I am taking the entire specialization, so this is my 4th review. By this point I have just given up on how poorly created and managed this course is. Just here to rate it 1 star and discourage anyone from taking it. Just do the HuggingFace NLP course better. And just a quick example of how bad this course is … week 3 has a duplicate video, basically 2 different videos playing the exact same video … automating packaging with azure container registry and automating packaging with docker hub … so, no big deal, technical error … however, 2 different learners have pointed this out 8 months ago and 5 months ago at the writing of this review and it still hasn’t been fixed. This is all you need to know about the level of this course and specialization. I know Noah is rolling out a second book on MLOps – Implementing MLOps in the Enterprise around Christmas `23 so I was thinking I’d threat myself and buy both his MLOps books for the holidays. This specialization completely discouraged me from doing that.

You need prior knowledge of Docker and Azure to understand a lot of videos. Azure account in not provided, so have to create and use your own. There is no instruction for the lab sessions, so you have to keep going back to the videos, which takes a lot of time.

This course provides a basic understanding of MLFlow and HuggingFace platforms. It can be better if: - More hands-on lab with grading - More concise content. I feel Week 2, Week 3 and Week 4 can be combined into 1-2 weeks. Many contents are duplicated.

Barely scratching the surface. Extensive reference to free online material for self-study. Felt disappointing, had expected way more.

"Poking around" in demo videos like the ones found on YouTube does not make a course. I also doubt any serious company would use Gradio and HuggingFace Spaces for MLOps. These are designed for prototyping and sharing experiments, which are not MLOps.

This course helped me dive deep into essential tools for deploying and managing machine learning models at scale. I now feel confident in using MLflow for tracking experiments and Hugging Face for deploying state-of-the-art NLP models.

Great learning resources that will be useful long after completing the course, concise presentations, and clear explanations of all topics

Very good course, with effective exercises

It's an amazing experience.

Excellent content hepful for ML/Ops aspirers

.

Quite literally the lowest-quality course I have ever taken. The instructors are completely clueless (even despite the fact that at the level this course operates, these are simple things), at times, the mind-boggling level of amateurish fumbling during lectures becomes downright tragicomic. The course material lacks any coherent concept, is utterly superficial, and most of the time barely rises above kindergarten level, consisting mainly of basic practical demos. One of the lecturers has such a poor speaking style that it barely qualifies him to deliver presentations at all. Beyond these issues, I could go on at length detailing the severe flaws and shortcomings, but given that they ‘honored’ me with such a disaster of a course, I find even the effort of writing this review hardly worth it. I honestly cannot understand how Coursera could allow its name to be associated with something like this, it's a terrible reflection on the platform as well.