mirror of https://github.com/coqui-ai/TTS.git
modify according to PR reviews
This commit is contained in:
parent
42ba30eb8f
commit
9ab07f94e2
TTS
|
@ -372,10 +372,15 @@ def _voxcel_x(root_path, meta_file, voxcel_idx):
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
def baker(root_path: str, meta_file: str) -> List[List[str]]:
|
||||||
|
"""Normalizes the Baker meta data file to TTS format
|
||||||
|
|
||||||
# ======================================== Baker (chinese mandarin single speaker) ===========================================
|
Args:
|
||||||
def baker(root_path, meta_file):
|
root_path (str): path to the baker dataset
|
||||||
"""Normalizes the Baker meta data file to TTS format"""
|
meta_file (str): name of the meta dataset containing names of wav to select and the transcript of the sentence
|
||||||
|
Returns:
|
||||||
|
List[List[str]]: List of (text, wav_path, speaker_name) associated with each sentences
|
||||||
|
"""
|
||||||
txt_file = os.path.join(root_path, meta_file)
|
txt_file = os.path.join(root_path, meta_file)
|
||||||
items = []
|
items = []
|
||||||
speaker_name = "baker"
|
speaker_name = "baker"
|
||||||
|
|
|
@ -105,7 +105,7 @@ def _num2chinese(num :str, big=False, simp=True, o=False, twoalt=False) -> str:
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
def _number_replace(match) -> str:
|
def _number_replace(match: re.Match) -> str:
|
||||||
"""function to apply in a match, transform all numbers in a match by chinese characters
|
"""function to apply in a match, transform all numbers in a match by chinese characters
|
||||||
|
|
||||||
Args:
|
Args:
|
||||||
|
|
|
@ -220,7 +220,6 @@ def synthesis(model,
|
||||||
model outputs.
|
model outputs.
|
||||||
speaker_id (int): id of speaker
|
speaker_id (int): id of speaker
|
||||||
style_wav (str | Dict[str, float]): Uses for style embedding of GST.
|
style_wav (str | Dict[str, float]): Uses for style embedding of GST.
|
||||||
style_wav (str): Uses for style embedding of GST.
|
|
||||||
truncated (bool): keep model states after inference. It can be used
|
truncated (bool): keep model states after inference. It can be used
|
||||||
for continuous inference at long texts.
|
for continuous inference at long texts.
|
||||||
enable_eos_bos_chars (bool): enable special chars for end of sentence and start of sentence.
|
enable_eos_bos_chars (bool): enable special chars for end of sentence and start of sentence.
|
||||||
|
|
|
@ -127,7 +127,7 @@ class Synthesizer(object):
|
||||||
gst_style_input = None
|
gst_style_input = None
|
||||||
if self.tts_config.use_gst:
|
if self.tts_config.use_gst:
|
||||||
if self.tts_config.gst["gst_style_input"] not in ["", {}]:
|
if self.tts_config.gst["gst_style_input"] not in ["", {}]:
|
||||||
style_wav = self.tts_config.gst["gst_style_input"]
|
gst_style_input = self.tts_config.gst["gst_style_input"]
|
||||||
|
|
||||||
for sen in sens:
|
for sen in sens:
|
||||||
# synthesize voice
|
# synthesize voice
|
||||||
|
|
Loading…
Reference in New Issue