Is There a Hyper-Period in Quantamalgarithms?

We used the mnis dela set, and we did a unsupervised clustering on the mnist images after some simple picier. And somehow wo can look at at each situration how the accuracy is improving. And then we launch the same experiment using camins. And we played with this ro parmiter somehow. So what we observe is that it's kind of learning in the same way. It's converging anyway. If the ro is small, of course, it becomes exactly as the camins. But if the ro is too big, it's still convert but you can see that it has more trouble to do it. And it reached a far lower accuracy