Datasets:

ArXiv:
License:
NamCyan commited on
Commit
9ff26ff
1 Parent(s): edc339a
Files changed (1) hide show
  1. the-vault-function.py +19 -8
the-vault-function.py CHANGED
@@ -123,13 +123,16 @@ class TheVaultFunction(datasets.GeneratorBasedBuilder):
123
  if "all" in split_set:
124
  split_set = _SPLIT_CONFIGS[1:]
125
 
 
126
  if "train" in split_set:
127
  split_set.remove('train')
128
  split_set.extend(["train/small", "train/medium"])
 
129
 
130
  if "all" in languages:
131
  languages = _LANG_CONFIGS[1:]
132
 
 
133
  for split in split_set:
134
  split_files = []
135
  for language in languages:
@@ -141,14 +144,22 @@ class TheVaultFunction(datasets.GeneratorBasedBuilder):
141
  files = dl_manager.download(data_files)
142
  split_files.extend(files)
143
 
144
- generators.append(
145
- datasets.SplitGenerator(
146
- name=split.replace("/", "_"),
147
- gen_kwargs={
148
- "files": split_files,
149
- },
150
- ),
151
- )
 
 
 
 
 
 
 
 
152
  return generators
153
 
154
  def _generate_examples(self, files):
 
123
  if "all" in split_set:
124
  split_set = _SPLIT_CONFIGS[1:]
125
 
126
+ load_full_train = False
127
  if "train" in split_set:
128
  split_set.remove('train')
129
  split_set.extend(["train/small", "train/medium"])
130
+ load_full_train = True
131
 
132
  if "all" in languages:
133
  languages = _LANG_CONFIGS[1:]
134
 
135
+ train_split_files = []
136
  for split in split_set:
137
  split_files = []
138
  for language in languages:
 
144
  files = dl_manager.download(data_files)
145
  split_files.extend(files)
146
 
147
+ if load_full_train and "train" in split:
148
+ train_split_files.extend(split_files)
149
+ else:
150
+ generators.append(
151
+ datasets.SplitGenerator(
152
+ name=split.replace("/", "_"),
153
+ gen_kwargs={
154
+ "files": split_files,
155
+ },
156
+ ),
157
+ )
158
+
159
+ if load_full_train and train_split_files:
160
+ generators = [datasets.SplitGenerator(name="train", gen_kwargs={"files": train_split_files})] + generators
161
+
162
+
163
  return generators
164
 
165
  def _generate_examples(self, files):