Configuration

settings.yaml

GraphRAG is configured via settings.yaml. Key areas:

models: Chat + embedding model configuration (OpenAI)
input/output/cache/reporting: storage locations (file-based by default)
extract_graph, community_reports: core workflows
query: prompts and behavior for drift/local/global/basic

Example (excerpt):


models:
  default_chat_model:
    type: openai_chat
    api_key: ${OPENAI_API_KEY}
    model: gpt-4o-mini-2024-07-18
  default_embedding_model:
    type: openai_embedding
    api_key: ${OPENAI_API_KEY}
    model: text-embedding-3-small
 
input:
  storage: { type: file, base_dir: "input" }
output:
  type: file
  base_dir: "output"
cache:
  type: file
  base_dir: "cache"
reporting:
  type: file
  base_dir: "logs"
 
extract_graph:
  prompt: "prompts/extract_graph.txt"
  entity_types: [organization, person, geo, event]
 
community_reports:
  graph_prompt: "prompts/community_report_graph.txt"
  text_prompt: "prompts/community_report_text.txt"
  max_length: 2000
  max_input_length: 8000

Prompts

Prompt templates are in prompts/ and are referenced from settings.yaml for the extract, report, and query workflows.

Environment Variables

Set OPENAI_API_KEY in .env. See ./env for guidance.