system HF staff commited on
Commit
ecf512e
1 Parent(s): 71942ac

Update files from the datasets library (from 1.13.3)

Browse files

Release notes: https://github.com/huggingface/datasets/releases/tag/1.13.3

Files changed (2) hide show
  1. README.md +5 -0
  2. common_language.py +2 -0
README.md CHANGED
@@ -124,6 +124,9 @@ A typical data point comprises the `path` to the audio file, and its label `lang
124
  {
125
  'client_id': 'itln_trn_sp_175',
126
  'path': '/path/common_voice_kpd/Italian/train/itln_trn_sp_175/common_voice_it_18279446.wav',
 
 
 
127
  'sentence': 'Con gli studenti è leggermente simile.',
128
  'age': 'not_defined',
129
  'gender': 'not_defined',
@@ -137,6 +140,8 @@ A typical data point comprises the `path` to the audio file, and its label `lang
137
 
138
  `path` (`string`): The path to the audio file
139
 
 
 
140
  `language` (`ClassLabel`): The language of the recording (see the `Languages` section above)
141
 
142
  `sentence` (`string`): The sentence the user was prompted to speak
124
  {
125
  'client_id': 'itln_trn_sp_175',
126
  'path': '/path/common_voice_kpd/Italian/train/itln_trn_sp_175/common_voice_it_18279446.wav',
127
+ 'audio': {'path': '/path/common_voice_kpd/Italian/train/itln_trn_sp_175/common_voice_it_18279446.wav',
128
+ 'array': array([-0.00048828, -0.00018311, -0.00137329, ..., 0.00079346, 0.00091553, 0.00085449], dtype=float32),
129
+ 'sampling_rate': 48000},
130
  'sentence': 'Con gli studenti è leggermente simile.',
131
  'age': 'not_defined',
132
  'gender': 'not_defined',
140
 
141
  `path` (`string`): The path to the audio file
142
 
143
+ - `audio` (`dict`): A dictionary containing the path to the downloaded audio file, the decoded audio array, and the sampling rate. Note that when accessing the audio column: `dataset[0]["audio"]` the audio file is automatically decoded and resampled to `dataset.features["audio"].sampling_rate`. Decoding and resampling of a large number of audio files might take a significant amount of time. Thus it is important to first query the sample index before the `"audio"` column, *i.e.* `dataset[0]["audio"]` should **always** be preferred over `dataset["audio"][0]`.
144
+
145
  `language` (`ClassLabel`): The language of the recording (see the `Languages` section above)
146
 
147
  `sentence` (`string`): The sentence the user was prompted to speak
common_language.py CHANGED
@@ -110,6 +110,7 @@ class CommonLanguage(datasets.GeneratorBasedBuilder):
110
  {
111
  "client_id": datasets.Value("string"),
112
  "path": datasets.Value("string"),
 
113
  "sentence": datasets.Value("string"),
114
  "age": datasets.Value("string"),
115
  "gender": datasets.Value("string"),
@@ -164,6 +165,7 @@ class CommonLanguage(datasets.GeneratorBasedBuilder):
164
  yield key, {
165
  "client_id": client_id,
166
  "path": path,
 
167
  "sentence": sentence,
168
  "age": age,
169
  "gender": gender,
110
  {
111
  "client_id": datasets.Value("string"),
112
  "path": datasets.Value("string"),
113
+ "audio": datasets.features.Audio(sampling_rate=48_000),
114
  "sentence": datasets.Value("string"),
115
  "age": datasets.Value("string"),
116
  "gender": datasets.Value("string"),
165
  yield key, {
166
  "client_id": client_id,
167
  "path": path,
168
+ "audio": path,
169
  "sentence": sentence,
170
  "age": age,
171
  "gender": gender,