Sebastian Gehrmann
commited on
Commit
•
112809d
1
Parent(s):
9438a80
make loader work in colab
Browse files- SIMPITIKI.py +27 -27
SIMPITIKI.py
CHANGED
@@ -41,40 +41,40 @@ _LICENSE = "CC-BY 4.0"
|
|
41 |
_URLs = {
|
42 |
"v1":{
|
43 |
"random": {
|
44 |
-
"train":"
|
45 |
-
"val":"
|
46 |
-
"test":"
|
47 |
},
|
48 |
"transformations": {
|
49 |
-
"train": "
|
50 |
-
"val": "
|
51 |
-
"seen_transformations_test": "
|
52 |
-
"unseen_transformations_test":"
|
53 |
},
|
54 |
"source_dataset": {
|
55 |
-
"itwiki_train":"
|
56 |
-
"itwiki_val": "
|
57 |
-
"itwiki_test":"
|
58 |
-
"tn_test":"
|
59 |
}
|
60 |
},
|
61 |
"v2":{
|
62 |
"random": {
|
63 |
-
"train":"
|
64 |
-
"val":"
|
65 |
-
"test":"
|
66 |
},
|
67 |
"transformations": {
|
68 |
-
"train": "
|
69 |
-
"val": "
|
70 |
-
"seen_transformations_test": "
|
71 |
-
"unseen_transformations_test":"
|
72 |
},
|
73 |
"source_dataset": {
|
74 |
-
"itwiki_train":"
|
75 |
-
"itwiki_val": "
|
76 |
-
"itwiki_test":"
|
77 |
-
"tn_test":"
|
78 |
}
|
79 |
|
80 |
|
@@ -144,7 +144,7 @@ class SIMPITIKI(datasets.GeneratorBasedBuilder):
|
|
144 |
# dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLs
|
145 |
# 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.
|
146 |
# By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
|
147 |
-
|
148 |
my_urls = _URLs[self.config.name]
|
149 |
downloaded_files = dl_manager.download_and_extract(my_urls)
|
150 |
return [
|
@@ -182,7 +182,7 @@ class SIMPITIKI(datasets.GeneratorBasedBuilder):
|
|
182 |
},
|
183 |
),
|
184 |
|
185 |
-
|
186 |
datasets.SplitGenerator(
|
187 |
name='challenge_seen_transformations_val',
|
188 |
# These kwargs will be passed to _generate_examples
|
@@ -236,7 +236,7 @@ class SIMPITIKI(datasets.GeneratorBasedBuilder):
|
|
236 |
"split": "challenge_itwiki_test",
|
237 |
},
|
238 |
),
|
239 |
-
|
240 |
datasets.SplitGenerator(
|
241 |
name='challenge_tn_test',
|
242 |
# These kwargs will be passed to _generate_examples
|
@@ -256,7 +256,7 @@ class SIMPITIKI(datasets.GeneratorBasedBuilder):
|
|
256 |
|
257 |
with open(filepath, encoding="utf-8") as f:
|
258 |
for id_, row in enumerate(f):
|
259 |
-
data = json.loads(row)
|
260 |
yield id_, {
|
261 |
"text": data["text"],
|
262 |
"simplified_text": data["simplified_text"],
|
@@ -265,6 +265,6 @@ class SIMPITIKI(datasets.GeneratorBasedBuilder):
|
|
265 |
"gem_id": f"gem-SIMPITIKI-{split}-{id_}",
|
266 |
}
|
267 |
|
268 |
-
|
269 |
|
270 |
|
|
|
41 |
_URLs = {
|
42 |
"v1":{
|
43 |
"random": {
|
44 |
+
"train":"v1/random_split/train.jsonl",
|
45 |
+
"val":"v1/random_split/val.jsonl",
|
46 |
+
"test":"v1/random_split/test.jsonl"
|
47 |
},
|
48 |
"transformations": {
|
49 |
+
"train": "v1/transformations_split/train.jsonl",
|
50 |
+
"val": "v1/transformations_split/val.jsonl",
|
51 |
+
"seen_transformations_test": "v1/transformations_split/seen_transformations_test.jsonl",
|
52 |
+
"unseen_transformations_test":"v1/transformations_split/unseen_transformations_test.jsonl"
|
53 |
},
|
54 |
"source_dataset": {
|
55 |
+
"itwiki_train":"v1/source_dataset_split/itwiki_train.jsonl",
|
56 |
+
"itwiki_val": "v1/source_dataset_split/itwiki_val.jsonl",
|
57 |
+
"itwiki_test":"v1/source_dataset_split/itwiki_test.jsonl",
|
58 |
+
"tn_test":"v1/source_dataset_split/tn_test.jsonl"
|
59 |
}
|
60 |
},
|
61 |
"v2":{
|
62 |
"random": {
|
63 |
+
"train":"v2/random_split/train.jsonl",
|
64 |
+
"val":"v2/random_split/val.jsonl",
|
65 |
+
"test":"v2/random_split/test.jsonl"
|
66 |
},
|
67 |
"transformations": {
|
68 |
+
"train": "v2/transformations_split/train.jsonl",
|
69 |
+
"val": "v2/transformations_split/val.jsonl",
|
70 |
+
"seen_transformations_test": "v2/transformations_split/seen_transformations_test.jsonl",
|
71 |
+
"unseen_transformations_test":"v2/transformations_split/unseen_transformations_test.jsonl"
|
72 |
},
|
73 |
"source_dataset": {
|
74 |
+
"itwiki_train":"v2/source_dataset_split/itwiki_train.jsonl",
|
75 |
+
"itwiki_val": "v2/source_dataset_split/itwiki_val.jsonl",
|
76 |
+
"itwiki_test":"v2/source_dataset_split/itwiki_test.jsonl",
|
77 |
+
"tn_test":"v2/source_dataset_split/tn_test.jsonl"
|
78 |
}
|
79 |
|
80 |
|
|
|
144 |
# dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLs
|
145 |
# 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.
|
146 |
# By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
|
147 |
+
|
148 |
my_urls = _URLs[self.config.name]
|
149 |
downloaded_files = dl_manager.download_and_extract(my_urls)
|
150 |
return [
|
|
|
182 |
},
|
183 |
),
|
184 |
|
185 |
+
|
186 |
datasets.SplitGenerator(
|
187 |
name='challenge_seen_transformations_val',
|
188 |
# These kwargs will be passed to _generate_examples
|
|
|
236 |
"split": "challenge_itwiki_test",
|
237 |
},
|
238 |
),
|
239 |
+
|
240 |
datasets.SplitGenerator(
|
241 |
name='challenge_tn_test',
|
242 |
# These kwargs will be passed to _generate_examples
|
|
|
256 |
|
257 |
with open(filepath, encoding="utf-8") as f:
|
258 |
for id_, row in enumerate(f):
|
259 |
+
data = json.loads(row)
|
260 |
yield id_, {
|
261 |
"text": data["text"],
|
262 |
"simplified_text": data["simplified_text"],
|
|
|
265 |
"gem_id": f"gem-SIMPITIKI-{split}-{id_}",
|
266 |
}
|
267 |
|
268 |
+
|
269 |
|
270 |
|