MFA not aligning properly[BUG]
I've been trying to run MFA for a project and while it runs, only the first maybe 5 seconds of each clip is actually aligned. I had issues getting it to run in the first place, mainly with the dictionary used. I am using librispeech, and I kept getting errors that there were not tabs between the words and their phonemes. I looked for a while for a different version of librispeech, but they all had the same issue. The only way I could get it to recognize the phonemes properly was to manually replace the spaces with tabs in the text files and run MFA again. This allowed it to run, but again, the clips are not properly aligned.
I run MFA in conda in Windows 10 and use the librispeech dictionary. The clips have around 2-3 speakers in them, and the output are textgrid files.