I was disappointed because the lecture materials, paper summaries, and example codes all felt like they were written by GPT. I paid for this course specifically for the fine-tuning practice, but the instructor just displays the code and reads the comments as they are. It makes me feel like there's no difference between this and just asking GPT to write example code with comments. In the first practice session, there was no check for dataset imbalance, and the model was trained on data where all labels were 0; when it outputted 0, the instructor said it was correct... I'm seriously debating whether I should continue with the remaining lectures.