Skip to content

Latest commit

 

History

History
61 lines (51 loc) · 3.02 KB

readme.md

File metadata and controls

61 lines (51 loc) · 3.02 KB
Raw

tracker

GitHub logo
View on
GitHub




/IDE/readme.md

How do you want to work? What is your development workflow?

IDE The core point where "work" happens in is the box on the screen!

Code THe "work" is typing code, instructions, for softare - Python, Julia, R for instance.

Container But code needs software to interpret its instructions. Even the IDE may be software on a container.

Compute And containers need cpu, memory and maybe even GPU resources to run their software

What options for IDE are there and how do I get the IDE along with Compute + Container + Code

Options reviewed here will include:

  • Google Colaboratory for Notebooks: Colab
    • Like Google Docs for a notebook (.ipynb) file. It comes with free compute (even GPUs).
    • It times out after you quit using it or if you use it for too long (24 hours)
    • It can read and save to GitHub. It can even read and save to Google Drive.
    • Many IDE features on top of a notebook!
    • It comes with a container already loaded with common Python packages
    • Install more Python packages with pip
  • Vertex AI WorkBench: Vertex AI Workbench
    • Select Compute - CPU and GPU (optional)
    • Select a pre-built container, or custom container
    • Starts a Compute Engine VM and setup up JupyterLab for single click access
    • Already authenticated to the current user - access other GCP resources you have permission to
    • Install more Python packages with pip
    • Git clone, commit, push, pull built in
  • Cloud Workstations: Cloud Workstations
    • Enterprise solution for any number of users
    • Setup configuration for common uses: compute size, automatic timeouts (or always warm)
    • Select an IDE from a list or preconfigured hosted IDEs - already configured containers - or custom!
    • Authenticate as a GCP user with permissions granted to that user to other resources
    • Install more Python packages with pip
    • Git integration direction in the IDE you choose!
  • Compute Engine VM: Custom Compute Engine VM
    • Fully custom setup!
    • Even install a windowing environment on Linux to host an IDE directly on it.
    • Connect via SSH, even through Chrome RD
  • Local IDE, like VSCode: VS Code Local Setup
    • Run Python locally - data moves to local machine potentially
    • OR
    • SSH to remote machines: Vertex AI WorkBench, Cloud Worstation, Compute Engine VM