Presentation is amazing... Professor Ng always does fantastic job of communicating the material in a clear and easily understood way.

I took the course on the first run-through, and there were still some kinks in the grading process that were a little frustrating to deal with, but hopefully these will be ironed out for later versions.

Thank you, Professor Ng, and everyone else involved. You never disappoint!!

This course has excellent content. Unfortunately there seems to be a slight drop in quality compared to the other courses in this series, with respect to the programming assignments. I didn't find them to be very clearly explained or illuminating.

I'd recommend the jupyter notebooks be reworked with better explanations and more attention to the notational conventions.

Still an excellent introduction though.

The lectures were outstanding (as usual), but the programming assignments (except for the final Trigger Word assignment) were terrible. I spent almost all my time trying to figure out Keras syntax, without ever having a Keras tutorial or anything. If you are going to rely on Keras, you should probably add a tutorial or some references. A lot of wasted time. But other than that, this course was amazing.

Great course.. But, Lot more complex. Feels rushed though (even though for a 4 week course Andrew tried to cover the topic with lots of examples)... This sequence model course with RNN, Transformer Network etc. needs to be a specialization by itself. All of the RNN components with their dimensions, computations, etc. need to be discussed at length with examples and visual representations to do justice...

Very good course. I liked the speech recognition part more. I found the assignments involving Keras code difficult to do in both RNN and CNN courses. Without the help of discussion forum i wouldn't have completed the Keras assignments. Thank you all the fellow students and mentors for your contributions to the discussion forum. Thank you so much Andrew and team for putting this awesome specialization :)

As the other courses in this series, this is definitely another great course, and explains to details the various sequence models. I gave it 4 stars because I believe it might need some improvements. Compared to the previous courses it felt a little rushed, and had too much new information and long programming exercises for a single week. Maybe it would have been better if it was 4 weeks instead of 3.

Always a great course but I would expect to have more lessons on how to use Keras and Tensor Flow API in a better way for who needs to use them in real NLP applications. I still have some doubts on how to use them correctly (for example the use of time distributed layer in the last exercise 'trigger word detection' that we didn't use in the architecture for the exercise about attention mechanism)

I really enjoyed this course and learned a lot. The descriptions of GRUs and LSTMs were a little scant, however, and I found myself rewatching the videos trying to get my head around them. The course could be improved by going into a little more detail about the different gates and what it means to train them, or what sorts of information or patterns might be relevant for the training of a gate.

Although I loved the course and learnt a lot, I don't feel as confident trying to implement some sort of sequence modelling in practice compared to the other courses. And yet I still got full marks for this section. I think the course could have been spread out to 4 weeks with a few extra examples (maybe some stock market prediction examples). Regardless, thank you so much for the teachings!

I really enjoyed learning this course and gained a lot of knowledge from it. The only challenge that I found was some of the steps of the assignments were not clear. I could resolve them by checking the forum. I would also like to know more about document summary and document similarity, but there was not much content for it. Overall, thanks a lot for putting this valuable content together.

Great material and amazing Andrew Ng (5 stars) but very pure editoral review (videos with a lot of repeats of canceled phrases, pauses, quiz understanding, grader problems, very poorness of mentors support because they can do nothing to help, neither contact deeplearning.ai, in summary it's looks like alpha version of course not release and diserve 3 or even 2 stars), so in total 4 stars.

This course is almost as good as the prior four, but some of the lectures lack detail, there are mistakes in some quizzes, and the programming assignments at times are crammed too full of information. You can end up passing through this class without really understanding what's going on, whereas the CNN class does a much better job of forcing you to understand things before you pass.

Excellent lecture content.

Some of the programming assignments are quite poor. Sometimes there are minor mistakes in function descriptions, and other times the whole assignment architecture/plan is not well thought out. If the staff doesn't have resources to improve this, then allow the community to create branches and submit merge requests :)

Overall, I'm happy with this course.

The lectures were great, making an advanced subject accessible. The course materials were mostly good -- the exception being the optional (non-graded) assignment in Week 1, which was not well-structured, and failed to reinforce the concepts it was intended to. There were challenges with connectivity to the Jupyter notebook server, which caused much frustration and wasted time.

Another great course by Prof. Ng. The reason for 4 stars is that I found the assignments to gloss over a lot of new Keras ideas (for Keras beginners) at the expense of spending more time on how the ideas were being implemented. I think the course should be spread out over more weeks, say 5, and spend the extra time going into more depth around the Keras model architectures.

In the lecture videos there have been quite a few repetitions and in the programming exercises the necessary Keras background has not been delivered. For this I have to subtract one star.

The course's contents are very inspiring, challenging and interesting at the same time. I'm really looking forward to applying the techniques learned so far to problems in my business life.

The course could have covered topics like time-series modeling for prediction (sales, demand, a machine failure in a factory, etc) that is much more applicable than some of the assignments proposed here (half of them seemed to be just for fun). Also, I am a little dissapointed that the course didn't cover chatbots, which is one of the most widely used applications for RNNs.

The course was nice but more mathematics could be taught in the lectures, especially backpropagation in recurrent network. Also I feel there could be one more week in this course where recent models like Transformers and BERT can be taught. Overall a nice course to get familiar with Word Embeddings, LSTM, GRU, and some other topics like Translation and Speech Recognition.

The discussion felt really complicated at points. Also I was disappointed not to be able to complete the optional assignment for LSTM back propagation. Since it is ungraded, it would have been nice to at least see the correct implementation to learn from. Also there were several errors in the expected values or instructions in the assignments, that were really confusing.

The instructor Andrew Sir is excellent in conveying topics, but I just found the last part a bit dry compared to the previous 4.

And the course was a bit too long, even though it said 3 weeks.

But the hands on programming practices in this course, especially is second to none. Top Notch.

One would need to revisit and do it all over again to make it stay inside your head.

Ths course really shows cutting edge technology such as using deep networks consisting of LSTMs, GRUs etc.. I especially liked the audio trigger word recognition.

The translation with attention exercise is really much harder to understand than any other exercise from that specialization. I admit I have managed to implement it more using intuition than real understanding.

It has been quite a good course to explain the tedious concepts of RNN.

The only reason for a 4 star is there is definitely quite some room to improve upon the content and quality to bring it up to the mark of the previous 4 courses. There are quite a few bugs in the assignments which need to be rectified for the benefit of everyone, hope that it shall be done soon!

Concur with other reviewers: this class was good, covering a lot of interesting material and with well-structured quizzes & assignments. But the lectures seemed to skip past the sorts of in-depth explanations I wanted, instead just getting to the end point of "this is what this looks like". So good, but not quite as good as previous courses in the specialization.

This course is great to get intuitive understanding of Word Embeddings, RNNs, LSTMs, GRUs and Attention Models.

You will have great explainer videos and some excellent programming exercises. The course does not make you an expert, but it does make you familiar with the above mentioned architectures, so you can independently code and try them on your own solutions.

Really cool applications to work on, but the videos got a little too much into specific applications that may not be relevant most of the time. It was all interesting, but it made this course a lot longer each week. I could have done without a lot of the specifics of certain applications, just because it will be hard to apply/remember the concepts anyways.