Getting IndexError: list index out of range error with libritts-train-clean-100 dataset

[mlrober]

Hi Team,

Thanks for the wonderful project for helping with this tool.

While i was running this tool on MAC OSX binary, i have done the following steps:

1. I have used Libritts-train-clean-100 dataset to align.
2. when i execute the below command i'm getting the error as:
   Command:
   mfa align -d -c ./LibriTTS FastSpeech2/lexicon/librispeech-lexicon.txt english ./FastSpeech2/preprocessed_data/LibriTTS

Error:

Traceback (most recent call last):
File "/Users/rajasekar.a/tts/vir/bin/mfa", line 8, in
sys.exit(main())
File "/Users/rajasekar.a/tts/vir/lib/python3.9/site-packages/montreal_forced_aligner/command_line/mfa.py", line 381, in main
run_adapt_model(args, unknown, acoustic_languages)
File "/Users/rajasekar.a/tts/vir/lib/python3.9/site-packages/montreal_forced_aligner/command_line/align.py", line 184, in run_align_corpus
from montreal_forced_aligner.command_line.mfa import align_parser, fix_path, unfix_path, acoustic_languages,
File "/Users/rajasekar.a/tts/vir/lib/python3.9/site-packages/montreal_forced_aligner/command_line/align.py", line 131, in align_corpus
except Exception as _:
File "/Users/rajasekar.a/tts/vir/lib/python3.9/site-packages/montreal_forced_aligner/aligner/pretrained.py", line 146, in export_textgrids
convert_ali_to_textgrids(self.align_config, output_directory, ali_directory, self.dictionary,
File "/Users/rajasekar.a/tts/vir/lib/python3.9/site-packages/montreal_forced_aligner/multiprocessing/alignment.py", line 588, in convert_ali_to_textgrids
parsed = parse_ctm(word_ctm_path, corpus, dictionary, mode='word')
File "/Users/rajasekar.a/tts/vir/lib/python3.9/site-packages/montreal_forced_aligner/textgrid.py", line 44, in parse_ctm
cur = current_labels[cur_ind]
IndexError: list index out of range

Can someone help me to resolve this issue?

[Weijia221B]

I got the same error, have no idea why it happened

[chaksam]

Getting same error while processing LJSpeech dataset on Ubuntu system:

---

INFO - Setting up corpus information...
INFO - Number of speakers in corpus: 1, average number of utterances per speaker: 13100.0
INFO - Parsing dictionary without pronunciation probabilities without silence probabilities
INFO - Creating dictionary information...
INFO - Setting up training data...
Generating base features (mfcc)...
Calculating CMVN...
INFO - Done with setup!
INFO - Performing first-pass alignment...
INFO - Calculating fMLLR for speaker adaptation...
INFO - Performing second-pass alignment...
Traceback (most recent call last):
File "/home/anaconda3/envs/aligner/bin/mfa", line 8, in
sys.exit(main())
File "/home/anaconda3/envs/aligner/lib/python3.8/site-packages/montreal_forced_aligner/command_line/mfa.py", line 379, in main
run_align_corpus(args, unknown, acoustic_languages)
File "/home/anaconda3/envs/aligner/lib/python3.8/site-packages/montreal_forced_aligner/command_line/align.py", line 179, in run_align_corpus
align_corpus(args, unknown_args)
File "/home/anaconda3/envs/aligner/lib/python3.8/site-packages/montreal_forced_aligner/command_line/align.py", line 127, in align_corpus
a.export_textgrids(args.output_directory)
File "/home/anaconda3/envs/aligner/lib/python3.8/site-packages/montreal_forced_aligner/aligner/pretrained.py", line 146, in export_textgrids
convert_ali_to_textgrids(self.align_config, output_directory, ali_directory, self.dictionary,
File "/home/anaconda3/envs/aligner/lib/python3.8/site-packages/montreal_forced_aligner/multiprocessing/alignment.py", line 588, in convert_ali_to_textgrids
parsed = parse_ctm(word_ctm_path, corpus, dictionary, mode='word')
File "/home/anaconda3/envs/aligner/lib/python3.8/site-packages/montreal_forced_aligner/textgrid.py", line 40, in parse_ctm
cur = current_labels[cur_ind]
IndexError: list index out of range

[hal3003]

Hi there,

I'm encountering the same issue on Fedora 34. I have a custom data set, consisting of four recordings and plain text transcriptions. I generate a dictionary using
mfa g2p english_g2p ~/test_data/input/003.txt ~/test_data/003.dict and the run the alignment using mfa align ~/test_data/input/ ~/test_data/003.dict english ~/test_data/output --clean --verbose.
The output of that command including the error message looks like this:

All required kaldi binaries were found!
Cleaning old directory!
/home/user/Documents/MFA/input/align.log
INFO - Setting up corpus information...
INFO - Number of speakers in corpus: 1, average number of utterances per speaker: 1.0
INFO - Parsing dictionary without pronunciation probabilities without silence probabilities
INFO - Creating dictionary information...
INFO - Setting up training data...
Generating base features (mfcc)...
Calculating CMVN...
INFO - Done with setup!
INFO - Performing first-pass alignment...
INFO - Calculating fMLLR for speaker adaptation...
INFO - Performing second-pass alignment...
Traceback (most recent call last):
File "/home/user/.virtualenvs/mfa/bin/mfa", line 8, in
sys.exit(main())
File "/home/user/.virtualenvs/mfa/lib/python3.9/site-packages/montreal_forced_aligner/command_line/mfa.py", line 379, in main
run_align_corpus(args, unknown, acoustic_languages)
File "/home/user/.virtualenvs/mfa/lib/python3.9/site-packages/montreal_forced_aligner/command_line/align.py", line 179, in run_align_corpus
align_corpus(args, unknown_args)
File "/home/user/.virtualenvs/mfa/lib/python3.9/site-packages/montreal_forced_aligner/command_line/align.py", line 127, in align_corpus
a.export_textgrids(args.output_directory)
File "/home/user/.virtualenvs/mfa/lib/python3.9/site-packages/montreal_forced_aligner/aligner/pretrained.py", line 146, in export_textgrids
convert_ali_to_textgrids(self.align_config, output_directory, ali_directory, self.dictionary,
File "/home/user/.virtualenvs/mfa/lib/python3.9/site-packages/montreal_forced_aligner/multiprocessing/alignment.py", line 588, in convert_ali_to_textgrids
parsed = parse_ctm(word_ctm_path, corpus, dictionary, mode='word')
File "/home/user/.virtualenvs/mfa/lib/python3.9/site-packages/montreal_forced_aligner/textgrid.py", line 108, in parse_ctm
cur = current_labels[cur_ind]
IndexError: list index out of range

Notably, the error occurs only for one of the files I'm trying to align.

[mlrober]

Hi Admin,

Can you help us on ths issue

[mmcauliffe]

Ah sorry yes, I tracked down the issue to one of the words ending in "-", I believe it's something like "mr goodwill-" that's causing the issue in the current release version. I should have a fix out for it soon, but got sucked down the rabbit hole of refactoring the output code and optimizing it. If you need a fix asap, you can try downgrading to version 2.0.0a21, that should be before I changed how compounds were being parsed, so should work.

…

On Tue., Sep. 21, 2021, 8:03 a.m. mlrober, ***@***.***> wrote: Hi Admin, Can you help us on ths issue — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#329 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAVJOT4JM7RSYQQYPY3VG53UDCNETANCNFSM5DZBGRUQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.