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

Approximate audio length from file size

This commit is contained in:
Eren Gölge 2021-10-18 08:54:02 +00:00
parent b4b890df03
commit a0a5d580e9
4 changed files with 5 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
TTS/tts/configs/tacotron_config.py
View File

@ -106,7 +106,7 @@ class TacotronConfig(BaseTTSConfig):
Weight decay coefficient. Defaults to `1e-6`. Weight decay coefficient. Defaults to `1e-6`.
grad_clip (float): grad_clip (float):
Gradient clipping threshold. Defaults to `5`. Gradient clipping threshold. Defaults to `5`.
seq_len_notm (bool): seq_len_norm (bool):
enable / disable the sequnce length normalization in the loss functions. If set True, loss of a sample enable / disable the sequnce length normalization in the loss functions. If set True, loss of a sample
is divided by the sequence length. Defaults to False. is divided by the sequence length. Defaults to False.
loss_masking (bool): loss_masking (bool):

2
TTS/tts/datasets/dataset.py
View File

@ -330,7 +330,7 @@ class TTSDataset(Dataset):
if by_audio_len: if by_audio_len:
lengths = [] lengths = []
for item in self.items: for item in self.items:
lengths.append(os.path.getsize(item[1])) lengths.append(os.path.getsize(item[1]) / 16 * 8) # assuming 16bit audio
lengths = np.array(lengths) lengths = np.array(lengths)
else: else:
lengths = np.array([len(ins[0]) for ins in self.items]) lengths = np.array([len(ins[0]) for ins in self.items])

2
TTS/tts/models/base_tacotron.py
View File

@ -242,4 +242,4 @@ class BaseTacotron(BaseTTS):
self.decoder.set_r(r) self.decoder.set_r(r)
if trainer.config.bidirectional_decoder: if trainer.config.bidirectional_decoder:
trainer.model.decoder_backward.set_r(r) trainer.model.decoder_backward.set_r(r)
print(f"\n > Number of output frames: {self.decoder.r}") print(f"\n > Number of output frames: {self.decoder.r}")

4
TTS/tts/models/base_tts.py
View File

@ -20,9 +20,9 @@ from TTS.utils.audio import AudioProcessor
class BaseTTS(BaseModel): class BaseTTS(BaseModel):
"""Abstract `tts` class. Every new `tts` model must inherit this. """Base `tts` class. Every new `tts` model must inherit this.
It defines `tts` specific functions on top of `Model`. It defines common `tts` specific functions on top of `Model` implementation.
Notes on input/output tensor shapes: Notes on input/output tensor shapes:
Any input or output tensor of the model must be shaped as Any input or output tensor of the model must be shaped as