Datasets:

Multilinguality:
multilingual
Size Categories:
100K<n<1M
Annotations Creators:
expert-generated
Source Datasets:
original
ArXiv:
Tags:
License:
anton-l HF staff commited on
Commit
99befc2
1 Parent(s): 2b71d4d

Fix streaming filenames

Browse files
Files changed (1) hide show
  1. multilingual_librispeech.py +4 -3
multilingual_librispeech.py CHANGED
@@ -22,6 +22,7 @@ import os
22
 
23
  import datasets
24
  from datasets.tasks import AutomaticSpeechRecognition
 
25
 
26
 
27
  _CITATION = """\
@@ -222,12 +223,12 @@ def download_audio_archives(dl_manager, root_dir, split):
222
  # each split contains many .tar.gz archives with its audio files
223
  # audio_filenames.txt contains the names of these archives
224
  split_dir = os.path.join(root_dir, split)
225
- audio_filenames_path = dl_manager.download_and_extract(os.path.join(split_dir, "audio_filenames.txt"))
226
 
227
- with open(audio_filenames_path, "r", encoding="utf-8") as file:
228
  audio_filenames = [line.strip() for line in file.readlines()]
229
 
230
  archive_paths = dl_manager.download([os.path.join(split_dir, "audio", filename) for filename in audio_filenames])
231
  audio_archives = [dl_manager.iter_archive(archive_path) for archive_path in archive_paths]
232
 
233
- return audio_archives
 
22
 
23
  import datasets
24
  from datasets.tasks import AutomaticSpeechRecognition
25
+ from datasets.utils.streaming_download_manager import xopen
26
 
27
 
28
  _CITATION = """\
 
223
  # each split contains many .tar.gz archives with its audio files
224
  # audio_filenames.txt contains the names of these archives
225
  split_dir = os.path.join(root_dir, split)
226
+ audio_filenames_path = dl_manager.download(os.path.join(split_dir, "audio_filenames.txt"))
227
 
228
+ with xopen(audio_filenames_path, "r", encoding="utf-8") as file:
229
  audio_filenames = [line.strip() for line in file.readlines()]
230
 
231
  archive_paths = dl_manager.download([os.path.join(split_dir, "audio", filename) for filename in audio_filenames])
232
  audio_archives = [dl_manager.iter_archive(archive_path) for archive_path in archive_paths]
233
 
234
+ return audio_archives