Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

fix Lint check

This commit is contained in:
Edresson 2020-07-29 21:21:42 -03:00 committed by erogol
parent a2ee48c28e
commit f91b9eeda1
2 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
mozilla_voice_tts/tts/datasets/preprocess.py
View File

@ -229,14 +229,14 @@ def vctk(root_path, meta_files=None, wavs_path='wav48'):
items = [] items = []
meta_files = glob(f"{os.path.join(root_path,'txt')}/**/*.txt", recursive=True) meta_files = glob(f"{os.path.join(root_path,'txt')}/**/*.txt", recursive=True)
for meta_file in meta_files: for meta_file in meta_files:
txt, speaker_id, txt_file = os.path.relpath(meta_file,root_path).split(os.sep) _, speaker_id, txt_file = os.path.relpath(meta_file, root_path).split(os.sep)
file_id = txt_file.split('.')[0] file_id = txt_file.split('.')[0]
if isinstance(test_speakers, list): # if is list ignore this speakers ids if isinstance(test_speakers, list): # if is list ignore this speakers ids
if speaker_id in test_speakers: if speaker_id in test_speakers:
continue continue
with open(meta_file) as file_text: with open(meta_file) as file_text:
text = file_text.readlines()[0] text = file_text.readlines()[0]
wav_file = os.path.join(root_path, wavs_path, speaker_id,file_id+'.wav') wav_file = os.path.join(root_path, wavs_path, speaker_id, file_id+'.wav')
items.append([text, wav_file, speaker_id]) items.append([text, wav_file, speaker_id])
return items return items

4
mozilla_voice_tts/tts/models/tacotron2.py
View File

@ -58,7 +58,7 @@ class Tacotron2(TacotronAbstract):
else: else:
# if speaker_embedding_dim is not None we need use speaker embedding per sample # if speaker_embedding_dim is not None we need use speaker embedding per sample
self.embeddings_per_sample = True self.embeddings_per_sample = True
# speaker and gst embeddings is concat in decoder input # speaker and gst embeddings is concat in decoder input
if num_speakers > 1: if num_speakers > 1:
decoder_in_features = decoder_in_features + speaker_embedding_dim # add speaker embedding dim decoder_in_features = decoder_in_features + speaker_embedding_dim # add speaker embedding dim
@ -73,7 +73,7 @@ class Tacotron2(TacotronAbstract):
if not self.embeddings_per_sample: if not self.embeddings_per_sample:
self.speaker_embedding = nn.Embedding(num_speakers, speaker_embedding_dim) self.speaker_embedding = nn.Embedding(num_speakers, speaker_embedding_dim)
self.speaker_embedding.weight.data.normal_(0, 0.3) self.speaker_embedding.weight.data.normal_(0, 0.3)
# base model layers # base model layers
self.encoder = Encoder(encoder_in_features) self.encoder = Encoder(encoder_in_features)
self.decoder = Decoder(decoder_in_features, self.decoder_output_dim, r, attn_type, attn_win, self.decoder = Decoder(decoder_in_features, self.decoder_output_dim, r, attn_type, attn_win,