Thread by @DynamicWebPaige

✨👩‍💻 Spent the morning diving deep into @GoogleCloud to create demos for my upcoming geoscience ML workshop.

🧠 The learning curve for any new cloud platform is steep; but I found so many nifty new features!

Ex: free deep learning image cloud.google.com/deep-learning-…

#LearnInPublic

✨📒 Tbh, though: have mostly been living in @GoogleColab.

You can run @TensorFlow, @scikit_learn, etc. for free directly in a browser using CPU/GPU/TPU; and there are a wealth of example notebooks available to play with.

Ex: this deep-dive on ethical ML
pair-code.github.io/facets/

✨📒If you have used @ProjectJupyter or @DataBricks notebooks, @GoogleColab feels very similar.

colab.research.google.com/github/google/…

👩‍💻 Other warm-fuzzies: logging in is frictionless; packages are (mostly) pre-downloaded; and notebooks can be pulled from @GitHub + saved to @GoogleDrive.

✨💾 Other nifty thing: in addition to curl-ing in files and !pip installing libraries, you can mount your @GoogleDrive instance from @GoogleColab with:

from google.colab import drive
drive.mount('/content/drive')

📂 The directories are then explorable in the 'Files' tab!