config.yaml

http_service:
  host: 0.0.0.0
  port: 7860
  debug: false
model_config:
  chinese_roberta_wwm_ext_large: bert/chinese-roberta-wwm-ext-large
  bert_base_japanese_v3: bert/bert-base-japanese-v3
  bert_large_japanese_v2: bert/bert-large-japanese-v2
  deberta_v2_large_japanese: bert/deberta-v2-large-japanese
  deberta_v3_large: bert/deberta-v3-large
  deberta_v2_large_japanese_char_wwm: bert/deberta-v2-large-japanese-char-wwm
  wav2vec2_large_robust_12_ft_emotion_msp_dim: emotional/wav2vec2-large-robust-12-ft-emotion-msp-dim
  clap_htsat_fused: emotional/clap-htsat-fused
  erlangshen_MegatronBert_1_3B_Chinese: bert/Erlangshen-MegatronBert-1.3B-Chinese
  vits_chinese_bert: bert/vits_chinese_bert
  hubert_soft_0d54a1f4: hubert/hubert_soft/hubert-soft-0d54a1f4.pt
  dimensional_emotion_npy: dimensional_emotion_npy
  dimensional_emotion_model: dimensional_emotion_model/models.yaml
  g2pw_model: G2PWModel
  chinese_hubert_base: hubert/chinese_hubert_base
tts_config:
  models_path: models
  auto_load: true
  models: []
admin:
  id: 0
  username: admin
  password: admin
system:
  device: cuda
  upload_folder: upload
  cache_path: cache
  clean_interval_seconds: 3600
  cache_audio: false
  api_key_enabled: false
  api_key: aaa
  is_admin_enabled: true
  admin_route: /admin
  data_path: data
log_config:
  logs_path: logs
  logs_backupcount: 30
  logging_level: DEBUG
language_identification:
  language_identification_library: langid
  espeak_library: null
  language_automatic_detect: ''
reading_config:
  interlocutor:
    model_type: VITS
    id: 0
    preset: default
  narrator:
    model_type: VITS
    id: 0
    preset: default
vits_config:
  dynamic_loading: false
  id: 0
  format: wav
  lang: auto
  length: 1.0
  noise: 0.33
  noisew: 0.4
  segment_size: 50
  use_streaming: false
w2v2_vits_config:
  id: 0
  format: wav
  lang: auto
  length: 1.0
  noise: 0.33
  noisew: 0.4
  segment_size: 50
  emotion: 0
hubert_vits_config:
  id: 0
  format: wav
  length: 1.0
  noise: 0.33
  noisew: 0.4
bert_vits2_config:
  id: 0
  format: wav
  lang: auto
  length: 1.0
  noise: 0.33
  noisew: 0.4
  segment_size: 50
  sdp_ratio: 0.2
  emotion: 0
  text_prompt: Happy
  style_text: None
  style_weight: 0.7
  use_streaming: false
  torch_data_type: ''
gpt_sovits_config:
  hz: 50
  is_half: false
  id: 0
  lang: auto
  format: wav
  segment_size: 30
  top_k: 5
  top_p: 1.0
  temperature: 1.0
  use_streaming: false
  batch_size: 5
  speed: 1.0
  presets:
    default1:
      refer_wav_path: null
      prompt_text: null
      prompt_lang: auto
    default2:
      refer_wav_path: null
      prompt_text: null
      prompt_lang: auto