services:
  qwenflashnext:
    build: .
    image: qwenflashnext:latest
    container_name: qwenflashnext
    restart: unless-stopped
    gpus: all
    ipc: host
    ports:
      - "8081:8081"
    volumes:
      - ./models:/models:ro
    environment:
      - LLAMA_ARG_HOST=0.0.0.0
      - LLAMA_ARG_PORT=8081
    env_file:
      - .env
    command: >
      --model /models/${QUANT:-UD-Q4_K_XL}/Qwen3.8-Flash-Next-${QUANT:-UD-Q4_K_XL}-00001-of-${SHARDS:-00004}.gguf
      --alias ${MODEL_ALIAS:-qwen3.8-flash-next}
      --ctx-size ${CTX_SIZE:-65536}
      --n-gpu-layers 999
      --override-tensor "per_layer_token_embd=CPU"
      --flash-attn on
      --jinja
      --temp 1.0 --top-p 0.95 --top-k 20 --min-p 0.0
      --parallel ${PARALLEL:-1}
      --threads ${THREADS:-16}
      --metrics
    healthcheck:
      test: ["CMD", "curl", "-fsS", "http://127.0.0.1:8081/health"]
      interval: 30s
      timeout: 10s
      retries: 60
      start_period: 900s