Course content is excellent. Organization is not good. It seems like there were some errors in the order of a few quizzes as well as links to the wrong sections. If you take this course and find yourself lost on a quiz or a lab; skip it watch more content, then go back. It was probably out of order.

good intro

Willing to rate first couple of courses well for intro. By now expected much more meat - details, coding exercises - not the high level skimming of the tools. Almost an advert for GCP. Anyone with minimum java/python reading level can take this class - no intermediate programming involved.

Good course. Learning experience is mediocre due to many issues in the Lab's .

As a person who already knew ML and tensorflow, the way the instructor referred to things, mixed concepts, and simplified some of them, made everything a lot more confusing that it really is. I know its easy to dismiss it in this kind of lessons as "that's a complex subject, so it's normal" but it really comes down to the instructor not being effective as he should be.

Although the course is positioned as intro to machine learning with TensorFlow, it includes too many low-level details that you are unlikely be able to conceive without prior knowledge in machine learning and TensorFlow. The parts devoted to feature engineering and hyperparameter tuning are fully chaotic.

Muito complexo para quem não tem bagagem algum com python e machine learning.

The course seemed little superficial to me

Decent content is made very difficult to take in due to poorest editing I have ever seen on Coursera. There may not be many alternatives to this course but in terms of quality this course is not a fair value for money. There are multiple duplications, haphazard cuts, 3-5 second videos that are tied together out of context. It seems the videos are taken from a classroom presentation and shoehorned into an online course format. Very disappointing.

Organization of the videos is poor. The videos are too many in number and a large number of them are extremely short in duration. But it really good to learn tensorflow and usage in distributed fashion in GCP.

Lak is a great teacher, I love his other courses, but this one is not up to the mark. This course has been basically recorded in a live training session and just uploaded the slides and audio as a video.

There are places where he is clearly pointing at stuff on the slide in his live session, but in the video there is no way of knowing what he is exactly talking about.

Please recreate this course from scratch specifically for Coursera format. This is a very important course and students should get all the concepts.

Course Content was good, but poorly organised and span of the videos were very short. Difficult to focus as videos were very short. lot of videos are less than two minutes. Kindly make the videos for atleast 5-10 mins long. If possible, combine the videos, it would help the course takers in future.

not practical at all

I'd like the updated version of the course and lab with the new Tensorflow.org tutorials with Keras etc. Sorry about the 2 stars, apart of that I learned important things in the lab and course. I'll give a 5 stars with the updated version. Thanks.

Data too huge to process within timeline of lab

super exciting topic but there are some issues with the labs. That's super frustrating

Not updated, python 2.7

Too many small clips, should build a 4 modules * 10 videos *10 minutes each

GCP also changed, not find the right place to execute the code. Local docker encounter some issue, so I give up the code part. Really pity, that is what i am looking for

There are many problems with the Datalab and mismatching between "mlengine" in the course and "ai-platforms" at GCP

Subtitles with lot of problems.

It's ok as an introductory course, but videos are not aligned with labs, and I expected to extract deeper knowledge from it.

The labs are incredibly slow - I also think you need a better background in the technology being taught to understand what is going on. Again, a lot of the quiz q&a aren't covered by the material - also - trying to get support via the discussion forums is a fruitless exercise, my post was ignored - and the course is done.

The content was good, but the labs are awful. It is already done, so you only have to play it. It isn't helpful and I don't think anyone can learn this way. The ideal is to be like the other labs, where you had to input the code yourself. This might be a very hard subject to keep in mind.

This course/specialization is seriously insufficient in order to pass the GCP certification. A lot of knowledge from the Architecture course would still be necessary. The disclaimer about work experience does not cover that.

The most of the subtitles of this course are either out-sync or totally incorrect. Please fix it ASAP!!

the worst course in the Coursera