FileNotFoundError: [Errno 2] No such file or directory: '/app/voices/tts/tts_models--multilingual--multi-dataset--xtts_v2.0.2/config.json'

[joestump]

I'm not sure if this is a regression or not, but reverting to 0.18.2 fixed this issue for me. I found this comment av/harbor#10 (comment) on the Harbor project that was identical to my issue, however despite multiple deletes/restarts I couldn't get latest to load a full model.

The result is only model.pth is in the /app/voices/tts/tts_models--multilingual--multi-dataset--xtts_v2.0.2/ directory.

I tried running docker exec -it openedaispeech /app/download_voices_tts-1-hd.sh xtts_v2.0.2 and it would re-download, but again only the model.pth would show up. I reverted to 0.18.2 and it downloaded first thing no issues.

[matatonic]

Hi, first off, don't use xtts_v2.0.2 unless you have a very specific reason to, just use 'xtts' which will be the latest.
Something is certainly off here, the download scripts haven't changed since 0.14.0 and 0.18.2 is the latest release.