cahya commited on
Commit
cd30a7d
·
1 Parent(s): 5fb8e5c
Files changed (1) hide show
  1. fleurs.py +5 -5
fleurs.py CHANGED
@@ -15,7 +15,7 @@
15
 
16
  import os
17
  from collections import OrderedDict
18
- from text_processor.text_processor import TextProcessor
19
  import datasets
20
 
21
  logger = datasets.logging.get_logger(__name__)
@@ -52,7 +52,7 @@ for langs in _FLEURS_LANG:
52
  _ALL_CONFIGS.append("all")
53
 
54
  # TODO(FLEURS)
55
- _DESCRIPTION = "1.FLEURS is the speech version of the FLORES machine translation benchmark, covering 2000 n-way parallel sentences in n=102 languages."
56
  _CITATION = ""
57
  _HOMEPAGE_URL = ""
58
 
@@ -104,7 +104,7 @@ class Fleurs(datasets.GeneratorBasedBuilder):
104
  # "audio": datasets.Audio(sampling_rate=48_000),
105
  "audio": datasets.features.Audio(sampling_rate=48_000),
106
  "transcription": datasets.Value("string"),
107
- "raw_transcription": datasets.Value("string"),
108
  "lang_id": datasets.ClassLabel(names=langs),
109
  "language": datasets.Value("string"),
110
  "lang_group_id": datasets.ClassLabel(
@@ -185,7 +185,7 @@ class Fleurs(datasets.GeneratorBasedBuilder):
185
  ]
186
 
187
  def _get_data(self, lines, lang_id):
188
- tp = TextProcessor()
189
  data = {}
190
  for line in lines:
191
  if isinstance(line, bytes):
@@ -204,7 +204,7 @@ class Fleurs(datasets.GeneratorBasedBuilder):
204
 
205
  data[file_name] = {
206
  "id": int(_id),
207
- "raw_transcription": raw_transcription,
208
  "transcription": transcription,
209
  "num_samples": int(num_samples),
210
  "lang_id": _FLEURS_LANG.index(lang_id),
 
15
 
16
  import os
17
  from collections import OrderedDict
18
+ # from text_processor.text_processor import TextProcessor
19
  import datasets
20
 
21
  logger = datasets.logging.get_logger(__name__)
 
52
  _ALL_CONFIGS.append("all")
53
 
54
  # TODO(FLEURS)
55
+ _DESCRIPTION = "2.FLEURS is the speech version of the FLORES machine translation benchmark, covering 2000 n-way parallel sentences in n=102 languages."
56
  _CITATION = ""
57
  _HOMEPAGE_URL = ""
58
 
 
104
  # "audio": datasets.Audio(sampling_rate=48_000),
105
  "audio": datasets.features.Audio(sampling_rate=48_000),
106
  "transcription": datasets.Value("string"),
107
+ "sentence": datasets.Value("string"),
108
  "lang_id": datasets.ClassLabel(names=langs),
109
  "language": datasets.Value("string"),
110
  "lang_group_id": datasets.ClassLabel(
 
185
  ]
186
 
187
  def _get_data(self, lines, lang_id):
188
+ # tp = TextProcessor()
189
  data = {}
190
  for line in lines:
191
  if isinstance(line, bytes):
 
204
 
205
  data[file_name] = {
206
  "id": int(_id),
207
+ "sentence": raw_transcription,
208
  "transcription": transcription,
209
  "num_samples": int(num_samples),
210
  "lang_id": _FLEURS_LANG.index(lang_id),