It feels like more than half the videos are the instructor just stepping through lab code. And 20+ percent of content is just links to papers to read. I mean, I don't mind reading good papers. Just put them all in references page or something and refer to them from course content. Don't try to pad the course content with just links to papers. Another 20+ percent of the content are labs that run on Google Colab. But be careful not to spend too much time reading the code and tinkering around! You might run out of compute time you get with Colab's free plan! I guess you could pay for compute if you don't want to wait for the free compute credit to refresh, but that's extra cost you're likely not aware of when you sign up for the course. But the most infuriating thing about this course is that it is littered with errors. Code with bad variable naming that the instructor just unquestioningly uses in video to give confusing explanation, quiz question with missing code, quiz question with grader marking the wrong answer correct, labs that don't run as intended or not at all, etc. I reported most of the errors through Coursera's report function or through Deeplearning.ai's community forums. But I don't know that I should have to pay for course content that has so many errors. Maybe it would be worth the hassle if it was free until all the kinks are worked out or something.

Coursera & the course provider shows very poor standards in assignments. None of my assignment submitted without complaint. Course provider made all assignment with very poor ethics, all is good, you pass in the colab but on submission grader not taking your model. And no bug/error shown or proved why this this thing happen. Course provider didn't provide valuable information which we need to complete the assignment, he leaves much on the user and on the colab to manage.

The simple fact google colabs resources are not included makes this course one star