zqwerty
commited on
Commit
·
62f4783
1
Parent(s):
70baea8
add challenge set CM+T
Browse files- CrossWOZ.py +29 -47
CrossWOZ.py
CHANGED
@@ -176,46 +176,38 @@ class CrossWOZ(datasets.GeneratorBasedBuilder):
|
|
176 |
# dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLs
|
177 |
# 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.
|
178 |
# By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
|
179 |
-
|
180 |
return [
|
181 |
-
# datasets.SplitGenerator(
|
182 |
-
# name=datasets.Split.TRAIN,
|
183 |
-
# # These kwargs will be passed to _generate_examples
|
184 |
-
# gen_kwargs={
|
185 |
-
# "filepath": os.path.join(data_dir, "train.json"),
|
186 |
-
# "split": "train",
|
187 |
-
# },
|
188 |
-
# ),
|
189 |
-
# datasets.SplitGenerator(
|
190 |
-
# name=datasets.Split.TEST,
|
191 |
-
# # These kwargs will be passed to _generate_examples
|
192 |
-
# gen_kwargs={
|
193 |
-
# "filepath": os.path.join(data_dir, "test.json"),
|
194 |
-
# "split": "test"
|
195 |
-
# },
|
196 |
-
# ),
|
197 |
-
# datasets.SplitGenerator(
|
198 |
-
# name=datasets.Split.VALIDATION,
|
199 |
-
# # These kwargs will be passed to _generate_examples
|
200 |
-
# gen_kwargs={
|
201 |
-
# "filepath": os.path.join(data_dir, "val.json"),
|
202 |
-
# "split": "dev",
|
203 |
-
# },
|
204 |
-
# ),
|
205 |
-
# datasets.SplitGenerator(
|
206 |
-
# name="challenge.CM+T",
|
207 |
-
# # These kwargs will be passed to _generate_examples
|
208 |
-
# gen_kwargs={
|
209 |
-
# "filepath": os.path.join(data_dir, "test.json"),
|
210 |
-
# "split": "challenge.CM+T",
|
211 |
-
# },
|
212 |
-
# ),
|
213 |
datasets.SplitGenerator(
|
214 |
-
name=
|
215 |
# These kwargs will be passed to _generate_examples
|
216 |
gen_kwargs={
|
217 |
-
"filepath":
|
218 |
-
"split": "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
219 |
},
|
220 |
),
|
221 |
]
|
@@ -271,21 +263,11 @@ class CrossWOZ(datasets.GeneratorBasedBuilder):
|
|
271 |
}
|
272 |
}
|
273 |
|
274 |
-
if split == 'challenge':
|
275 |
-
with open(filepath, encoding="utf-8") as f:
|
276 |
-
for id_, row in enumerate(f):
|
277 |
-
data = json.loads(row)
|
278 |
-
yield id_, {
|
279 |
-
"sentence": data["sentence"],
|
280 |
-
"label": data["label"],
|
281 |
-
"gem_id": data["gem_id"]
|
282 |
-
}
|
283 |
-
|
284 |
key = 0
|
285 |
with open(filepath, encoding="utf-8") as f:
|
286 |
data = json.load(f)
|
287 |
for dialog_id, dialog in data.items():
|
288 |
-
if split == "
|
289 |
continue
|
290 |
messages = []
|
291 |
for turn in dialog["messages"]:
|
|
|
176 |
# dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLs
|
177 |
# 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.
|
178 |
# By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
|
179 |
+
data_dir = dl_manager.download_and_extract("data.zip")
|
180 |
return [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
181 |
datasets.SplitGenerator(
|
182 |
+
name=datasets.Split.TRAIN,
|
183 |
# These kwargs will be passed to _generate_examples
|
184 |
gen_kwargs={
|
185 |
+
"filepath": os.path.join(data_dir, "train.json"),
|
186 |
+
"split": "train",
|
187 |
+
},
|
188 |
+
),
|
189 |
+
datasets.SplitGenerator(
|
190 |
+
name=datasets.Split.TEST,
|
191 |
+
# These kwargs will be passed to _generate_examples
|
192 |
+
gen_kwargs={
|
193 |
+
"filepath": os.path.join(data_dir, "test.json"),
|
194 |
+
"split": "test"
|
195 |
+
},
|
196 |
+
),
|
197 |
+
datasets.SplitGenerator(
|
198 |
+
name=datasets.Split.VALIDATION,
|
199 |
+
# These kwargs will be passed to _generate_examples
|
200 |
+
gen_kwargs={
|
201 |
+
"filepath": os.path.join(data_dir, "val.json"),
|
202 |
+
"split": "dev",
|
203 |
+
},
|
204 |
+
),
|
205 |
+
datasets.SplitGenerator(
|
206 |
+
name="challenge_CM+T",
|
207 |
+
# These kwargs will be passed to _generate_examples
|
208 |
+
gen_kwargs={
|
209 |
+
"filepath": os.path.join(data_dir, "test.json"),
|
210 |
+
"split": "challenge_CM+T",
|
211 |
},
|
212 |
),
|
213 |
]
|
|
|
263 |
}
|
264 |
}
|
265 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
266 |
key = 0
|
267 |
with open(filepath, encoding="utf-8") as f:
|
268 |
data = json.load(f)
|
269 |
for dialog_id, dialog in data.items():
|
270 |
+
if split == "challenge_CM+T" and dialog["type"] != "不独立多领域+交通":
|
271 |
continue
|
272 |
messages = []
|
273 |
for turn in dialog["messages"]:
|