Michael McAuliffe

Is kaldi installed in your aligner environment? Can you run `compute-mfcc-feats` by itself and see if there's a more informative error message?

The "spn" are generated from bracketed words or words not in the dictionary. In the root temporary directory (~/Documents/MFA/{corpus_name}), there will be an "oovs_found.txt". If you add those words to...

Hi, 0.8.1 is quite old, so I don't know how much I can support that. I would recommend either upgrading to the latest version: https://montreal-forced-aligner.readthedocs.io/en/latest/getting_started.html#getting-started-ref, or use the latest stable...

Neither, it just uses the GMM-HMM pipeline for training through LDA+SAT. The experiments that we did with nnet2 several years had very slight benefits if at all over the GMM...

Ah yes, it changed to the more general configuration flag `uses_speaker_adaptation`, so you would use it via `mfa align .... --uses_speaker_adaptation False` and that will disable it. I'll try to...

I'm not too sure, the old style of using frozen executables have some weird behavior. Can you try following the newer instructions for installation and see if that works better...

I wasn't planning on it. The frozen release model was a source of a lot of headaches for me. It's possible that things have gotten better in that space in...

Can you update to the latest version? There's an additional check for a common issue with an incompatible platform that the binaries weren't compiled for. If that doesn't pop up...

Are you on the latest version of MFA? I'm not able to replicate this, and each file has a unique integer ID internally now (specifically to get around this and...

Yeah, that would be my thought as well. The original use case is geared towards datasets that have lots of speakers, rather than just one. In principle, it shouldn't be...