Homework size are TOO large! One star given. One additional for that this course is highly challenging.

Not much value in auditing this class w/o access to the coursework itself.

Core Issues: 1. No Actual Teaching The course does not teach. It speed-reads jargon-heavy scripts without breaking down concepts, demonstrating techniques, or scaffolding understanding. Learners are left to reverse-engineer logic from superficial narration. 2. Assignments Require Unintroduced Concepts Key techniques like gradient ascent, latent space traversal, and regularization are required in assignments but never explained. Indexing logic is omitted entirely, despite being critical for assignment success. 3. Auto-Grader Rewards Mimicry The grading system validates surface-level output without checking for semantic control or architectural understanding. Learners can pass by accident or mimicry, not mastery. 4. Optional Content Contains Core Architecture The most important insights—semantic control, projection-based disentanglement, real image editing—are buried in optional readings. These are never integrated into the course or assignments. 5. Bias and Entanglement Are Mentioned, Not Addressed Ethical concerns are gestured at but not operationalized. No tools are provided to detect, measure, or correct bias. Entanglement is acknowledged but never resolved. 6. Instructor Delivery Is Performative Instructional videos are rushed and dense, resembling scripted voiceovers. In contrast, welcome and congratulatory videos are slower and emotionally paced, revealing a prioritization of ritual over rigor. Conclusion: Course 1 does not teach GANs—it performs them. It offers no operational tools, no architectural clarity, and no filtration logic. Learners who succeed do so in spite of the course, not because of it. A full redesign is recommended to align instruction, assignments, and evaluation with the actual demands of the materia

After week 1, not sure what the objective of this course is, seems this is more about PyTorch.

The course material about the concept of GANs is actually quite good, but the assignments are near impossible for me as I am not a PyTorch expert. Trying to figure out the syntax, not really what I was looking for.

A couple of years back, I completed one of their earlier courses (Machine Learning) and it didn’t use any language, but essentially a MatLab like environment to manipulate matrices and vectors.

No sweat, I’ll just cancel the subscription and find another way to learn this stuff.

Had to pause multiple times to listen again or read the English translation at the bottom. Slowing down the lecture along with proper pauses and meaningful visual illustrations can improve this course in a big way.

Content of this course is good, but the way it is presented leaves much to be desired

The coding exercises seem completely unguided by the course, and feel like a waste of my time.

I'm not going to pay you for the time I spend studying pytorch.org

thanks for the hard work, but I feel a lot of places not explained clearly, and the assignment is also not that helpful

how can i create my first GAN without guidence.

Way, way, way too light on the details