M.I.M.I.R - Multi-agent Intelligent Memory & Insight Repository

# NornicDB ARM64 Metal + Heimdall (Cognitive Guardian) # # Batteries-included deployment with both embedding and SLM models. # No additional model downloads required. # # Build: # docker-compose -f docker/docker-compose.arm64-metal-heimdall.yml build # # Run: # docker-compose -f docker/docker-compose.arm64-metal-heimdall.yml up # # Prerequisites: # Place these models in the ./models directory: # - bge-m3.gguf (embedding model) # - qwen2.5-1.5b-instruct-q4_k_m.gguf (Heimdall SLM) services: nornicdb: build: context: .. dockerfile: docker/Dockerfile.arm64-metal-heimdall image: timothyswt/nornicdb-arm64-metal-bge-heimdall ports: - "7474:7474" # HTTP API + Bifrost UI - "7687:7687" # Bolt protocol volumes: - nornicdb-data:/data environment: # Auth (disable for development) NORNICDB_NO_AUTH: "true" # Heimdall is enabled by default in this image # Override these to customize: # NORNICDB_HEIMDALL_ENABLED: "true" # NORNICDB_HEIMDALL_MAX_TOKENS: "512" # NORNICDB_HEIMDALL_TEMPERATURE: "0.1" restart: unless-stopped healthcheck: test: ["CMD", "wget", "--spider", "-q", "http://localhost:7474/health"] interval: 30s timeout: 10s retries: 3 start_period: 15s volumes: nornicdb-data: