1、安装环境，请务必确认docker和docker-compose已经安装好了
[root@localhost]# cat /etc/redhat-release
CentOS Linux release 7.6.1810 (Core)
[root@localhost]# docker --version
Docker version 19.03.1, build 74b1e89
[root@localhost]# docker-compose -version
docker-compose version 1.24.0, build 0aa59064

2、下载opsmanage的zip包
https://github.com/welliamcao/OpsManage

3、把zip包上传到linux系统任何目录，例如：/data/pkg。

4、编辑aut

目前遇到的问题/使用障碍

• 目前已经支持了6种数据库，是否会支持impala？ 或者hive等其他。感觉实现方式上不是很困难。建议考虑。
• 对于数据库的备份，还原等，有一个开源项目用起来还可以，支持多种数据库的备份还原，建议考虑。phpbu
• 对于mysql或者pg数据库集群的使用，在添加实例的时候需要指定实例的状态是master 还是slave，在实际工作中，主从节点的角色是变化的，人为切换或者故障切换，此时实例的状态就和管理平台上的状态不对应，会造成疑问。想知道此问题有没有进一步的方案或计划？
• 有没有一种办法，能以网状图或者其他图表的形式，把一个集群内的所有节点关联起来，这样对于集群管理来说，应该是非常好的使用体验。
• 有没有考虑支持集群节点自发现？ 比如手

Hi again, Mark.

What is the inputKey field for?

(You can spawn multiple hosts with this key).

Ok, where do I put the input key? What does 'multiple hosts' mean here? Does that mean I can spawn up multiple copies of hostyoself that serve up the same files at the same address and are available via the same sharable URL?

The 2x down-sampling is one of the important operations in reference models. But, a convolution or a pooling with stride=2, padding='SAME' may result in different outputs over different deep learning libraries (e.g., TensorFlow, CNTK, Theano, Caffe, Torch, ...) due to their different padding behaviors.

For example (TensorNets syntax; but can be regarded as pseudo codes for other libraries),

Have you changed the basic TENSORBOX to run on multiclass? If you did, do you have some documentation on how to use your updated version of TENSORBOX with multiclass? If you didn't, on which network do you base your multiclass detection, is it YOLO? And what was the purpose of using TENSORBOX?
Thank you...

I think we should add an image resizer into the store.

This one looks to have shot up to popularity with 8k stars: https://github.com/esimov/caire

It does smart things like scale/resize images while retaining the scale for faces, it can even extend the width of images etc.

Not sure of the best user interface, we could use a string like a human-readable sentence - or JSON as the input. JSO