Thread by @svpino

A summary of the setup I use for work and how I use each one of these:

▫️ Google Colab PRO
▫️ @awscloud's SageMaker
▫️ Mac Pro running @code

Here are the details: 🧵👇

I always start new things with Google Colab.

Opening Colab is fast, and I can go from an idea to a running script in no time.

I can share the notebook and open it anywhere without worrying about version control systems.

2/9

Any new experiments go into Colab. Whenever I need to test something, I do it in Colab.

Paying for the PRO version is a no-brainer for me:

▫️ Faster GPUs
▫️ More RAM
▫️ More Disk
▫️ Longer runtime

3/9

As soon as things get serious, I move to @awscloud's SageMaker.

It is way more expensive, but I have virtually unlimited resources.

For example, running a 48-hour training job can't be done in Colab.

4/9

SageMaker is the place I use to deploy full machine learning systems.

All the work I do that's client-facing is there.

5/9

Finally, I run Visual Studio @code locally.

Anything other than training models happens on my computer.

Pretty much everything is dockerized, including my whole development environment (@code's extensions and everything.)

6/9

Unfortunately, my kick-ass GPU is not good for machine learning (thanks, Apple!)

I do have a lot of RAM, which helps when loading and processing large files. I can also keep 789 Chrome tabs open with no problem.

7/9

I've been using this 3-environment setup for quite some time now, and I'm used to it.

It's fast, flexible, and gives me everything I need.

8/9

I want to tell you that writing these threads takes a lot of work, but I'm not going to lie: I really enjoy it, so this doesn't feel like work to me.

This should give you a good idea of the type of content that I post.

Follow me for more of it.

9/9