change the docstring style from numpydoc to google, test=tts

paddlespeech/t2s/datasets/data_table.py

@@ -22,26 +22,17 @@
 
 class DataTable(Dataset):
  """Dataset to load and convert data for general purpose.
-
- Parameters
- ----------
- data : List[Dict[str, Any]]
- Metadata, a list of meta datum, each of which is composed of 
- several fields
- fields : List[str], optional
- Fields to use, if not specified, all the fields in the data are 
- used, by default None
- converters : Dict[str, Callable], optional
- Converters used to process each field, by default None
- use_cache : bool, optional
- Whether to use cache, by default False
-
- Raises
- ------
- ValueError
- If there is some field that does not exist in data. 
- ValueError
- If there is some field in converters that does not exist in fields.
+ Args:
+ data (List[Dict[str, Any]]): Metadata, a list of meta datum, each of which is composed of several fields
+ fields (List[str], optional): Fields to use, if not specified, all the fields in the data are used, by default None
+ converters (Dict[str, Callable], optional): Converters used to process each field, by default None
+ use_cache (bool, optional): Whether to use cache, by default False
+
+ Raises:
+ ValueError:
+ If there is some field that does not exist in data. 
+ ValueError:
+ If there is some field in converters that does not exist in fields.
  """
 
  def __init__(self,
@@ -95,15 +86,11 @@ def _convert(self, meta_datum: Dict[str, Any]) -> Dict[str, Any]:
  """Convert a meta datum to an example by applying the corresponding 
  converters to each fields requested.
 
- Parameters
- ----------
- meta_datum : Dict[str, Any]
- Meta datum
+ Args:
+ meta_datum (Dict[str, Any]): Meta datum
 
- Returns
- -------
- Dict[str, Any]
- Converted example
+ Returns:
+ Dict[str, Any]: Converted example
  """
  example = {}
  for field in self.fields:
@@ -118,16 +105,11 @@ def _convert(self, meta_datum: Dict[str, Any]) -> Dict[str, Any]:
 
  def __getitem__(self, idx: int) -> Dict[str, Any]:
  """Get an example given an index.
+ Args:
+ idx (int): Index of the example to get
 
- Parameters
- ----------
- idx : int
- Index of the example to get
-
- Returns
- -------
- Dict[str, Any]
- A converted example
+ Returns:
+ Dict[str, Any]: A converted example
  """
  if self.use_cache and self.caches[idx] is not None:
  return self.caches[idx]

paddlespeech/t2s/datasets/preprocess_utils.py

@@ -18,14 +18,10 @@
 def get_phn_dur(file_name):
  '''
  read MFA duration.txt
- Parameters
- ----------
- file_name : str or Path
- path of gen_duration_from_textgrid.py's result
- Returns
- ----------
- Dict
- sentence: {'utt': ([char], [int])}
+ Args:
+ file_name (str or Path): path of gen_duration_from_textgrid.py's result
+ Returns: 
+ Dict: sentence: {'utt': ([char], [int])}
  '''
  f = open(file_name, 'r')
  sentence = {}
@@ -48,10 +44,8 @@ def get_phn_dur(file_name):
 def merge_silence(sentence):
  '''
  merge silences
- Parameters
- ----------
- sentence : Dict
- sentence: {'utt': (([char], [int]), str)}
+ Args:
+ sentence (Dict): sentence: {'utt': (([char], [int]), str)}
  '''
  for utt in sentence:
  cur_phn, cur_dur, speaker = sentence[utt]
@@ -81,12 +75,9 @@ def merge_silence(sentence):
 def get_input_token(sentence, output_path, dataset="baker"):
  '''
  get phone set from training data and save it
- Parameters
- ----------
- sentence : Dict
- sentence: {'utt': ([char], [int])}
- output_path : str or path
- path to save phone_id_map
+ Args:
+ sentence (Dict): sentence: {'utt': ([char], [int])}
+ output_path (str or path):path to save phone_id_map
  '''
  phn_token = set()
  for utt in sentence:
@@ -112,14 +103,10 @@ def get_phones_tones(sentence,
  dataset="baker"):
  '''
  get phone set and tone set from training data and save it
- Parameters
- ----------
- sentence : Dict
- sentence: {'utt': ([char], [int])}
- phones_output_path : str or path
- path to save phone_id_map
- tones_output_path : str or path
- path to save tone_id_map
+ Args:
+ sentence (Dict): sentence: {'utt': ([char], [int])}
+ phones_output_path (str or path): path to save phone_id_map
+ tones_output_path (str or path): path to save tone_id_map
  '''
  phn_token = set()
  tone_token = set()
@@ -162,14 +149,10 @@ def get_spk_id_map(speaker_set, output_path):
 def compare_duration_and_mel_length(sentences, utt, mel):
  '''
  check duration error, correct sentences[utt] if possible, else pop sentences[utt]
- Parameters
- ----------
- sentences : Dict
- sentences[utt] = [phones_list ,durations_list]
- utt : str
- utt_id
- mel : np.ndarry
- features (num_frames, n_mels)
+ Args:
+ sentences (Dict): sentences[utt] = [phones_list ,durations_list]
+ utt (str): utt_id
+ mel (np.ndarry): features (num_frames, n_mels)
  '''
 
  if utt in sentences:

paddlespeech/t2s/datasets/vocoder_batch_fn.py

@@ -29,15 +29,11 @@ def __init__(
  hop_size=256,
  aux_context_window=0, ):
  """Initialize customized collater for DataLoader.
+ Args:
 
- Parameters
- ----------
- batch_max_steps : int
- The maximum length of input signal in batch.
- hop_size : int
- Hop size of auxiliary features.
- aux_context_window : int
- Context window size for auxiliary feature conv.
+ batch_max_steps (int): The maximum length of input signal in batch.
+ hop_size (int): Hop size of auxiliary features.
+ aux_context_window (int): Context window size for auxiliary feature conv.
 
  """
  if batch_max_steps % hop_size != 0:
@@ -56,18 +52,15 @@ def __init__(
  def __call__(self, batch):
  """Convert into batch tensors.
 
- Parameters
- ----------
- batch : list
- list of tuple of the pair of audio and features. Audio shape (T, ), features shape(T', C).
+ Args:
+ batch (list): list of tuple of the pair of audio and features. Audio shape (T, ), features shape(T', C).
 
- Returns
- ----------
- Tensor
- Auxiliary feature batch (B, C, T'), where
- T = (T' - 2 * aux_context_window) * hop_size.
- Tensor
- Target signal batch (B, 1, T).
+ Returns: 
+ Tensor:
+ Auxiliary feature batch (B, C, T'), where
+ T = (T' - 2 * aux_context_window) * hop_size.
+ Tensor:
+ Target signal batch (B, 1, T).
 
  """
  # check length
@@ -104,11 +97,10 @@ def __call__(self, batch):
  def _adjust_length(self, x, c):
  """Adjust the audio and feature lengths.
 
- Note
- -------
- Basically we assume that the length of x and c are adjusted
- through preprocessing stage, but if we use other library processed
- features, this process will be needed.
+ Note:
+ Basically we assume that the length of x and c are adjusted
+ through preprocessing stage, but if we use other library processed
+ features, this process will be needed.
 
  """
  if len(x) < c.shape[0] * self.hop_size:
@@ -162,22 +154,14 @@ def __call__(self, batch):
  # voc_pad = 2 this will pad the input so that the resnet can 'see' wider than input length
  # max_offsets = n_frames - 2 - (mel_win + 2 * hp.voc_pad) = n_frames - 15
  """Convert into batch tensors.
-
- Parameters
- ----------
- batch : list
- list of tuple of the pair of audio and features. 
- Audio shape (T, ), features shape(T', C).
-
- Returns
- ----------
- Tensor
- Input signal batch (B, 1, T).
- Tensor
- Target signal batch (B, 1, T).
- Tensor
- Auxiliary feature batch (B, C, T'), where
- T = (T' - 2 * aux_context_window) * hop_size.
+ Args:
+ batch (list): list of tuple of the pair of audio and features. Audio shape (T, ), features shape(T', C).
+
+ Returns:
+ Tensor: Input signal batch (B, 1, T).
+ Tensor: Target signal batch (B, 1, T).
+ Tensor: Auxiliary feature batch (B, C, T'), 
+ where T = (T' - 2 * aux_context_window) * hop_size.
 
  """
  # check length

paddlespeech/t2s/exps/transformer_tts/preprocess.py

@@ -31,15 +31,12 @@
 
 
 def get_lj_sentences(file_name, frontend):
- '''
- read MFA duration.txt
- Parameters
- ----------
- file_name : str or Path
- Returns
- ----------
- Dict
- sentence: {'utt': ([char], [int])}
+ '''read MFA duration.txt
+
+ Args:
+ file_name (str or Path)
+ Returns:
+ Dict: sentence: {'utt': ([char], [int])}
  '''
  f = open(file_name, 'r')
  sentence = {}
@@ -59,14 +56,11 @@ def get_lj_sentences(file_name, frontend):
 
 
 def get_input_token(sentence, output_path):
- '''
- get phone set from training data and save it
- Parameters
- ----------
- sentence : Dict
- sentence: {'utt': ([char], str)}
- output_path : str or path
- path to save phone_id_map
+ '''get phone set from training data and save it
+ 
+ Args:
+ sentence (Dict): sentence: {'utt': ([char], str)}
+ output_path (str or path): path to save phone_id_map
  '''
  phn_token = set()
  for utt in sentence: