ML Workspace

Over the past several years I have been working on my PhD. In the past year, my dissertation has been my focus. As my dissertation defense approaches, there are several tools that were critical to help me approach the finish line. These tools include:

• Scrivener
• Bookends
• ML Workspace

I have played with various python virtual environments including pyenv and virtualenv. They were alright but I did not feel like they were the tools I needed. I switched over to Google Collab and I liked it. I quickly discovered a limitation on a vacation flight to HI. Because I wanted to use my flight time to code for my dissertation I realized that an online tool with no internet sucks.

When I got back from my trip, I started looking for a new tool that would allow me to reproduce my devement environment both off and online and be OS agnostic. I stumbled across a docker image called ML Workspace and it changed the way I work.

The ML in ML Workspace stands for machine learning. According to the ML Workspace GitHub page, it is described as an all-in-one web-based IDE specialized for machine learning and data science. In more simple terms, ML Workspace is a docker container with all the tools nessasary to complete machine learning projects.

There are numerous tools within the container. These tools include VSCode, jupyterlab, git, and others. VSCode and theother tools all run from the browser. It is great python developmeny environment. at one point during my development process i started playing with the idea of moving my code to R. it wss extremely easy to add an R kernel to jupyterlab. Overall, ML Workspace provides me an ideal work environment for my python and jupyter notebook based projects.

If you are looking for a rock solid deveopmemt environment for ML projects, go to thr project´s GitHub page to lear how to get started.