A honest review of the Data and Deployment Specialization
What is the TensorFlow: Data & Deployment Specialization?
The TensorFlow: Data and Deployment specialization is deeplearning.ai' s new group of courses and is intended to teach you how to practically use your DL skills to make practical applications. It is co-taught by Andrew Ng and Laurence Moroney, both seasoned veterans in AI and DL. I will give a detailed review of the first course below. I am trying to keep the content coming while doing the course myself!
1. Browser-Based models with TensorFlow.jS
This course is intended to teach you how to set up a browser-based model using TensorFlow's browser equivalent- TfjS. It shows the necessary JavaScript required for training models. It must be noted that tfjs is not direct TensorFlow! It's a different version, and they have few differences, which are correctly taught in the course. If you are used to Andrew Ng lectures, this isn't like that. The course assumes that you know Deep Learning, so most talks aren't over even 3 minutes.
Prerequisites
- Knowledge (at least basic understanding) of TensorFlow.
- Fluency with programming basics.
Would help
- Knowledge of basic HTML
- Knowledge of jS
Review of Course material
As with any other deeplearning.ai course, this one's also filled with great content. It is high paced, but easy-to-understand teaching language is used. Quizzes aren't too hard, and it's actually an excellent course material-wise.
Review of Programming Assignments
I found the programming assignments relatively fun yet straightforward. The code is explained well in comments, and I really recommend you download the GitHub repository's course code. The community is absolutely fantastic, and people have found solutions to most of the bugs you may encounter in the system.
Last Word
What the course-makers lack in updates, the community makes up for it using discussions. Non-critical bugs like 'accuracy' to be replaced with 'acc' have been pointed out in the forum. I got a good experience in writing tfjs code, and although the course is brief, you can smoothly go forward yourself and do whatever you want to advance yourself in tfjs. I took a certification in this course, as I always feel proud to see an AI/ML certificate in my name, you can try if you want.
Bugs
- BUG: The model got terrifyingly low accuracy (about 60%) when running in Chrome.
HOW I FIXED IT: The easiest fix was changing browsers. I already had both Chrome and Chrome Canary (the dev version of Chrome), so I tried again with Canary and got way better accuracy (about 95%) - BUG: My computer's graphics literally blacked the screen out while trying to train one of the models in the course! But that's probably because my computer is old.
HOW I FIXED IT: I cut a few layers and units off my model, and it worked like a charm with average model accuracy :)
To do or not to do, that is the question
It's a go from DJ!


Comments
Post a Comment