maximedb commited on
Commit
b7831a3
·
1 Parent(s): ff82e4e

Update paws-x-all.py

Browse files
Files changed (1) hide show
  1. paws-x-all.py +7 -7
paws-x-all.py CHANGED
@@ -132,23 +132,23 @@ class PAWSX(datasets.GeneratorBasedBuilder):
132
  _TEST_FILE_NAME = [os.path.join(data_dir, f"x-final/{l}/test_2k.tsv") for l in _LANGUAGES]
133
  _VAL_FILE_NAME = [os.path.join(data_dir, f"x-final/{l}/dev_2k.tsv") for l in _LANGUAGES]
134
  else:
135
- _TEST_FILE_NAME = os.path.join(data_dir, f"x-final/{self.config.name}/test_2k.tsv")
136
- _VAL_FILE_NAME = os.path.join(data_dir, f"x-final/{self.config.name}/dev_2k.tsv")
137
 
138
  if self.config.name == "en":
139
- _TRAIN_FILE_NAME = os.path.join(data_dir, f"x-final/{self.config.name}/train.tsv")
140
  elif self.config.name == "all":
141
  _TRAIN_FILE_NAME = [os.path.join(data_dir, f"x-final/{l}/translated_train.tsv") for l in _LANGUAGES if l != "en"]
142
  _TRAIN_FILE_NAME += [os.path.join(data_dir, f"x-final/en/train.tsv")]
143
  else:
144
- _TRAIN_FILE_NAME = os.path.join(data_dir, f"x-final/{self.config.name}/translated_train.tsv")
145
 
146
  return [
147
  datasets.SplitGenerator(
148
  name=datasets.Split.TRAIN,
149
  # These kwargs will be passed to _generate_examples
150
  gen_kwargs={
151
- "filepath": _TRAIN_FILE_NAME,
152
  "split": datasets.Split.TRAIN,
153
  },
154
  ),
@@ -156,7 +156,7 @@ class PAWSX(datasets.GeneratorBasedBuilder):
156
  name=datasets.Split.TEST,
157
  # These kwargs will be passed to _generate_examples
158
  gen_kwargs={
159
- "filepath": _TEST_FILE_NAME,
160
  "split": datasets.Split.TEST,
161
  },
162
  ),
@@ -164,7 +164,7 @@ class PAWSX(datasets.GeneratorBasedBuilder):
164
  name=datasets.Split.VALIDATION,
165
  # These kwargs will be passed to _generate_examples
166
  gen_kwargs={
167
- "filepath": _VAL_FILE_NAME,
168
  "split": datasets.Split.VALIDATION,
169
  },
170
  ),
 
132
  _TEST_FILE_NAME = [os.path.join(data_dir, f"x-final/{l}/test_2k.tsv") for l in _LANGUAGES]
133
  _VAL_FILE_NAME = [os.path.join(data_dir, f"x-final/{l}/dev_2k.tsv") for l in _LANGUAGES]
134
  else:
135
+ _TEST_FILE_NAME = [os.path.join(data_dir, f"x-final/{self.config.name}/test_2k.tsv")]
136
+ _VAL_FILE_NAME = [os.path.join(data_dir, f"x-final/{self.config.name}/dev_2k.tsv")]
137
 
138
  if self.config.name == "en":
139
+ _TRAIN_FILE_NAME = [os.path.join(data_dir, f"x-final/{self.config.name}/train.tsv")]
140
  elif self.config.name == "all":
141
  _TRAIN_FILE_NAME = [os.path.join(data_dir, f"x-final/{l}/translated_train.tsv") for l in _LANGUAGES if l != "en"]
142
  _TRAIN_FILE_NAME += [os.path.join(data_dir, f"x-final/en/train.tsv")]
143
  else:
144
+ _TRAIN_FILE_NAME = [os.path.join(data_dir, f"x-final/{self.config.name}/translated_train.tsv")]
145
 
146
  return [
147
  datasets.SplitGenerator(
148
  name=datasets.Split.TRAIN,
149
  # These kwargs will be passed to _generate_examples
150
  gen_kwargs={
151
+ "filepaths": _TRAIN_FILE_NAME,
152
  "split": datasets.Split.TRAIN,
153
  },
154
  ),
 
156
  name=datasets.Split.TEST,
157
  # These kwargs will be passed to _generate_examples
158
  gen_kwargs={
159
+ "filepaths": _TEST_FILE_NAME,
160
  "split": datasets.Split.TEST,
161
  },
162
  ),
 
164
  name=datasets.Split.VALIDATION,
165
  # These kwargs will be passed to _generate_examples
166
  gen_kwargs={
167
+ "filepaths": _VAL_FILE_NAME,
168
  "split": datasets.Split.VALIDATION,
169
  },
170
  ),