# Directories to index
watch_paths:
  - ./data/scifact
  - ./data/nfcorpus #modify this

# File extensions to include
include_extensions:
  - ".pdf"
  - ".docx"
  - ".txt"
  - ".md"
  - ".pptx"
  - ".xlsx"
  - ".py"
  - ".js"
  - ".ipynb"


# --- Add these ---
top_k: 5                  # final results returned to user
candidate_k: 20           # candidates fetched before reranking

query_expansion: true     # WordNet synonym expansion
max_synonyms: 5           # max synonyms to append

reranking_enabled: true   # cross-encoder reranking
reranker_model: "cross-encoder/ms-marco-MiniLM-L-6-v2"

# Directories to skip
skip_directories:
  - ".git"
  - "node_modules"
  - "__pycache__"
  - ".venv"

# Where to store index data
data_dir: "./data"

embedding_model: "all-MiniLM-L6-v2"
# embedding_model: BAAI/bge-small-en-v1.5

debounce_seconds: 5