My first hands-on project with Rhyme!
In this lockdown, after getting bored for a long time, I decided to try out different MOOCs. As I am bored most of the time when I’m not coding, I have accounts in many different online learning platforms that I generally don’t use much (e.g., Udemy, Coursera, edX). After browsing the catalog for over a day, I decided I should do a guided project instead of a full-fledged course.
Why a “Guided Project”?
On August 28, 2019, Coursera announced the acquisition of Rhyme Softworks, which provides custom cloud computing frameworks that can be used to complete projects online. It provides the platform with a special kind of short course called “Guided Projects” that are follow-along projects you complete online.
There were many different reasons why I thought a guided project is better:
- It’s different and seemed interesting
- I love hands-on projects and sites like Freecodecamp and Codeacademy
- It is probably the best way to learn to code
- Lastly, it is small and completed in a few hours
The Guided Project
Finally, I decided I should do this course named “Build Multi-Layered Perceptron Models with Keras”. I always wanted to know more about machine learning and Keras, and this seemed like an opportunity for the same.
Rhyme interface
Initially, I had a few hiccups when I tried to log on to the platform and get things going.
The Jupyter notebook would just not open and asked for some sort of token for authentication. I realized that as it is a cloud desktop, I can restart Jupyter from the terminal, and voila! It worked. I just had to change the port of the notebook a bit.
As one can see, the whole webpage is divide into two parts; one contains the cloud computing unit where one can write and execute code, and the other part includes the video lesson for the project. Various other helpful features allow the user to interact with the instructor or teaching assistant directly (if they are online) annotated in the image. The best part of the project is that they mail you the finished product (whatever code you wrote on the machine) along with the datasets and other functions used in the notebook.
The Instructor
The instructor, Snehan Kekre, was friendly and explained the motive behind the guided project. It was easy to follow along and write the lines of code. The instructor explained the code as he wrote (also I was coding along with him), which made it much easier to follow than other traditional lectures.
My Experience
The guided project consisted of three parts/phases:
- An introduction page — This part details the problem addressed in the project and the approach to deal with it.
- The cloud computing framework and the project — This is the main part of the project. Here you follow the instructor and code.
- A short quiz to test your understanding — The questionnaire asks about the different aspects of the project; it may be about some specific code or, in general, about the problem statement in general.
It took me about 1.5 hours to complete the whole project (including the quiz). The quiz was quite easy. Then I had to fill a survey and got my certificate. The project was easy to follow through and quite informative. It was like the hands-on lab sessions we have in our college from time to time (for some courses). I feel like it has made me a bit confident about MLPs in general and also implementing an MLP using Keras. After the project, I installed Keras and tried to code the whole problem from scratch on my PC, and it was awesome!
Pros
- Informative — Although the project was quite short, it did boost my knowledge and confidence.
- Pretty Hands-on — Reading a theoretical model and implementing it to solve a real-life problem are two very different things. The hands-on approach shines a light on the subtle nuances and best practices that should be kept in mind while coding a theoretical project.
- No downloads/software required — As the code you write is implemented on a cloud computing unit, you can run different programs, including high-end libraries like TensorFlow and Keras, that might not work on your native desktop or computer.
- Side by side lectures — As the lecture and coding go in together, it is quite easy to navigate and understand even the most complicated code as opposed to listening to a lecture and then later implementing the same code separately.
- Online — Lastly, as the project is online, it doesn’t require you to go to someplace specifically to learn and the lessons can be accessed from the comfort of your own home.
Cons
- Too Easy — The major problem I have with the project is it was too easy, including the quiz at the end. The quiz consisted of just five questions that were quite easy to answer.
- Pre-requisite Knowledge — The project assumes a specific pre-requisite knowledge about the topic, which may be a setback for people interested in a guided project but who don’t have the background knowledge to take on one.
Takeaway
As mentioned earlier, one needs some background knowledge before jumping into a guided project. Apart from that, it is was a rather enriching experience for me, and I hope it will be the same for others who undertake such projects. I liked the experience so much that I have completed another project, “Logistic Regression with Python and Numpy” and have enrolled in some more. The guided projects provide insight into the practical side of things, which is a necessity in today’s world.
Before you go…
Connect with me on Instagram and Facebook.
If you liked this story, hit the clap button (you can clap up to 50 times with just one button!) and follow me for more such articles!
Share, recommend, and comment away! Engagement helps us communicate and be more human!
