Text-to-Speech
Fairseq
French
audio
File size: 610 Bytes
fd9f8a3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
bpe_tokenizer:
  bpe: sentencepiece
  sentencepiece_model: spm_char.model
features:
  eps: 0.01
  f_max: 8000
  f_min: 20
  hop_len_t: 0.011609977324263039
  hop_length: 256
  n_fft: 1024
  n_mels: 80
  n_stft: 513
  sample_rate: 22050
  type: spectrogram+melscale+log
  win_len_t: 0.046439909297052155
  win_length: 1024
  window_fn: hann
global_cmvn:
  stats_npz_path: gcmvn_stats.npz
speaker_set_filename: speakers.txt
transforms:
  '*':
  - global_cmvn
vocab_filename: spm_char.txt
vocoder:
  type: hifigan
  checkpoint: hifigan.bin
  config: hifigan.json
hub:
  phonemizer: ipa
  lang: fr
  speaker: 1291