As someone who recently took up this course, I must say it has been an incredible journey into the world of generative AI! The course covers a perfect blend of theoretical and hands-on learning, making complex concepts like LLM architectures and generative models feel much more approachable. What stood out to me was how clearly the differences between various generative models were explained, including their unique training approaches. The module on building a chatbot using the Hugging Face transformers library was especially exciting. It’s amazing how we got to implement something so practical and see it come to life! The course didn’t just stop at theory; it also guided us through implementing a data loader with PyTorch's DataLoader class, which gave a solid understanding of real-world application. Overall, this course is perfect for anyone wanting to dive deep into generative AI, particularly in the field of NLP. Highly recommend it!

Extremely comprehensive course, goes in great detail about all the different Generative AI architectures and models, and the different tokenization methods used that enable us to have NLP, it explains them thoroughly, and then provides you with a lab so that you get to experiment and apply this info in a practical setting. Personally I respect courses that truly give you fundamental information, and not just narrate general information, through experience I noticed IBM courses out of all course providers care about making their course as useful as possible, many other course creators spend more time setting up cameras in professional studios with flashy editing, to make their course look good, rather than being informative. IBM in the other hand is all about knowledge, and nothing more.

I recently completed the course "Generative AI and LLMs: Architecture and Data Preparation," and it was an outstanding experience! The content was well-structured and provided deep insights into the architecture and data handling essential for working with large language models. The instructors were knowledgeable and engaging, making complex concepts easy to grasp. I highly recommend this course to anyone looking to enhance their understanding of generative AI!

I am pretty much new to NLP data preparation. However this course made me comfortable with Date preparation activities.

A clear introduction on GAI and LLM, with some exercises to get familiar with the implementation

I love this course, it is what I was looking for a long time, Thank you IBM team !

Aborda los conceptos clave y te enseña a implementarlos, muy buen curso.

A Faire soigneusement beaucoup d informations .

This course is excellent for learning AI.

Excellent Course!

Amazing Course

Good content

I agree that the course material contained a lot of relevant information and found it highly informative. However, having a human instructor walk through the code step by step would have elevated it to a phenomenal level. The robotic video presentations and interspersed readings weren't as engaging; a fully human-led video-audio format throughout would have been much more effective.

I highly recommend using a human to deliver the lectures, which might enhance student engagement. Great introductory course.

I think it is too easy to get certification. The difficulty of the examination should be increased.

The content in the lectures is complex but the slides are not made available to download. Also the Cheat Sheets and other similar materials are presented in weird "windows" that also do not make them available for download. This is a first for me in a Coursera course and I'm find it not very conducive to learning. These material should be easily available. Not certain I will complete the full Specialization if the materials are not made available.

I found the course on LLMs to be a solid introduction, particularly appreciating the cheatsheet and experiments included. However, the requirement to purchase a $49 certificate to complete the course felt excessive. The course producer should learn from many other courses on Coursera, completing the course should be free with the option to purchase the certificate as an add-on.

Tokenizer & dataloader are quite important parts but I am surprised by how shallow they are touched and how easy are the quiz questions.

very superficial

This is an extremely un-educational course and a hugely frustrating experience from a seasoned learner's perspective. It is fundamentally a voice going through whatever info appears on a series of slide stacks just like a robot would. There is no effort whatsoever in trying to make you understand anything and no actual time to do it while you hear the voice running through the text. The only practical way to acquire the knowledge contained in the slides is to stop every 30 seconds, read, look up somewhere else (google, forums...) to actually consolidate the knowledge, and then click the play button again. I am a proficient and quick learner and have a good background on neural networks and gen-ai use and programming and I was completely incapable of following the explanations given in this course after the first 5 minutes: boring, robotic, ineffectual, and very counter-productive. This has been a very frustrating experience from a learner's perspective and my advice would be to take this course out and rethink it from scratch, as it certainly does not serve its purpose at all. I left one star because the knowledge is actually there in the slides. The information is contained in the slides, it is simply never actually conveyed to the audience by the reading robotic entity in charge.