EveryVoiceTTS / EveryVoiceTTS/EveryVoice
Filelist produces incorrect synthesized outputs with incorrect headers
Open
bug
question
- Dominant language
- Python
- Stars
- 45
- Forks
- 4
- Avg merge
- 1d 8h
- Merged PRs (30d)
- 14
Description
When passing a filelist to `--filelist` in the `everyvoice synthesize` command with headers `utterance|basename` instead of, say, `characters|basename`, it seems to alter the input text sequence. We should investigate and confirm the steps to reproduce.
Contributor guide
Assessment
This issue has not been assessed yet.