.env.example

# --- HF Spaces Deployment Settings ---

# --- Application Settings ---
APP_NAME=QuerySphere
APP_VERSION=1.0.0
PORT=7860
HOST=0.0.0.0
DEBUG=false
MAX_FILE_SIZE_MB=100
MAX_BATCH_FILES=5


# LLM Provider Selection
OLLAMA_ENABLED=false
USE_OPENAI=true


# OpenAI API Key (set this in HF Space Secrets tab)
OPENAI_API_KEY=sk-your-actual-key-here
OPENAI_MODEL=gpt-4o-mini


# --- Generation Parameters ---
DEFAULT_TEMPERATURE=0.1
TOP_P=0.9
MAX_TOKENS=1000
CONTEXT_WINDOW=8192


# --- Embedding Settings ---
EMBEDDING_MODEL=sentence-transformers/all-MiniLM-L6-v2
EMBEDDING_DIMENSION=384
EMBEDDING_DEVICE=cpu
EMBEDDING_BATCH_SIZE=16


# --- Chunking Settings ---
FIXED_CHUNK_SIZE=384
FIXED_CHUNK_OVERLAP=20
FIXED_CHUNK_STRATEGY=fixed


# --- Retrieval Settings ---
TOP_K_RETRIEVE=10
TOP_K_FINAL=5
FAISS_NPROBE=16
VECTOR_WEIGHT=0.6
BM25_WEIGHT=0.4
BM25_K1=1.5
BM25_B=0.75
ENABLE_RERANKING=true
RERANKER_MODEL=cross-encoder/ms-marco-MiniLM-L-6-v2


# --- Storage Settings ---
VECTOR_STORE_DIR=./data/vector_store
METADATA_DB_PATH=./data/metadata.db
AUTO_BACKUP=false
BACKUP_DIR=./data/backups


# --- Cache Settings ---
ENABLE_CACHE=true
CACHE_TYPE=memory
CACHE_TTL=3600
CACHE_MAX_SIZE=500


# --- Logging Settings ---
LOG_LEVEL=INFO
LOG_DIR=./logs


# --- Performance Settings ---
MAX_WORKERS=2
ASYNC_BATCH_SIZE=5


# --- RAGAS Evaluation ---
ENABLE_RAGAS=true
RAGAS_ENABLE_GROUND_TRUTH=false