid
Embeddings Config - Hybrid embeddings configuration
- Hybrid
Entity Config - Hybrid entity extraction configuration
- Hybrid
Graph Config - Hybrid graph construction configuration
- Hybrid
Pipeline Config - Hybrid pipeline configuration Combines semantic and algorithmic approaches with weighted fusion
- Hybrid
Retrieval Config - Hybrid retrieval configuration (RRF fusion)
- Hybrid
Weights Config - Hybrid weight configuration
- LLMParams
Config - LLM generation parameters
- Lazy
GraphRAG Config - LazyGraphRAG configuration Concept-based retrieval without prior summarization (Microsoft Research, 2025) Achieves 0.1% of full GraphRAG indexing cost and 700x cheaper query costs
- Local
Models Config - Local model configuration (Ollama)
- Mode
Config - Pipeline mode/approach configuration Determines which pipeline implementation to use
- Models
Config - Model configuration for LLM and embeddings
- Neo4j
Config - Neo4j graph database configuration
- Ollama
SetConfig - Ollama-specific configuration for local LLM and embeddings
- Performance
Config - Performance tuning configuration
- Pipeline
Config - Pipeline execution configuration
- PostgreSQL
Config - PostgreSQL database configuration
- Semantic
Embeddings Config - Semantic embeddings configuration (neural models)
- Semantic
Entity Config - Semantic entity extraction configuration (LLM-based)
- Semantic
Graph Config - Semantic graph construction configuration (embedding-based)
- Semantic
Pipeline Config - Semantic/Neural pipeline configuration Uses deep learning models for embeddings, entity extraction, and retrieval
- Semantic
Retrieval Config - Semantic retrieval configuration (vector search)
- SetConfig
- Complete GraphRAG configuration loaded from TOML
- Storage
Config - Storage backend configuration
- Text
Extraction Config - Text extraction and chunking configuration