Setting Up the Environment

Setup everything in a few minutes

Thanks to the magic of Docker and VSCode we can launch a running environment really quick! 🎉

You can choose how to do it:

👉 Setting the environment in my own computer
👉 Setting the environment using Github Codespaces

Local Environment

Thanks to the magic of the DevContainers we can simply, clone the repo, open VSCode and launch a full ready environment clicking a single button. 😄

You need Docker Installed

In order to use DevContainers you need to have Docker installed in your computer. If you don't have it, you can download it from here.

Install the VSCode pluggin "Dev-Containers"

In order to use DevContainers you need to have the VSCode pluggin Dev Containers.

You can do it manually:

Or by pressing:

Press Ctrl+P
Paste ext install ms-vscode-remote.remote-containers
Press Enter

1. Clone the repository

Run this command

git clone `https://github.com/<your_user>/mlops-course.git`

2. Open VSCode

Launch VSCode and open the folder of the repo.

On linux or Mac: Run this command

cd mlops-course/
code .

3. Launch the Devcontainer

Click on the bottom-right blue button Reopen in Container (wait, first time takes some time to run).

4. Open MLFlow

A MLFlow server should start automatically (thanks to the magic of Docker Compose) at http://localhost:5000/. If it did not started automatically (or if you launched VSCode on the web), then follow this instructions:

👉 Click on: Ports
👉 Check there is a port 5000 open (if not add a port 5000)
👉 Click on: 🌐

Congrats!

🎉 You are ready to go! 🎉

Github Codespaces

GitHub Codespaces is like having a ready-to-use computer for writing and testing code, but it's all online. Imagine if your coding tools, like your text editor and programming languages, were available on the internet. With GitHub Codespaces, you can write and run your code in a web browser without needing to install everything on your personal computer. It's handy for collaborating with others and quickly starting coding projects, kind of like having your coding workspace in the cloud.

You need a Github Account

In order to access to GitHub Codespaces you need to log in into your account or register if you have no account. Github provides 60 hours of computing at CodeSpaces for free! 🤑

1. Create a copy of the repository (fork)

👉 Go to the oficial repo at https://github.com/Matesanz/mlops-course
👉 Click on Fork > Create a new Fork

2. Launch a CodeSpace in VSCode

👉 Go to your fork at https://github.com/<your_user>/mlops-course
👉 Click on: <> Code > Open in > Open in Visual Studio Code

Accept all the pop-ups that appear

Alternative: you can also launch VSCode in the web

💡 If you have not installed VSCode at your computer this is a (even more) straightforward way to go.

👉 Go to your fork at https://github.com/<your_user>/mlops-course
👉 Click on Code > Create codespace on main

👉 Wait for the CodeSpace to launch

Don't forget to stop your CodeSpace once you've finished!

Remember that you "only" have 60 hours / month of free CodeSpaces. But no worries if you forget to do it, after a time of inactivity (if you closed the tab), the CodeSpace will automatically turn off. 👍

3. Open MLFlow

A MLFlow server should start automatically (thanks to the magic of Docker Compose) at http://localhost:5000/. If it did not started automatically (or if you launched VSCode on the web), then follow this instructions:

👉 Click on: Ports
👉 Check there is a port 5000 open (if not add a port 5000)
👉 Click on: 🌐

Congrats!

🎉 You are ready to go! 🎉