Modern Natural Language Processing in Python (Udemy.com)

Awesome! your presentation was masterclass. Good content and well-structured presentation. I really enjoyed this course, learnt a lot, informative and fun approach to presenting the material. The instructor was extremely knowledgeable, making complex topics easy to understand through engaging, practical examples and also the trainer was incredibly supportive and responsive, provided clear, actionable advice. Thank you so much to instructor.

This is the only tutorial in Udemy in which we can find the English to French translation model. Very nice explanation about the Transformer model and all its detailed implementation.

This is an advanced course and covers recent concepts like Transformer, Positional Encoding, Encoder, Decoder etc. It requires some prior knowledge of CNN, RNN, LSTM, seq-to-seq learning. The course also covers building a Transformer from scratch using the "Attention is all you need" research paper as basis with English to French translation as a use case. However, the course contents are very small and explanations are "verbal". It should be extended with more visuals and use of white board to convey the concepts well. Also, content should be simplified by using some examples while explaining the concepts. I found Hedu Math of Intelligence YT videos and Illustrated Transformer blog by Jay Alammar helpful to understand these concepts. The coding part should also be simplified by explaining the logic & design, and showing running samples of expected output like matrix/vector shapes, sample data etc. It becomes very hard to keep up with the intuitions and code implementations. Overall, the course covers an advanced topic which is hard to apprehend initially and requires some prior knowledge; but it is also a great opportunity to simplify the concepts given that there are not many resources available for the same.

A good course on cutting edge topics. Some parts of the transformer were quite heavy and difficult to follow. I know it is difficult to visualize the internals of such a complex system, but I would have preferred visualization of the dimensions of each output of the different steps. It would have made it easier to follow.

This course is a good refresher of the CNN module in Deep Learning A-Z. That aside, The NLP content comprising the Sentiment analyzer and the Transformer offers excellent intuitive and practical knowledge. The Transformer architecture is quite difficult to understand, but it is well explained.

The project works are good subjects, which I think the instructor has worked a lot to prepare. There could be more explained about the datasets, how people can get started with them, what if you select another language dataset, how would it impact to using the regex to sort special characters of each language? There is apart relatively little information about the NLP research, which has progressed a lot recently, so it would be perhaps a good idea to explain apart word2vec/GloVE, about BERT etc. Anyway, I think the course has a great potential and in the future will be even more valuable!

The content is awesome. Viewers would also usually expect the pronunciation to be clear. It was a wonderful effort though kudos!

It's as I expect from the SDS team. Really clear and concise explanations, in a step-by-step approach. You should already have an understanding of basic CNN and NLP to get the most out of the course.

A bit too much apriori knowledge required to understand fully. Had to go elsewhere to read up.

GOOD

good

Voice is not much clear . Specially for non native english students