The project of transformer network is a little bit difficult to work with. But the lectures are very well.

the course is very comprehensive however more details about how backpropagation is done would be great !!!

The quiz section needs to improve. It would be helpful that notes be also given as an outline of the ppts.

A bit too much material without going in depth anywhere. Also, code-grader sometimes behaves very strange

Some inconsistencies with the HW level (week 2 basically told you what to write) but very good lectures.

Keras exercises tough to get through based on provided information, but the content was superb as always

Really good course. Exercises are not always connected to the core of the deep learning problem though..

The course is great , in the assignments it is better to stick to single package and TF is preferable.

Highly recommended course to understand the concept of speech recognition models, very real use cases.

Great lectures, but "fill-in-blanks-and-rewrite-math-formula-using-python" assignments were annoying

Very good course. Only, sometimes a bit more cryptic respect to the other ones in the specialization

Great lectures. The quality of exercises are also amazing, although some typos should be fixed asap.

Good course. The lesson on transformer needs to get better, be more logical with more explanations.

Some concepts need to be explained more detailed, otherwise I feel hard to keep up with the lecture

Great class. However, they are some typos/mistakes in the explanations of programming assignments.

learned a lot but I wish their should be some sort of referencing material for further exploration

Lectures were very insightful. Assignments were not as strong. Final week on XFormers was rushed.

Lecture content is pretty good. Exercises are much more finicky (less robust) than prior courses.

more excercise should be there in last course, but overall content was very good

thanks Andrew ng

The Attention model still needs more explanation. I haven't completely grasped attention models

More errors than most in the projects. Likely due to the rush of getting this final course out

Very good course, informative and easy to follow. Only problem I had is with its auto grader.

Really good course but compared to others optional assigments lack instructions in this one.

My only complain is Keras. A crash course/lecture on Keras would benefit this course a lot!

The course is of high quality. However, it is not as good and clear as the previous courses.