This is what happens when a language model inference has insufficient window size for penalizing repetition.
Hi, it wasn't my fault, I myself was shocked to be honest, it was a bug, I didn't on purpose
reply