Best course on AI

Great course !

Awesome course

very helpful

best course

Nice..!!

It is a nice course to get you into Pytorch and with some insightful views of how some ML algorithms work but adding to the most upvoted review, the synth voice dialogue sometimes doesn't make sense, the inflections on the speech are weird at times, it spells things that come from a text based explanation rather than someone speaking (things like spelling "I E for -for example- and C N N for convolutional neural network among many, many others)... sometimes the voice is talking about one thing and something else is highlighted on the video, time mismatch...

Many grammar mistakes, stuff left in the examples and quizes that doesn't make sense... definitely needs a redaction and content check.

The course material contains some really fantastic information, graphics, and programming assignments. However, the presentation of this material is absolutely terrible! It seems they intentionally tried to make the presentations as boring as possible. The lectures are monotone, the 15 second opening scene is annoying, and the content focuses 70% on the concepts of Deep Learning (which is fine) and 30% on PyTorch. So when you finish you do not feel very skilled with PyTorch.

Finally, ALL of the student complain that the programming environment is very often offline. You cannot do many of the assignments because the "Cognitive Classroom" is usually not working. However, the last lecture f each week contains the Jupyter notebooks for the assignments. You can download and then run them in some other environment like Google Colaboratory or IBM Watson Cloud. Also, most of the programs contain a programming omission that the students have to fix every time. The instructors have not fixed the problem which has been reported to them. So pay attention for the "Pillow Error" in Week 3 because you will be fixing it yourself in most assignments for the next 4 weeks.

Bad !!!!! Many mistakes, questions too easy !!! I am really disapointed

Excellent Course. I love the way the course was presented. There were a lot of practical and visual examples explaining each module. It is highly recommended!

By this course I can understand the basic concept for building neural network or deep lerning model using PyTorch. Very Good course to beginner.

It would be better to add a section explaining the experiment code of the famous paper.

Quiz questions are very easy. Graded Programming Assignments would have been better.

clear instruction, great ilustration and process description. Thank you so much

incredible course covering from basics to a satisfaction level

Highly recommended course for students

Easy-to-follow course for pytorch

I really enjoy this course!!!

very well structured course.

Aprendí muchísimo. Gracias.

Labs were detailed one.

Fantastic explanation

It was Good !!

I appreciate this course. Its really amazing course and if you are a beginner in Deep Learning and want to use and learn Pytorch then this course is really good to start.

One thing about this course is that some important topics like RNN, R-CNN , text and sentiment analysis, time series are not included in this course which I think should be included.

This course is difficult to rate as a learning experience. There are some very good parts yet there is also some very poor material. I would say that if you are already very familiar with machine learning and Python BEFORE taking this course, you can still draw some useful learnings on how PyTorch can be applied to various problems, and how to create convolutional neural networks with it; but if you are uncertain about some of the key concepts, this course may only end up making things worse for you.

To give an idea of the problems, there are issues like:

- When explaining the train/validation/test data logic and how validation data can be used to prevent overfitting, the videos keep calling training data test data.

- Pytorch is used for some really fancy stuff like defining functions and datasets, but then those functions are not parametrized in any sensible way – meaning if you want to compare loss functions from two different initialisations of the model weights, you are expected to define a new function so you can just change the variable “LOSS” to “LOSS2”, rather than just passing the loss function as a parameter or just initializing or returning it. Given the Pytorch logic is not your regular Python stuff, a best practice should be provided – it is definitely not writing a new function every time.

So be warned: if you know what you are doing, and simply want to learn how to do it with Pytorch, this may still be a decent course for you, just ignore all the stuff where the instructors make mistakes (and they are plenty, also in incorrect quiz answers). But if you feel at all uncertain, I suggest you hone your machine learning skills elsewhere, because otherwise this course will leave you totally confounded on even the very basics of machine learning.

On the upside then, you learn Pytorch through repetition. In the beginning, the logic appears very intimidating, but then you gradually learn the logic and you can do some very impressive stuff quite easily in the end. Be prepared for the amount of repetition, however - first the stuff is shown on a video, then you run the exactly same stuff in a lab, and unfortunately the Skills Lab is not at all efficient for some of the stuff - I ended up downloading the notebooks and using them on my Watson Studio account for much faster performance.