Week 2 was wonderful to learn and understand. I wish to explore more on LSTM and Attention based models. Video lectures were not enough for me. Still I am happy for what I learnt.

Overall it is a good course. Much need to improve on the feedback of grading of which output is not really helping. Referring to global variables in assignments are really tricky.

Overall a very good course , i wish the model construction ( i.e. the rationale in stacking up the layers) is covered in bit more detail, otherwise an excellent course, no doubt.

The assignments can be made a little difficult. Can Include assignments of Image captioning, that would have been great. Overall a great course. Sir Andrew Ng, superb teaching!!

great course as always by andrew. would give it a 5/5 but had some sloppy mistakes regarding the formulas. overal impression : exceptional and highly recommended. thanks andrew!

Compared to the previous courses, this one was a little hard for me. Because Dr. Andrew tried to squeezed a bunch of information in this course. It's a wonderful course though.

Some of the assignments are very difficult to pass. I think I understand the concepts but the instructions for the assignments could be a lot clearer and provide more guidance.

The course is great. The assignment sessions are a little confusing. It would be good to show the complete codes for the test cases so that test errors can be easier to debug.

I did not feel like I got a super clear understanding on how to use various library functions in keras, I did gain a really good understanding of the underlying models though!

Great course overall with lots of technical details. Week 4 content on transformers is high-level and makes it difficult and time-consuming to complete the hands-on exercise.

Excellent content, although the exercises - at this stage of specialization - should focus more on NN architecture, than on pure computational and matrix slicing tasks, imho.

Even though it has some flaws in audio clips and homeworks, programming assignments are fun to do. Great real world applications on this whole specialization is just perfect!

The content was quite good as with previous courses in this specialization. Just dropping a star because week 1 material wasn't as clear to follow as the rest of the content.

Compared to the first two sections, I don't think this section is better than those. Anyway, I learnt some concepts about sequence models which I need to dive into in future

A great course on NLP including Transformers. One star down for quite a brief introduction to transformers (I understand that time passes but you better update the course)

The videos are of high quality as always, while the programming exercises had some error in it. Compare with the previous few courses this one seems not polished so well.

I learned a lot, but it really felt like there could have been a whole other week's worth of material, especially talking about newer innovations like transformers, etc.

This course and assignments helped me understand concepts in NLP. However, it is a little less comprehensive compared to the previous courses in the deep learning series

Thank you Deep learning team for putting together this course. The course has really helped me understand the various possibilities with the knowledge of deep learning.

The first week was a bit too tough compared to the second and third. So, I felt it was a bit hurried. It could have been distributed into two separate weeks, perhaps.

This course was very useful. I would make the programming assignments for the first session a bit clear. Other than that, everything was easy to understand and clear.

Good, but not as in depth as the other lecture series I found. It is faster paced and skips over much more of the detail at which they go into in the earlier modules.

If the programming exercises weren't plagued by some bugs, I would have given 5 star. It's a very complex subject matter, but Andrew takes you through it by the hand.

I feel week 2 and week 3 materials were covered a bit too quick. Would appreciate more explanation of the implementation details of beam search and activation model.

This last course (and especially the Week 4 material about transformers) feels much more rushed and less detailed than the previous courses in that specialization.