services:
  #  ollama:
  #    image: ollama/ollama:latest
  #    container_name: ollama
  #    restart: unless-stopped
  #    volumes:
  #      - ollama-data:/root/.ollama
  #    ports:
  #      - 11434:11434
  #    devices:
  #      - /dev/kfd:/dev/kfd
  #      - /dev/dri:/dev/dri
  #    environment:
  #      # AMD GPU architecture override — tune for your card:
  #      # RDNA2 (RX 6800/6900): 10.3.0
  #      # RDNA3 (RX 7900/7800): 11.0.0
  #      # GCN  (RX 580 etc):    9.0.0
  #      - HSA_OVERRIDE_GFX_VERSION=11.0.0
  #      - OLLAMA_HOST=0.0.0.0:11434

  silero-tts:
    image: docker.zereb.ru/silero-tts:latest
    container_name: silero-tts
    restart: unless-stopped
    ports:
      - 5000:5000
    environment:
      - OMP_NUM_THREADS=4
      - MKL_NUM_THREADS=4

#  watchtower:
#    image: containrrr/watchtower:latest
#    container_name: watchtower
#    restart: unless-stopped
#    volumes:
#      - /var/run/docker.sock:/var/run/docker.sock
#    environment:
#      - WATCHTOWER_CLEANUP=true
#      - WATCHTOWER_POLL_INTERVAL=86400
#      - WATCHTOWER_INCLUDE_STOPPED=false

  llama-cpp:
    image: docker.zereb.ru/llama-cpp-vulkan:latest
    restart: unless-stopped
    ports:
      - 5001:8080
    devices:
      - /dev/kfd:/dev/kfd
      - /dev/dri:/dev/dri
    volumes:
      - ./models:/models
    command: >
      --jinja
      -fa 1
      -m /models/Qwen3.5-4B-Q8_0.gguf
      --host 0.0.0.0
      -ngl 999
      --ctx-size 16384
      --batch-size 1024
      --ubatch-size 512

volumes:
  ollama-data: