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

Refactor TTSDataset to use TTSTokenizer

This commit is contained in:
Eren Gölge 2021-11-16 13:33:21 +01:00
parent 84091096a6
commit b2bb954a51
1 changed files with 17 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
TTS/tts/datasets/dataset.py
View File

@ -69,6 +69,9 @@ class TTSDataset(Dataset):
samples (list): List of dataset samples. samples (list): List of dataset samples.
tokenizer (TTSTokenizer): tokenizer to convert text to sequence IDs. If None init internally else
use the given. Defaults to None.
tokenizer (TTSTokenizer): tokenizer to convert text to sequence IDs. If None init internally else tokenizer (TTSTokenizer): tokenizer to convert text to sequence IDs. If None init internally else
use the given. Defaults to None. use the given. Defaults to None.
@ -202,6 +205,20 @@ class TTSDataset(Dataset):
token_ids = self.tokenizer.text_to_ids(text) token_ids = self.tokenizer.text_to_ids(text)
return np.array(token_ids, dtype=np.int32) return np.array(token_ids, dtype=np.int32)
@staticmethod
def _parse_sample(item):
language_name = None
attn_file = None
if len(item) == 5:
text, wav_file, speaker_name, language_name, attn_file = item
elif len(item) == 4:
text, wav_file, speaker_name, language_name = item
elif len(item) == 3:
text, wav_file, speaker_name = item
else:
raise ValueError(" [!] Dataset cannot parse the sample.")
return text, wav_file, speaker_name, language_name, attn_file
def load_data(self, idx): def load_data(self, idx):
item = self.samples[idx] item = self.samples[idx]