ApeRAG Manual

상위: ApeRAG

1. Start

git clone https://github.com/apecloud/ApeRAG.git
cd ApeRAG
cp envs/env.template .env
docker-compose up -d --pull always

neo4j 실행 (Optional)

docker compose --profile neo4j up -d neo4j

Docray 실행 (Optional)

# CPU 버전 (기본)
DOCRAY_HOST=http://aperag-docray:8639 docker compose --profile docray up -d

# GPU 버전 (더 빠름, GPU 있으면 권장)
DOCRAY_HOST=http://aperag-docray-gpu:8639 docker compose --profile docray-gpu up -d

# .env 파일에 추가
DOCRAY_HOST=http://aperag-docray:8639

2. Register & Login

http://localhost:3000

• ID
• Emal
• Password

3. Model Settings

http://localhost:3000/workspace/providers

OpenAI API KEY 입력

Models 클릭

Add Model 클릭

Embedding 모델 추가

Embedding 클릭

Collection 부분 활성화

4. Collection Settings

http://localhost:3000/workspace/collections

Add collection 클릭

Collection 이름 및 옵션 설정

Upload file

5. Monitoring

http://localhost:5555

• ID : admin
• Passwd : admin

Celery-worker monitoring

Task monitoring

6. Appendix

6.1. Flower setup

Flower (Celery monitoring) 사용하려면 .env & docker-compose.yaml 수정 필요

.env 환경변수 추가

CELERY_BROKER_URL=redis://default:password@aperag-redis:6379/0

docker-compose.yaml

flower:
    <<: *api
    image: ${REGISTRY:-docker.io}/apecloud/aperag:${VERSION:-v0.0.0-nightly}
    container_name: aperag-flower
    env_file:
      - .env
      - envs/docker.env.overrides
    environment:
      - NODE_IP=aperag-flower
      - CELERY_BROKER_URL=${CELERY_BROKER_URL}   # 추가
    ports:
      - "5555:5555"
    command: ["/app/scripts/start-celery-flower.sh"]
    healthcheck:
      test: ["CMD-SHELL", "curl -fsS http://localhost:5555/ | grep -qi flower || exit 1"]
      interval: 10s
      timeout: 5s
      retries: 12
      start_period: 15s

6.2. Celery-worker

worker 수 늘려서 속도 높일 수 있음

docker compose up -d --scale celeryworker=2