version: '3.8'
services:
  mk-llm-api:
    build: .
    environment:
      - MODEL_PATH=${MODEL_PATH:-./models/mistral-finetuned-mk}
      - MODEL_ID=${MODEL_ID:-mk-llm}
      - LOAD_IN_4BIT=${LOAD_IN_4BIT:-false}
      - LOAD_IN_8BIT=${LOAD_IN_8BIT:-false}
      - TORCH_DTYPE=${TORCH_DTYPE:-float16}
      - HOST=0.0.0.0
      - PORT=8000
    ports:
      - "8000:8000"
    deploy:
      resources:
        reservations:
          devices:
            - capabilities: [gpu]