search

r9y9 / deepvoice3_pytorch

PyTorch implementation of convolutional neural networks-based text-to-speech synthesis models
https://r9y9.github.io/deepvoice3_pytorch/
Other
1.97k stars 485 forks source link

problem in vctk preprocessing. #133

Closed vega189674 closed 5 years ago

vega189674 commented 5 years ago

Hello. I have encountered a problem while I run vctk_preprocess.py. I can not understand why the file is not loaded. this is a log. How can I solve this problem? Or may I ignore this problem?

... Traceback (most recent call last): File "./extract_feats.py", line 1439, in keep_silences, full_features) File "./extract_feats.py", line 528, in extract_intermediate_features with open("do_ehmm", "r") as f: FileNotFoundError: [Errno 2] No such file or directory: 'do_ehmm' cp ./latest_features/merlin/misc/scripts/alignment/phone_align/full-context-labels/mono/.lab ../../data/lab/p225 cp: './latest_features/merlin/misc/scripts/alignment/phone_align/full-context-labels/mono/.lab' を stat できません: そのようなファイルやディレクトリはありません rm -rf ./latest_features 1%|▍ | 1/108 [00:42<1:15:00, 42.06s/it]../../data/VCTK-Corpus/wav48/p226 ../../data/VCTK-Corpus/txt/p226 python ./extract_feats.py -w ../../data/VCTK-Corpus/wav48/p226 -t ../../data/VCTK-Corpus/txt/p226 ^CTraceback (most recent call last): File "prepare_vctk_labels.py", line 43, in ...

And that's the directory where it is running.

src/ehmm/bin ├── ConvertFeatsFileToBinaryFormat ├── FeatureExtraction ├── Makefile ├── ScaleBinaryFeats ├── check_lab.pl ├── comp_dcep.pl ├── do_ehmm ├── edec ├── ehmm ├── phfromutt.pl ├── phseq.scm ├── rem_ssil.pl ├── rm_nan_utts ├── scale_feat.pl ├── seqproc.pl ├── sl2nm.pl └── sym2nm.pl

r9y9 commented 5 years ago

Did you confirm if https://gist.github.com/kastnerkyle/cc0ac48d34860c5bb3f9112f4d9a0300 works?

https://github.com/r9y9/deepvoice3_pytorch/tree/master/vctk_preprocess#note

vega189674 commented 5 years ago

Thank you very much. I had overlooked it. There was an error when executing extruct_feats.py , so I will fix it.