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

Comment synthesis.py

This commit is contained in:
Eren Gölge 2021-10-21 13:53:45 +00:00
parent 1987aaaaed
commit aea90e2501
1 changed files with 35 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

48
TTS/tts/utils/synthesis.py
View File

@ -172,7 +172,7 @@ def speaker_id_to_torch(speaker_id, cuda=False):
def embedding_to_torch(d_vector, cuda=False): def embedding_to_torch(d_vector, cuda=False):
if d_vector is not None: if d_vector is not None:
d_vector = np.asarray(d_vector) d_vector = np.asarray(d_vector)
d_vector = torch.from_numpy(d_vector).unsqueeze(0).type(torch.FloatTensor) d_vector = torch.from_numpy(d_vector).type(torch.FloatTensor)
if cuda: if cuda:
return d_vector.cuda() return d_vector.cuda()
return d_vector return d_vector
@ -210,20 +210,42 @@ def synthesis(
d_vector=None, d_vector=None,
backend="torch", backend="torch",
): ):
"""Synthesize voice for the given text. """Synthesize voice for the given text using Griffin-Lim vocoder or just compute output features to be passed to
the vocoder model.
Args: Args:
model (TTS.tts.models): model to synthesize. model (TTS.tts.models):
text (str): target text The TTS model to synthesize audio with.
CONFIG (dict): config dictionary to be loaded from config.json.
use_cuda (bool): enable cuda. text (str):
ap (TTS.tts.utils.audio.AudioProcessor): audio processor to process The input text to convert to speech.
model outputs.
speaker_id (int): id of speaker CONFIG (Coqpit):
style_wav (str | Dict[str, float]): Uses for style embedding of GST. Model configuration.
enable_eos_bos_chars (bool): enable special chars for end of sentence and start of sentence.
do_trim_silence (bool): trim silence after synthesis. use_cuda (bool):
backend (str): tf or torch Enable/disable CUDA.
ap (TTS.tts.utils.audio.AudioProcessor):
The audio processor for extracting features and pre/post-processing audio.
speaker_id (int):
Speaker ID passed to the speaker embedding layer in multi-speaker model. Defaults to None.
style_wav (str | Dict[str, float]):
Path or tensor to/of a waveform used for computing the style embedding. Defaults to None.
enable_eos_bos_chars (bool):
enable special chars for end of sentence and start of sentence. Defaults to False.
do_trim_silence (bool):
trim silence after synthesis. Defaults to False.
d_vector (torch.Tensor):
d-vector for multi-speaker models in share :math:`[1, D]`. Defaults to None.
backend (str):
tf or torch. Defaults to "torch".
""" """
# GST processing # GST processing
style_mel = None style_mel = None