lstm-workshop-vm

A virtual machine Deep-learning-ready for the LSTM workshop

Setup

1. Install VirtualBox
2. Install Vagrant
3. If you have Git, clone this repository. Otherwise download the content as a ZIP archive and unpack it somewhere.
4. Open a command line interface/shell (PowerShell or Cmd on Windows, Terminal on Mac, your favourite terminal emulator on Linux)
5. Change directory to the location of this repository's files.
6. Run: vagrant up --provision the first time you start up the VM. Use just vagrant up for subsequent calls.
7. Get some coffee, eat a sandwich or surf the web or something, this will take about 10 minutes depending on your internet connection and computer.
8. Run vagrant ssh to log onto the VM from the shell.
9. Inside the box, run ./run_jupyter.sh to start the jupyter notebook server.
10. Open your browser at http://localhost:8888 and log in with password workshop
11. When you want to stop the VM, use Ctrl+c to stop the server, then exit to leave the ssh session, and vagrant halt to stop the VM.

Notes

RAM

The VM has 2 GiB of RAM by default. If you want to increase performance, you may increase the v.memory in [Vagrantfile] to about half of what you have on your system. DO NOT SET THIS HIGHER THAN 75% OF YOUR SYSTEM'S RAM.

CPUs

The VM has 2 cores by default. If you want to increase performance, you may increase the v.cpus in [Vagrantfile] to half of the number of logical cores on your system. DO NOT SET THIS HIGHER THAN HALF THE NUMBER OF LOGICAL CORES ON YOUR SYSTEM.