#!/bin/sh

# 使用Dockerfile构建ml_backend完整镜像
docker build -t label_studio_ml_backend .

# 拉取最新的本体镜像
docker pull heartexlabs/label-studio:latest

cd dist/label-studio
# 可修改docker-compose.yml中相应配置，如端口
chmod 777 mydata
chmod 777 localstorage
docker-compose up -d

cd dist/label-studio-ml-backend
# 该目录下的docker-compose.yml，需要配置 LABEL_STUDIO_URL 和 LABEL_STUDIO_API_KEY (登录label-studio前端进行获取)，其他自行参考进行配置
docker-compose up -d

---

# 打包为本地文件
#   - docker save -o label_studio_ml_backend.tar label_studio_ml_backend
#   - docker save -o label_studio.tar heartexlabs/label-studio:latest

# dist
# ├── label-studio
# │   ├── docker-compose.yml
# │   ├── label_studio_v1.13.1.tar
# │   ├── localstorage
# │   └── mydata
# └── label-studio-ml-backend
#     ├── docker-compose.yml
#     ├── label_studio_ml_backend.tar
#     ├── model.py
#     └── models
#         └── best.pt