Right now, I’m currently in Berlin, a place I’ve wanted to visit for over a year now. I took Friday off work and headed here for the weekend. While here, I (of course) made a few stops to view some carillons, but I was only able to play the one in the Tiergarten. One of them, in the Französischer Dom, is a super strange instrument as its almost entirely enclosed, meaning its very hard to actually hear the bells from the outside, and extremely loud in the tower itself.

In terms of research, my work has mostly been the same since yesterday, with further testing for model training parameters. However, as most of this is done now, I’m looking into other ways of showing model accuracy such as a confusion matrix. This would help me visualize which letters are being incorrectly transliterated by the model the most.