aadelucia commited on
Commit
5228c13
1 Parent(s): 00b849e
Files changed (1) hide show
  1. bernice-pretrain-data.py +2 -1
bernice-pretrain-data.py CHANGED
@@ -67,6 +67,7 @@ _URLS = {
67
  "indic": ["indic_tweet_ids.txt.gz"]
68
  }
69
 
 
70
  # TODO: Name of the dataset usually matches the script name with CamelCase instead of snake_case
71
  class BernicePretrainData(datasets.GeneratorBasedBuilder):
72
  """Tweet IDs for the 2.5 billion multilingual tweets used to train Bernice, a Twitter encoder."""
@@ -141,7 +142,7 @@ class BernicePretrainData(datasets.GeneratorBasedBuilder):
141
  # TODO: This method handles input defined in _split_generators to yield (key, example) tuples from the dataset.
142
  # The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
143
  for filepath in filepaths:
144
- with gzip.open(filepath, "rb") as f:
145
  for line_number, instance in enumerate(f):
146
  tweet_id, lang, year = instance.strip().split("\t")
147
  yield tweet_id, {
 
67
  "indic": ["indic_tweet_ids.txt.gz"]
68
  }
69
 
70
+
71
  # TODO: Name of the dataset usually matches the script name with CamelCase instead of snake_case
72
  class BernicePretrainData(datasets.GeneratorBasedBuilder):
73
  """Tweet IDs for the 2.5 billion multilingual tweets used to train Bernice, a Twitter encoder."""
 
142
  # TODO: This method handles input defined in _split_generators to yield (key, example) tuples from the dataset.
143
  # The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
144
  for filepath in filepaths:
145
+ with open(filepath, encoding="utf-8") as f:
146
  for line_number, instance in enumerate(f):
147
  tweet_id, lang, year = instance.strip().split("\t")
148
  yield tweet_id, {