Hey, everyone I am doing the deeplearning specialization by Andrew ng, I completed the first 3 courses. The next 2 courses are CNNs and then Sequence models. I was wondering if I could do the Sequence models course first and then CNN cause I have an NLP course starting this semester and I wanted to supplement my understanding through this course. Are there any prerequisites in CNNs course that are required for the sequence models course ?
You can do that (take sequence course first then CNN) if you are already familiar with TensorFlow, which is introduced in Course 4 (CNN).
At the end of course 2’s 3rd week there was a link provided to a course called “custom and distributed training with tensorflow“ I completed that course. It covered concepts like gradient tape, Graph mode and custom training, will that be enough to skip CNNs for now ?
Maybe yes. Give it a try. TensorFlow is not that difficult, you can learn while taking RNN course.
From my experience the 3 first courses are very fundamental and necessary to proceed. The rest 2 although are very essential for learning DL, each is very specific on its own field. So, i think you can go for Sequence models that will help you on your semester, and then come back if you have time.
The course custom and distribution training with tensorflow is actually part of tensorflow advanced technique specialisation, because of the same recommendedation when I was doing DLS, I ended up doing the last two courses of DLS and Tensorflow advanced technique specialisation simultaneously.
But I recommend you to complete DLS first, tensorflow might not be difficult once your basics of Deep Learning becomes strong.
Although I recommend to follow course 4 and course 5 sequence only as CNN explains many of layers which might be used in sequence.
Then go with a simpler Tensorflow Developer professional which is pretty easy, then only do Tensorflow Advanced technique specialisation if you are interested in computer vision techniques but assignment aren’t easy for this specialisation.
Good Luck!