Opiniones y comentarios de aprendices correspondientes a Introduction to TensorFlow for Artificial Intelligence, Machine Learning, and Deep Learning por parte de deeplearning.ai

I would say it is a introduction to Keras rather than Tensorflow. The course not really touch tensorflow, it all about the high level API which is Keras in TF.

The grader is very buggy. Coded a 1-Conv2d & 1-MaxPool model and caused the grader to run out of memory. However, a more complicated model passes the solution.

the notebooks have a poor explanation of what should be done and unless you delete the last two cells every time you won't be able to submit

While useful, the experience with submitting the assignements was really inferior to what I'm used to with Coursera and DeepLearning.AI

This course needs to be more beginner friendly....it directly jumps to advanced concepts without clearing the intermediates

Really superficial overview of tensorflow and deep learning. Very few concepts were explained in any real depth.

Resources shown in the video were not included as web links, so the course couldn't be completed

A little too complex for beginners. Content must be explained from a novice point of view

Some reading exercises had missing links and some code used a deprecated function.

Not a great experience with the assignments, especially the last one.

This course teach how to use Keras more than using Tensorflow

It´s well explained but way too basic and short.

这课挺水的， 不如 deep learning specialization多矣。

notebooks need work from the instructors

Course was not rigorous enough

Too basic.

meh

This specialization is false advertising. It does NOT prepare you for the Tensorflow certification exam. It’s especially disappointing after taking the fantastic specialization by Andrew Ng, and makes this specialization feel like a cheap cash grab by Coursera and DeepLearning.ai. This series of courses fails to prepare you for three reasons:

1 – The certification exam is done on whatever is the current version of Tensorflow (v2.6 as of writing). You can’t expect a specialization like this to update every minor release, but much of the coding is still on the v1.X version!

2 – The certification exam requires you to work in the PyCharm IDE. The IDE doesn’t even get a mention in this specialization and it is all done through Google Colab.

3 – The material is covered at a very superficial level. I was hoping to walk out of it feeling confident in using Tensorflow on novel problems, but I’ve barely learned anything about Tensorflow that I didn’t already get from Andrew Ng’s specialization. There’re a few minutes of lectures (some weeks less than 10 minutes). The programming assignments are either pathetically easy, or lack any guidance on what to do (seriously sometimes there’s no instructions at all, you have to guess what to do by the variable names), or both.

Save your time and money and go elsewhere to learn Tensorflow.

This course is FULL of errors (both in code and math), inconsistencies and wrong explanations. I tried to document them, but I just gave up, because it is so many of them... For example the math which explains multiclass classification (Week 2 Video “Coding a Computer Vision Neural Network”) is wrong – the output of multiclass classification should be pseudoprobabilities, not numbers ranging 1 to 9… There are also unsolved problems reported in GitHub (https://github.com/lmoroney/dlaicourse). It seems like they really don’t care about correctness, completeness and quality of this course… If you want to learn TensorFlow I highly discourage to use this course - you will just learn wrong things and would have to unlearn them later... Also graded exercises are in TensorFlow 1.x and materials are prepared for TensorFlow 2.x which means that sometimes the code from materials does not work in graded exec, eg. logs.get('accuracy') does not work in tf 1.x and you need to use logs.get('acc'). I did this course only to get some practice and pass TensorFlow exam, because I'm academic who works with PyTorch.

First of all, it's an introduction to Keras and not Tensorflow.

Secondly, the explanations the author gives are lacking/misleading.

For example,in week one the net didn't learn exactly the hypothesis 2x-1 for other reasons than the ones he mentioned (oh,and the net did not give some kind of a probabilistic interpretation to the data...).

I am not sure why Andrew NG (who is the best instructor in the world to my mind) allowed this kind of instructor to be branded as deeplearning.ai.

This course claims to be over 28 hours, however, I was able to finish it (watching every video, reading every bit of text, doing every exercise) in less than 6 hours. The lab work is a complete copy-and-paste of the examples covered in the course. If you want to "buy a credential" take this course, but if you want to actually learn anything, take another course.

This is a poor course. The course assignments often fail for mysterious reasons (Grader timed out, Grader ran out of memory), and the course points to outside resources more than its own explanation.

I found the explanation deeply lacking. I would suggest not taking this course, and I will not be taking the remaining courses in this specialization.

the instructor did not give us enough explanation for the code written, it is just reading it. many things he added in his code does not make sense for the beginners like me. in addition that he said that it is out of the scope of the course. so why did you add it in your code. i feel the instructor should explain things more than that

This course does not go very much into detail and way too much time is given for easy exercises and homework. The homework contains a lot of bugs, which need to be fixed since students waste a lot of time debugging errors which are not due to their fault. Furthermore, the homework is just about copying the code given in the videos.

The videos don't contain much usefull information, but only a demonstration of the most basic concepts of tensorflow. Practical assignments does not aim at teaching you any skills, but copying code from one notebook to another. And after this course I actually have some concerns about the author qualification on the subject.