language identification via fasttext

cleaning GPU resources on pipeline unloading, because it also pushes tensors to GPU: voices

tts: unloading on idle

compat mode: language from voice as first letter

response_format other than wav (like mp3) streamable and not

fastapi.UploadFile: sync vs async

/v1/audio/speech: raw implementation

kkr/app service

fixed mypy issues

pyproject.toml refactoring

.aider files to be ignored

kokoro: can speak in eo

README.md

comment polishment

disabled pushing logger.warning()-s to sentry because of transformers: 'Device set to use'

/metrics

logging to sentry works

fix: verbose_json format

LEGO_ZONE_FROM_DOMAIN=1 to extract zone name from challenge fqdn

Prepare release v4.22.2

chore: update goreleaser configuration

acme-dns: use new registred account (#2445)

Detach v4.22.1

readme: the habr publication

readme: the habr publication

LICENSE

readme fix