Technical Documentation
The server that this Wiki instance is installed on is called 'Beethoven' Here you will find technical documentation about the way 'Beethoven' is organised. Beethoven is intended as a research, development and experimentation environment for data-science applied to the intelligence domain. A second host called 'Mozart' is earmarked as a staging and testing server. Currently 'Mozart's RAM and HDD are mounted in 'Beethoven'. We will be requesting more for both hosts. The fat/heavy laptop has a NVdia GTX graphics card with a compute capability of 7.5 and a 1TB SSD. This host is call liszt and is preferred for GPU critical use such as tensorflow. Host beethoven is good enough for general use and host the nextcloud and wiki instances. It has two non-SSD devices which are suitable for backup and archiving. When installing docker start with portainer and nginxproxymanager.
- Why this setup?
- Linux PopOS installation
- Updating & Upgrading
- Installing & configuring TLJH
- Virtual machines
- Adding and/or removing encrypted drives
- User customisation
- Conda installation
- Tensorflow and other libraries & tools
- Creating a GPU enabled tensorflow docker image
- Jupyter and Tensorflow startup scripts
- Various docker containers
- Configuring the proxymanager
- Dockerizing the UGS Tool
- UGS Tool technical documentation
- Basic git/gitea use (using JupyterLab & TLJH as example)
- Nextcloud access
- Remote access
- One-liners