Datasets:
GEM
/

Languages:
Chinese
Multilinguality:
unknown
Size Categories:
unknown
Language Creators:
unknown
Annotations Creators:
none
Source Datasets:
original
License:
zqwerty commited on
Commit
e111ca7
1 Parent(s): a6a5745

use relative path

Browse files
Files changed (1) hide show
  1. CrossWOZ.py +7 -17
CrossWOZ.py CHANGED
@@ -42,9 +42,9 @@ _HOMEPAGE = "https://github.com/thu-coai/CrossWOZ"
42
  _LICENSE = "Apache License, Version 2.0"
43
 
44
  _URLs = {
45
- "train": "https://github.com/thu-coai/CrossWOZ/blob/master/data/crosswoz/train.json.zip",
46
- "val": "https://github.com/thu-coai/CrossWOZ/blob/master/data/crosswoz/val.json.zip",
47
- "test": "https://github.com/thu-coai/CrossWOZ/blob/master/data/crosswoz/test.json.zip"
48
  }
49
 
50
 
@@ -213,19 +213,12 @@ class CrossWOZ(datasets.GeneratorBasedBuilder):
213
  # dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLs
214
  # It can accept any type or nested list/dict and will give back the same structure with the url replaced with path to local files.
215
  # By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
216
- downloaded_files = dl_manager.download_and_extract(_URLs)
217
- print(downloaded_files)
218
- for _, file in downloaded_files.items():
219
- print(_, file)
220
- print(os.path.dirname(file))
221
- print(os.listdir(os.path.dirname(file)))
222
- print(os.listdir(file))
223
  return [
224
  datasets.SplitGenerator(
225
  name=datasets.Split.TRAIN,
226
  # These kwargs will be passed to _generate_examples
227
  gen_kwargs={
228
- "filepath": os.path.join(downloaded_files["train"], "{}.json".format(os.path.basename(downloaded_files["train"]))),
229
  "split": "train",
230
  },
231
  ),
@@ -233,7 +226,7 @@ class CrossWOZ(datasets.GeneratorBasedBuilder):
233
  name=datasets.Split.TEST,
234
  # These kwargs will be passed to _generate_examples
235
  gen_kwargs={
236
- "filepath": os.path.join(downloaded_files["test"], "{}.json".format(os.path.basename(downloaded_files["test"]))),
237
  "split": "test"
238
  },
239
  ),
@@ -241,7 +234,7 @@ class CrossWOZ(datasets.GeneratorBasedBuilder):
241
  name=datasets.Split.VALIDATION,
242
  # These kwargs will be passed to _generate_examples
243
  gen_kwargs={
244
- "filepath": os.path.join(downloaded_files["val"], "{}.json".format(os.path.basename(downloaded_files["val"]))),
245
  "split": "dev",
246
  },
247
  ),
@@ -253,9 +246,6 @@ class CrossWOZ(datasets.GeneratorBasedBuilder):
253
  """ Yields examples as (key, example) tuples. """
254
  # This method handles input defined in _split_generators to yield (key, example) tuples from the dataset.
255
  # The `key` is here for legacy reason (tfds) and is not important in itself.
256
- print(filepath)
257
- print(os.path.dirname(filepath))
258
- print(os.listdir(os.path.dirname(filepath)))
259
  def convert_goal(raw_goal):
260
  goal = []
261
  for subgoal in raw_goal:
@@ -304,4 +294,4 @@ class CrossWOZ(datasets.GeneratorBasedBuilder):
304
  "messages": messages,
305
  "final_goal": convert_goal(dialog["final_goal"])
306
  }
307
- key += 1
 
42
  _LICENSE = "Apache License, Version 2.0"
43
 
44
  _URLs = {
45
+ "train": "train.json.zip",
46
+ "val": "val.json.zip",
47
+ "test": "test.json.zip"
48
  }
49
 
50
 
 
213
  # dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLs
214
  # It can accept any type or nested list/dict and will give back the same structure with the url replaced with path to local files.
215
  # By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
 
 
 
 
 
 
 
216
  return [
217
  datasets.SplitGenerator(
218
  name=datasets.Split.TRAIN,
219
  # These kwargs will be passed to _generate_examples
220
  gen_kwargs={
221
+ "filepath": _URLs["train"],
222
  "split": "train",
223
  },
224
  ),
 
226
  name=datasets.Split.TEST,
227
  # These kwargs will be passed to _generate_examples
228
  gen_kwargs={
229
+ "filepath": _URLs["test"],
230
  "split": "test"
231
  },
232
  ),
 
234
  name=datasets.Split.VALIDATION,
235
  # These kwargs will be passed to _generate_examples
236
  gen_kwargs={
237
+ "filepath": _URLs["val"],
238
  "split": "dev",
239
  },
240
  ),
 
246
  """ Yields examples as (key, example) tuples. """
247
  # This method handles input defined in _split_generators to yield (key, example) tuples from the dataset.
248
  # The `key` is here for legacy reason (tfds) and is not important in itself.
 
 
 
249
  def convert_goal(raw_goal):
250
  goal = []
251
  for subgoal in raw_goal:
 
294
  "messages": messages,
295
  "final_goal": convert_goal(dialog["final_goal"])
296
  }
297
+ key += 1