.env-example

# DB2VEC ENVIRONMENT CONFIGURATION
# ===============================
# This file contains environment variables used by db2vec
# Copy this file to ".env" and customize as needed

# INPUT/OUTPUT CONFIGURATION
# --------------------------
# Path to the database dump file to process (.sql/.surql)
DUMP_FILE=./surreal.surql

# Target vector database type
# Options: redis|chroma|milvus|qdrant|surrealdb|pinecone
EXPORT_TYPE=redis

# DEBUG MODE
# ----------
# Print parsed JSON records before embedding
DEBUG=false

# VECTOR DATABASE CONNECTION
# -------------------------
# Vector database URL/host endpoint
VECTOR_HOST=redis://127.0.0.1:6379

# Database authentication (user/password or API key)
USER=root
PASS=
SECRET=
AUTH=false

# Database organization
DATABASE=default_database
TENANT=default_tenant
NAMESPACE=default_namespace

# Pinecone-specific settings
INDEXES=default_indexes
CLOUD=aws
REGION=us-east-1

# VECTOR CONFIGURATION
# -------------------
# Vector dimension size (must match your embedding model)
DIMENSION=768

# Distance metric: l2|ip|cosine|euclidean|dotproduct
METRIC=cosine

# DATA HANDLING
# ------------
# Max payload size (MB) per request
PAYLOAD_SIZE_MB=12

# Batch size for DB inserts
CHUNK_SIZE=10

# Group Redis records by table name (else use FT.CREATE/SEARCH)
GROUP_REDIS=false

# Use exclusion rules from config/exclude.json
USE_EXCLUDE=false

# EMBEDDING CONFIGURATION
# ---------------------
# Which embedding provider to use: ollama, tei, or google
EMBEDDING_PROVIDER=ollama

# Embedding model name/id
# Examples: nomic-embed-text, text-embedding-004, nomic-embed-text-v2-moe
EMBEDDING_MODEL=nomic-embed-text

# API Key for Google Gemini (required if EMBEDDING_PROVIDER=google)
# EMBEDDING_API_KEY=

# URL endpoint for Ollama or Google embeddings (optional)
# EMBEDDING_URL=

# Embedding performance tuning
EMBEDDING_MAX_CONCURRENCY=4
EMBEDDING_BATCH_SIZE=16
EMBEDDING_MAX_TOKENS=8000
OLLAMA_TIMEOUT=60

# Task type for Google Gemini
EMBEDDING_TASK_TYPE=SEMANTIC_SIMILARITY

# TEI (Text Embedding Inference) specific settings
TEI_BINARY_PATH=tei/tei-metal
TEI_LOCAL_PORT=8080

# PERFORMANCE
# ----------
# CPU threads for parallel tasks (0 = auto detect)
NUM_THREADS=0