docker-compose.yml

version: '3.8'

services:
  shimmy:
    image: ghcr.io/michael-a-kuykendall/shimmy:latest
    container_name: shimmy-server
    ports:
      - "11434:11434"  # Shimmy server port
    volumes:
      - ./models:/app/models              # Mount your models directory
      - shimmy-cache:/root/.cache         # Persistent cache for downloads
    environment:
      - SHIMMY_BASE_GGUF=/app/models      # Point to mounted models
      - SHIMMY_PORT=11434                 # Server port
      - SHIMMY_HOST=0.0.0.0              # Listen on all interfaces
      # TurboShimmy v2.1 — cut KV cache VRAM ~60% (recommended for 4 GB GPUs)
      # Uncomment to enable INT4 KV quantization:
      # - SHIMMY_KV_QUANT=int4
      # Uncomment to reduce prefill chunk size (prevents Windows TDR crashes on long prompts):
      # - SHIMMY_PREFILL_CHUNK=8
      # Uncomment to override max context length (default: model's native ctx from GGUF):
      # - SHIMMY_MAX_CTX=4096
    restart: unless-stopped
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia              # GPU support (optional)
              count: all
              capabilities: [gpu]

volumes:
  shimmy-cache:
    driver: local