I (finally) started Machine Learning
Two years ago, in the summer of 2023, I decided to learn to code. I was not completely new to coding, as I had taken a few courses during college. However, apart from reading some code and writing basic logic, I had no experience in assembling fully functional software or building cool stuff.
And I wanted to be able to do just that - build cool stuff.
So I started with a free and open-source course called The Odin Project. I learned JavaScript, HTML, and CSS, and the basics of web development.
It took me 9 months to learn enough to launch my first product, I called toastful - a wedding speech writing app using AI. A couple of months later, I started and launched another product called Pinsearch - a market research tool for Pinterest bloggers.
These products taught me a lot of things, including working with different APIs, data handling, presentation, and designing UI components. I had a lot of fun building these products, and they have made some decent money as well.
However, things got repetitive. I was doing almost the same things - building the same components, calling the same APIs, and solving the same problems, but for different applications.
So I asked myself, what could I learn next that's a bit more challenging than just backend API calls?
I have always been fascinated with Machine Learning. I did some research and Python coding back in 2018 when AI was not the coolest kid in town. Didn't fully understand it without a solid background in programming, so I gave up pretty quickly.
But now that I can code, I thought, why not revisit Machine Learning again and finally understand how it works?
So I reached out to my good friend Faizan from slashml for recommendations on where to get started. He pointed me towards fastai. Fastai is a company that makes machine learning easy by providing APIs to build and train models using PyTorch (a popular framework for building ML models). But they also have a really great course for learning ML called Practical Deep Learning for Coders; From Deep Learning Foundations to Stable Diffusion
I was a coder and wanted to learn 'practical' deep learning. Deep learning is a branch of Machine Learning using something bizarre called Neural Nets (I don't fully understand what they are right now, so I won't pretend otherwise and try to explain them).
I have been through the first few lessons and read the chapters (the course is based on a great book called Deep Learning for Coders with fastai & PyTorch - AI Applications Without a PhD), and I have been blown away by its simplicity and awesomeness. I mean, the things you can do with Machine Learning models can not possibly be done with traditional software. Sometimes, it seems pure magic, but when you look under the hood, all you will see is basic matrix multiplication and calculus.
I am planning to write short blog posts about things I learn during my journey to understand Machine Learning and finally build some cool stuff with it. So, keep an eye out for the next post.