feat: TTS Stack - Chatterbox Multilingual + Whisper STT

- Chatterbox TTS Server (Multilingual, 23 Sprachen, Voice Cloning)
- Whisper STT Server (faster-whisper-small, CPU)
- RTX 4060 GPU auf Tdarr VM (node2, 10.2.1.104)
- Voice Profile: chantal.wav
- Chantal Telegram Voice Integration
This commit is contained in:
feldjaeger 2026-04-17 12:37:44 +02:00
commit 6115f3bc09
4 changed files with 167 additions and 0 deletions

68
config-chatterbox.yaml Normal file
View file

@ -0,0 +1,68 @@
server:
host: 0.0.0.0
port: 8004
use_ngrok: false
use_auth: false
auth_username: user
auth_password: password
log_file_path: logs/tts_server.log
log_file_max_size_mb: 10
log_file_backup_count: 5
model:
repo_id: chatterbox-multilingual
tts_engine:
device: cuda
predefined_voices_path: voices
reference_audio_path: reference_audio
default_voice_id: Emily.wav
paths:
model_cache: model_cache
output: outputs
generation_defaults:
temperature: 0.8
exaggeration: 1.3
cfg_weight: 0.5
seed: 0
speed_factor: 1.0
language: en
audio_output:
format: wav
sample_rate: 24000
max_reference_duration_sec: 30
save_to_disk: false
ui_state:
last_text: '[gasp] Okay, we have fifteen minutes left! Fifteen minutes! [groan]
How have we been in here for forty-five minutes and only solved two puzzles?!
[clear throat] Let''s think logically. [sniff] This room smells like fear and
bad decisions. [cough] And also someone''s very strong cologne, Kevin. [sigh]
Okay, the clue says "the answer lies where time stands still." [chuckle] Great,
very helpful, very straightforward. [shush] Everyone stop talking for one second!
[gasp] Wait, the clock on the wall! It''s stuck at three fifteen! [laugh] That''s
it! That has to be it! [groan] It''s not the combination. Of course it''s not
the combination. [sniff] Why would anything in this room be logical? [sigh] Has
anyone checked under the carpet? [gasp] There''s a hidden compartment! [chuckle]
Kevin, you''re standing on a clue! Classic Kevin! [clear throat] Okay, it''s another
riddle. "What has keys but no locks, space but no room, and you can enter but
can''t go inside?" [groan] Who writes these things?! [cough] A keyboard! It''s
a keyboard! [laugh] There''s a keyboard painted on the wall! [gasp] Eight minutes!
We have eight minutes! [sigh] If we don''t escape this room, we''re never speaking
of this again. [shush] Focus, everyone! [chuckle] We''re so close I can taste
freedom! [sniff] That or more of Kevin''s cologne!
'
last_voice_mode: predefined
last_predefined_voice: Emily.wav
last_reference_file: Gianna.wav
last_seed: 3000
last_chunk_size: 240
last_split_text_enabled: true
hide_chunk_warning: false
hide_generation_warning: false
theme: light
last_preset_name: "\u26A1 Turbo: Escape Room Panic"
ui:
title: Chatterbox TTS Server
show_language_select: true
max_predefined_voices_in_dropdown: 50
debug:
save_intermediate_audio: false