system HF staff commited on
Commit
8d6952d
β€’
0 Parent(s):

Update files from the datasets library (from 1.3.0)

Browse files

Release notes: https://github.com/huggingface/datasets/releases/tag/1.3.0

.gitattributes ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bin.* filter=lfs diff=lfs merge=lfs -text
5
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.model filter=lfs diff=lfs merge=lfs -text
12
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
13
+ *.onnx filter=lfs diff=lfs merge=lfs -text
14
+ *.ot filter=lfs diff=lfs merge=lfs -text
15
+ *.parquet filter=lfs diff=lfs merge=lfs -text
16
+ *.pb filter=lfs diff=lfs merge=lfs -text
17
+ *.pt filter=lfs diff=lfs merge=lfs -text
18
+ *.pth filter=lfs diff=lfs merge=lfs -text
19
+ *.rar filter=lfs diff=lfs merge=lfs -text
20
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
21
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
22
+ *.tflite filter=lfs diff=lfs merge=lfs -text
23
+ *.tgz filter=lfs diff=lfs merge=lfs -text
24
+ *.xz filter=lfs diff=lfs merge=lfs -text
25
+ *.zip filter=lfs diff=lfs merge=lfs -text
26
+ *.zstandard filter=lfs diff=lfs merge=lfs -text
27
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,454 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ annotations_creators: []
3
+ language_creators:
4
+ - found
5
+ languages:
6
+ - en
7
+ licenses:
8
+ - unknown
9
+ multilinguality:
10
+ - monolingual
11
+ size_categories:
12
+ emoji:
13
+ - 100K<n<1M
14
+ emotion:
15
+ - 1K<n<10K
16
+ hate:
17
+ - 10K<n<100K
18
+ irony:
19
+ - 1K<n<10K
20
+ offensive:
21
+ - 10K<n<100K
22
+ sentiment:
23
+ - 10K<n<100K
24
+ stance_abortion:
25
+ - n<1K
26
+ stance_atheism:
27
+ - n<1K
28
+ stance_climate:
29
+ - n<1K
30
+ stance_feminist:
31
+ - n<1K
32
+ stance_hillary:
33
+ - n<1K
34
+ source_datasets:
35
+ emoji:
36
+ - extended|other-tweet-datasets
37
+ emotion:
38
+ - extended|other-tweet-datasets
39
+ hate:
40
+ - extended|other-tweet-datasets
41
+ irony:
42
+ - extended|other-tweet-datasets
43
+ offensive:
44
+ - extended|other-tweet-datasets
45
+ sentiment:
46
+ - extended|other-tweet-datasets
47
+ stance_abortion:
48
+ - extended|other-tweet-datasets
49
+ stance_atheism:
50
+ - extended|other-tweet-datasets
51
+ stance_climate:
52
+ - extended|other-tweet-datasets
53
+ stance_feminist:
54
+ - extended|other-tweet-datasets
55
+ stance_hillary:
56
+ - extended|other-tweet-datasets
57
+ task_categories:
58
+ - text-classification
59
+ task_ids:
60
+ emoji:
61
+ - multi-class-classification
62
+ emotion:
63
+ - multi-class-classification
64
+ - sentiment-classification
65
+ hate:
66
+ - intent-classification
67
+ irony:
68
+ - multi-class-classification
69
+ offensive:
70
+ - intent-classification
71
+ sentiment:
72
+ - multi-class-classification
73
+ - sentiment-classification
74
+ stance_abortion:
75
+ - intent-classification
76
+ - multi-class-classification
77
+ stance_atheism:
78
+ - intent-classification
79
+ - multi-class-classification
80
+ stance_climate:
81
+ - intent-classification
82
+ - multi-class-classification
83
+ stance_feminist:
84
+ - intent-classification
85
+ - multi-class-classification
86
+ stance_hillary:
87
+ - intent-classification
88
+ - multi-class-classification
89
+ ---
90
+
91
+ # Dataset Card for tweet_eval
92
+
93
+ ## Table of Contents
94
+
95
+ - [Dataset Description](#dataset-description)
96
+ - [Dataset Summary](#dataset-summary)
97
+ - [Supported Tasks](#supported-tasks-and-leaderboards)
98
+ - [Languages](#languages)
99
+ - [Dataset Structure](#dataset-structure)
100
+ - [Data Instances](#data-instances)
101
+ - [Data Fields](#data-instances)
102
+ - [Data Splits](#data-instances)
103
+ - [Dataset Creation](#dataset-creation)
104
+ - [Curation Rationale](#curation-rationale)
105
+ - [Source Data](#source-data)
106
+ - [Annotations](#annotations)
107
+ - [Personal and Sensitive Information](#personal-and-sensitive-information)
108
+ - [Considerations for Using the Data](#considerations-for-using-the-data)
109
+ - [Social Impact of Dataset](#social-impact-of-dataset)
110
+ - [Discussion of Biases](#discussion-of-biases)
111
+ - [Other Known Limitations](#other-known-limitations)
112
+ - [Additional Information](#additional-information)
113
+ - [Dataset Curators](#dataset-curators)
114
+ - [Licensing Information](#licensing-information)
115
+ - [Citation Information](#citation-information)
116
+
117
+ ## Dataset Description
118
+
119
+ - **Homepage:** [Needs More Information]
120
+ - **Repository:** [GitHub](https://github.com/cardiffnlp/tweeteval)
121
+ - **Paper:** [EMNLP Paper](https://arxiv.org/pdf/2010.12421.pdf)
122
+ - **Leaderboard:** [GitHub Leaderboard](https://github.com/cardiffnlp/tweeteval)
123
+ - **Point of Contact:** [Needs More Information]
124
+
125
+ ### Dataset Summary
126
+
127
+ TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. The tasks include - irony, hate, offensive, stance, emoji, emotion, and sentiment. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
128
+
129
+ ### Supported Tasks and Leaderboards
130
+
131
+ - `text_classification`: The dataset can be trained using a SentenceClassification model from HuggingFace transformers.
132
+
133
+ ### Languages
134
+
135
+ The text in the dataset is in English, as spoken by Twitter users.
136
+
137
+ ## Dataset Structure
138
+
139
+ ### Data Instances
140
+
141
+ An instance from `emoji` config:
142
+
143
+ ```
144
+ {'label': 12, 'text': 'Sunday afternoon walking through Venice in the sun with @user ️ ️ ️ @ Abbot Kinney, Venice'}
145
+ ```
146
+
147
+ An instance from `emotion` config:
148
+
149
+ ```
150
+ {'label': 2, 'text': "β€œWorry is a down payment on a problem you may never have'. \xa0Joyce Meyer. #motivation #leadership #worry"}
151
+ ```
152
+
153
+ An instance from `hate` config:
154
+
155
+ ```
156
+ {'label': 0, 'text': '@user nice new signage. Are you not concerned by Beatlemania -style hysterical crowds crongregating on you…'}
157
+ ```
158
+
159
+ An instance from `irony` config:
160
+
161
+ ```
162
+ {'label': 1, 'text': 'seeing ppl walking w/ crutches makes me really excited for the next 3 weeks of my life'}
163
+ ```
164
+
165
+ An instance from `offensive` config:
166
+
167
+ ```
168
+ {'label': 0, 'text': '@user Bono... who cares. Soon people will understand that they gain nothing from following a phony celebrity. Become a Leader of your people instead or help and support your fellow countrymen.'}
169
+ ```
170
+
171
+ An instance from `sentiment` config:
172
+
173
+ ```
174
+ {'label': 2, 'text': '"QT @user In the original draft of the 7th book, Remus Lupin survived the Battle of Hogwarts. #HappyBirthdayRemusLupin"'}
175
+ ```
176
+
177
+ An instance from `stance_abortion` config:
178
+
179
+ ```
180
+ {'label': 1, 'text': 'we remind ourselves that love means to be willing to give until it hurts - Mother Teresa'}
181
+ ```
182
+
183
+ An instance from `stance_atheism` config:
184
+
185
+ ```
186
+ {'label': 1, 'text': '@user Bless Almighty God, Almighty Holy Spirit and the Messiah. #SemST'}
187
+ ```
188
+
189
+ An instance from `stance_climate` config:
190
+
191
+ ```
192
+ {'label': 0, 'text': 'Why Is The Pope Upset? via @user #UnzippedTruth #PopeFrancis #SemST'}
193
+ ```
194
+
195
+ An instance from `stance_feminist` config:
196
+
197
+ ```
198
+ {'label': 1, 'text': "@user @user is the UK's answer to @user and @user #GamerGate #SemST"}
199
+ ```
200
+
201
+ An instance from `stance_hillary` config:
202
+
203
+ ```
204
+ {'label': 1, 'text': "If a man demanded staff to get him an ice tea he'd be called a sexists elitist pig.. Oink oink #Hillary #SemST"}
205
+ ```
206
+
207
+ ### Data Fields
208
+ For `emoji` config:
209
+
210
+ - `text`: a `string` feature containing the tweet.
211
+
212
+ - `label`: an `int` classification label with the following mapping:
213
+
214
+ `0`: ❀
215
+
216
+ `1`: 😍
217
+
218
+ `2`: πŸ˜‚
219
+
220
+ `3`: πŸ’•
221
+
222
+ `4`: πŸ”₯
223
+
224
+ `5`: 😊
225
+
226
+ `6`: 😎
227
+
228
+ `7`: ✨
229
+
230
+ `8`: πŸ’™
231
+
232
+ `9`: 😘
233
+
234
+ `10`: πŸ“·
235
+
236
+ `11`: πŸ‡ΊπŸ‡Έ
237
+
238
+ `12`: β˜€
239
+
240
+ `13`: πŸ’œ
241
+
242
+ `14`: πŸ˜‰
243
+
244
+ `15`: πŸ’―
245
+
246
+ `16`: 😁
247
+
248
+ `17`: πŸŽ„
249
+
250
+ `18`: πŸ“Έ
251
+
252
+ `19`: 😜
253
+
254
+ For `emotion` config:
255
+
256
+ - `text`: a `string` feature containing the tweet.
257
+
258
+ - `label`: an `int` classification label with the following mapping:
259
+
260
+ `0`: anger
261
+
262
+ `1`: joy
263
+
264
+ `2`: optimism
265
+
266
+ `3`: sadness
267
+
268
+ For `hate` config:
269
+
270
+ - `text`: a `string` feature containing the tweet.
271
+
272
+ - `label`: an `int` classification label with the following mapping:
273
+
274
+ `0`: non-hate
275
+
276
+ `1`: hate
277
+
278
+ For `irony` config:
279
+
280
+ - `text`: a `string` feature containing the tweet.
281
+
282
+ - `label`: an `int` classification label with the following mapping:
283
+
284
+ `0`: non_irony
285
+
286
+ `1`: irony
287
+
288
+ For `offensive` config:
289
+
290
+ - `text`: a `string` feature containing the tweet.
291
+
292
+ - `label`: an `int` classification label with the following mapping:
293
+
294
+ `0`: non-offensive
295
+
296
+ `1`: offensive
297
+
298
+ For `sentiment` config:
299
+
300
+ - `text`: a `string` feature containing the tweet.
301
+
302
+ - `label`: an `int` classification label with the following mapping:
303
+
304
+ `0`: negative
305
+
306
+ `1`: neutral
307
+
308
+ `2`: positive
309
+
310
+ For `stance_abortion` config:
311
+
312
+ - `text`: a `string` feature containing the tweet.
313
+
314
+ - `label`: an `int` classification label with the following mapping:
315
+
316
+ `0`: none
317
+
318
+ `1`: against
319
+
320
+ `2`: favor
321
+
322
+ For `stance_atheism` config:
323
+
324
+ - `text`: a `string` feature containing the tweet.
325
+
326
+ - `label`: an `int` classification label with the following mapping:
327
+
328
+ `0`: none
329
+
330
+ `1`: against
331
+
332
+ `2`: favor
333
+
334
+ For `stance_climate` config:
335
+
336
+ - `text`: a `string` feature containing the tweet.
337
+
338
+ - `label`: an `int` classification label with the following mapping:
339
+
340
+ `0`: none
341
+
342
+ `1`: against
343
+
344
+ `2`: favor
345
+
346
+ For `stance_feminist` config:
347
+
348
+ - `text`: a `string` feature containing the tweet.
349
+
350
+ - `label`: an `int` classification label with the following mapping:
351
+
352
+ `0`: none
353
+
354
+ `1`: against
355
+
356
+ `2`: favor
357
+
358
+ For `stance_hillary` config:
359
+
360
+ - `text`: a `string` feature containing the tweet.
361
+
362
+ - `label`: an `int` classification label with the following mapping:
363
+
364
+ `0`: none
365
+
366
+ `1`: against
367
+
368
+ `2`: favor
369
+
370
+
371
+
372
+ ### Data Splits
373
+
374
+ | name | train | validation | test |
375
+ | --------------- | ----- | ---------- | ----- |
376
+ | emoji | 45000 | 5000 | 50000 |
377
+ | emotion | 3257 | 374 | 1421 |
378
+ | hate | 9000 | 1000 | 2970 |
379
+ | irony | 2862 | 955 | 784 |
380
+ | offensive | 11916 | 1324 | 860 |
381
+ | sentiment | 45615 | 2000 | 12284 |
382
+ | stance_abortion | 587 | 66 | 280 |
383
+ | stance_atheism | 461 | 52 | 220 |
384
+ | stance_climate | 355 | 40 | 169 |
385
+ | stance_feminist | 597 | 67 | 285 |
386
+ | stance_hillary | 620 | 69 | 295 |
387
+
388
+ ## Dataset Creation
389
+
390
+ ### Curation Rationale
391
+
392
+ [Needs More Information]
393
+
394
+ ### Source Data
395
+
396
+ #### Initial Data Collection and Normalization
397
+
398
+ [Needs More Information]
399
+
400
+ #### Who are the source language producers?
401
+
402
+ [Needs More Information]
403
+
404
+ ### Annotations
405
+
406
+ #### Annotation process
407
+
408
+ [Needs More Information]
409
+
410
+ #### Who are the annotators?
411
+
412
+ [Needs More Information]
413
+
414
+ ### Personal and Sensitive Information
415
+
416
+ [Needs More Information]
417
+
418
+ ## Considerations for Using the Data
419
+
420
+ ### Social Impact of Dataset
421
+
422
+ [Needs More Information]
423
+
424
+ ### Discussion of Biases
425
+
426
+ [Needs More Information]
427
+
428
+ ### Other Known Limitations
429
+
430
+ [Needs More Information]
431
+
432
+ ## Additional Information
433
+
434
+ ### Dataset Curators
435
+
436
+ Francesco Barbieri, Jose Camacho-Collados, Luis Espiinosa-Anke and Leonardo Neves through Cardiff NLP.
437
+
438
+ ### Licensing Information
439
+
440
+ [Needs More Information]
441
+ ### Citation Information
442
+
443
+ ```
444
+ @inproceedings{barbieri2020tweeteval,
445
+ title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},
446
+ author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},
447
+ booktitle={Proceedings of Findings of EMNLP},
448
+ year={2020}
449
+ }
450
+ ```
451
+
452
+ ### Contributions
453
+
454
+ Thanks to [@gchhablani](https://github.com/gchhablani) and [@abhishekkrthakur](https://github.com/abhishekkrthakur) for adding this dataset.
dataset_infos.json ADDED
@@ -0,0 +1 @@
 
1
+ {"emoji": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 20, "names": ["\u2764", "\ud83d\ude0d", "\ud83d\ude02", "\ud83d\udc95", "\ud83d\udd25", "\ud83d\ude0a", "\ud83d\ude0e", "\u2728", "\ud83d\udc99", "\ud83d\ude18", "\ud83d\udcf7", "\ud83c\uddfa\ud83c\uddf8", "\u2600", "\ud83d\udc9c", "\ud83d\ude09", "\ud83d\udcaf", "\ud83d\ude01", "\ud83c\udf84", "\ud83d\udcf8", "\ud83d\ude1c"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "emoji", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3803187, "num_examples": 45000, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 4255921, "num_examples": 50000, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 396083, "num_examples": 5000, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emoji/train_text.txt": {"num_bytes": 3353167, "checksum": "eacb6b0ee1fe2803d72a009c2e731fe07659f604318a979951d2f07c23c564a1"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emoji/train_labels.txt": {"num_bytes": 102760, "checksum": "daee7da826683dbfa50ad3a29c60bc527e498f06c70eabee3745a99cc37ab3a5"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emoji/test_text.txt": {"num_bytes": 3705901, "checksum": "e4de11de1597842c431dd67868e83322f5a432564dfd8558889ed8ac6a1a5e09"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emoji/test_labels.txt": {"num_bytes": 114435, "checksum": "c1662b84788f36674ab8f0106f3e2e7d3e258ddf4959086ac7cc75b1e68dd1f6"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emoji/val_text.txt": {"num_bytes": 341079, "checksum": "3bc3742d6af404cea792671878684d110f3bc02fd79a2e34643789a521d81a26"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emoji/val_labels.txt": {"num_bytes": 11379, "checksum": "21ba456f688668d049ff0fb1fa04469ee684cf4e2467c71d2c3fe5ca2ba1bd1a"}}, "download_size": 7628721, "post_processing_size": null, "dataset_size": 8455191, "size_in_bytes": 16083912}, "emotion": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 4, "names": ["anger", "joy", "optimism", "sadness"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "emotion", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 338875, "num_examples": 3257, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 146649, "num_examples": 1421, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 38277, "num_examples": 374, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emotion/train_text.txt": {"num_bytes": 306630, "checksum": "2c62f67aeb3eac1aea0e5a9c3d0f4bc337992581f3f858061786a1fb4d79d95e"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emotion/train_labels.txt": {"num_bytes": 6514, "checksum": "987e767d8679e18abdf7de37a6d2bcd0a40a296ddd704e8d515cf0e3033c8d9c"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emotion/test_text.txt": {"num_bytes": 132523, "checksum": "7e1070f5d3e3fcece5bc73680bff9981e90d8f7b2f1009bfe7a01d059d1c6091"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emotion/test_labels.txt": {"num_bytes": 2842, "checksum": "245072348c711961785be6d395997f97cf7fcda3effeae7805664171dc75f913"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emotion/val_text.txt": {"num_bytes": 34556, "checksum": "e2e30c86b8cbb97944d6543aedc06eace3bb275cb2f381aba787b838b4f23ca5"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/emotion/val_labels.txt": {"num_bytes": 748, "checksum": "313730630160b7e0a6b4235b800c76683f4aeeb72d094eb69646630cd5cfe338"}}, "download_size": 483813, "post_processing_size": null, "dataset_size": 523801, "size_in_bytes": 1007614}, "hate": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 2, "names": ["non-hate", "hate"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "hate", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1223654, "num_examples": 9000, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 428938, "num_examples": 2970, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 154148, "num_examples": 1000, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/hate/train_text.txt": {"num_bytes": 1133852, "checksum": "6572bb3a42143128a5dfa99af8debeb0668e637c34b2d1e3140dac47316fe2c2"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/hate/train_labels.txt": {"num_bytes": 18000, "checksum": "4e8fde025a453a25c94632794254131dedeac4e57228ad64157c41571cc88f71"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/hate/test_text.txt": {"num_bytes": 399242, "checksum": "bc4762876a8dd8baa55c3cd7b03108e3231a5d691e80b8b1ef97c5be31b9da9a"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/hate/test_labels.txt": {"num_bytes": 5940, "checksum": "c14adca6b3627616a835c5ccea8a1cceb0235cd79417257f093eb0e16a69c62f"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/hate/val_text.txt": {"num_bytes": 144174, "checksum": "1ff78b1ed4c5ce43284b9eba32eb7d60c6d45d0d1b3b4d6df456ae01640764f1"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/hate/val_labels.txt": {"num_bytes": 2000, "checksum": "5092badf1a0e70036ea6264bcd0b78afc07d0f4a512fa6af34c2c4973600656b"}}, "download_size": 1703208, "post_processing_size": null, "dataset_size": 1806740, "size_in_bytes": 3509948}, "irony": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 2, "names": ["non_irony", "irony"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "irony", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 259191, "num_examples": 2862, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 75901, "num_examples": 784, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 86021, "num_examples": 955, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/irony/train_text.txt": {"num_bytes": 231594, "checksum": "a888125a44f7dfaa25b026318748d0e62cc9a300d20f66eafd62011a19eaea23"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/irony/train_labels.txt": {"num_bytes": 5724, "checksum": "fc69e6106c0f1f433a91536e08f83c71a391d7b219f7684d42f243a8089af77d"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/irony/test_text.txt": {"num_bytes": 68057, "checksum": "53103da934a7308eee82f05f2a9781a8ea3e88604fdc1e02d3101108505c64be"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/irony/test_labels.txt": {"num_bytes": 1568, "checksum": "08e2095e1725e74907a380614c220204e356bb46e3e8c93deb74e83e5b15ab38"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/irony/val_text.txt": {"num_bytes": 76760, "checksum": "8806cf3793e300a485cfae34892fc3a0a2f9a183deb06c750c6531515c83051e"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/irony/val_labels.txt": {"num_bytes": 1910, "checksum": "ccf429f63b4e8d0e7f425ca09445f7c31f7cea8a1b7c283b015b117c4002fd07"}}, "download_size": 385613, "post_processing_size": null, "dataset_size": 421113, "size_in_bytes": 806726}, "offensive": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 2, "names": ["non-offensive", "offensive"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "offensive", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1648069, "num_examples": 11916, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 135477, "num_examples": 860, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 192421, "num_examples": 1324, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/offensive/train_text.txt": {"num_bytes": 1529074, "checksum": "78a7a32e38b10af7d8970b008bf17f661c8d0a90dad145fa0fa6a944669650db"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/offensive/train_labels.txt": {"num_bytes": 23832, "checksum": "c0b7d6ebdaa4ebcf6fc557ef1e775d92eda160218a0e3b1dd48eb8234dc892a6"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/offensive/test_text.txt": {"num_bytes": 126921, "checksum": "25b08c3333c26190f1023961c4508ec9aab24d4722b1a3ea7a6040724c120547"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/offensive/test_labels.txt": {"num_bytes": 1720, "checksum": "41d05a7aa0b01f5dafab21b95adb4f979cb4226c046ff315702774d10dac1605"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/offensive/val_text.txt": {"num_bytes": 179188, "checksum": "816f36d180c35f15a5104838cb73856a0bef42043482fe738f3481b06242a55c"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/offensive/val_labels.txt": {"num_bytes": 2648, "checksum": "ed2deb776bd1c52fb8221fadd3360e32d9dfe46842d78053528126e46363a258"}}, "download_size": 1863383, "post_processing_size": null, "dataset_size": 1975967, "size_in_bytes": 3839350}, "sentiment": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 3, "names": ["negative", "neutral", "positive"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "sentiment", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 5425142, "num_examples": 45615, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 1279548, "num_examples": 12284, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 239088, "num_examples": 2000, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/sentiment/train_text.txt": {"num_bytes": 4970029, "checksum": "368f01052ea6fd8ffc408a2a2e6ac9669e31542581a0396ef16591ea26eb98a6"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/sentiment/train_labels.txt": {"num_bytes": 91230, "checksum": "122bfb1732fb6995b0e5c5f726c0ba457c469c3b6e60513007ce5037f23e65d4"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/sentiment/test_text.txt": {"num_bytes": 1156877, "checksum": "09a93a55c63fd93f97485ef7302889d7edb4091cd49733aa37da094f0bfa0675"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/sentiment/test_labels.txt": {"num_bytes": 24568, "checksum": "6afb4afe9374d1f983bcf9a7c79b108d0f37fdf020a83f30488309bed215db9d"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/sentiment/val_text.txt": {"num_bytes": 219137, "checksum": "e5b021e6fc45064c260b09814b803d8f56cada519c4d952d72f43d48a350a964"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/sentiment/val_labels.txt": {"num_bytes": 4000, "checksum": "b4566926c72e2e4e2916c864def94e76c4cdde52446af2c7ba4fc2006e057e51"}}, "download_size": 6465841, "post_processing_size": null, "dataset_size": 6943778, "size_in_bytes": 13409619}, "stance_abortion": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 3, "names": ["none", "against", "favor"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "stance_abortion", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 68698, "num_examples": 587, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 33175, "num_examples": 280, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 7661, "num_examples": 66, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/abortion/train_text.txt": {"num_bytes": 62828, "checksum": "a421d5b8fd9f972970b9275b83f65745bf81986d2a412b4caa2ba071f3efa916"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/abortion/train_labels.txt": {"num_bytes": 1174, "checksum": "e6786a594bd9a083c524a0f420c690351140b52af288f487cb4772d29675b014"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/abortion/test_text.txt": {"num_bytes": 30371, "checksum": "bf0e16a0b8ca4cf0ab90efbc560db3151c288fc842f5e3c6554e8589d521556a"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/abortion/test_labels.txt": {"num_bytes": 560, "checksum": "c90e6d36d863f876d6661620d37b613b4b07858a5277c8d6623713ee59ca451c"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/abortion/val_text.txt": {"num_bytes": 6997, "checksum": "0428ab3f2894936f2445a9020763c2bd19ed42986872168bb65886dede5843fd"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/abortion/val_labels.txt": {"num_bytes": 132, "checksum": "8df57a50823d5f3683ecf75d824a42e3b08eb52e25e3e2d6928f523097a0c050"}}, "download_size": 102062, "post_processing_size": null, "dataset_size": 109534, "size_in_bytes": 211596}, "stance_atheism": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 3, "names": ["none", "against", "favor"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "stance_atheism", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 54779, "num_examples": 461, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 25720, "num_examples": 220, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 6324, "num_examples": 52, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/atheism/train_text.txt": {"num_bytes": 50165, "checksum": "0e82f1d4a16d79a38a68aee761762cf8a846bc8f7f9395670ca44e2ecf2f58f7"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/atheism/train_labels.txt": {"num_bytes": 922, "checksum": "a764aac1a75ccb32c4ffc4c03c66dc365cb50f013d3e94549bf775636cbc8373"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/atheism/test_text.txt": {"num_bytes": 23516, "checksum": "16c5336b2cba606ca63a6afcc50241be63a8fccf021628c6505449439b9d54b3"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/atheism/test_labels.txt": {"num_bytes": 440, "checksum": "4ef7c9398d265cfac625092c834e43cef9da9cb318e563493abb64f65dfe1b52"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/atheism/val_text.txt": {"num_bytes": 5800, "checksum": "5fe14c4c01f87a45dba640dddbb1d1909a893f9565f159c48fa1ba35bb46c209"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/atheism/val_labels.txt": {"num_bytes": 104, "checksum": "638095b3582f927fd1481cdb8d1f9f670f8d27880baf32c0b26c5946fd8f8292"}}, "download_size": 80947, "post_processing_size": null, "dataset_size": 86823, "size_in_bytes": 167770}, "stance_climate": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 3, "names": ["none", "against", "favor"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "stance_climate", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 40253, "num_examples": 355, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 19929, "num_examples": 169, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 4805, "num_examples": 40, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/climate/train_text.txt": {"num_bytes": 36699, "checksum": "4803211832d318026323a8e5014cff1b95e1c8c3854378101e5d1a8c82582eb7"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/climate/train_labels.txt": {"num_bytes": 710, "checksum": "d6274f55bc95f5a7f2ae591b886c1414a7664aaf4e0c609f4ba6cf377929af18"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/climate/test_text.txt": {"num_bytes": 18235, "checksum": "41ee8ee2ad3c36e0629654fdb271f37775197c79be8b299adbeadd2003b63c53"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/climate/test_labels.txt": {"num_bytes": 338, "checksum": "193c9f2358f61d9efe558324ec89ecaf08e600a44b68128f47838c01d9f98dfd"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/climate/val_text.txt": {"num_bytes": 4401, "checksum": "fc5714703add266801ee2fd98296ea20ec0879e89cdb9f906d9812d9f640f2ba"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/climate/val_labels.txt": {"num_bytes": 80, "checksum": "0cb133ab9b137292f075210db45f7e293dc52798a4e21e59037bfcfe66c97aa6"}}, "download_size": 60463, "post_processing_size": null, "dataset_size": 64987, "size_in_bytes": 125450}, "stance_feminist": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 3, "names": ["none", "against", "favor"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "stance_feminist", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 70513, "num_examples": 597, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 33309, "num_examples": 285, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 8039, "num_examples": 67, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/feminist/train_text.txt": {"num_bytes": 64539, "checksum": "c176e6663973c8e78bfa92ba1e8874a70cc5358567d71584a90943bc6525eaab"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/feminist/train_labels.txt": {"num_bytes": 1194, "checksum": "abd4f196d801423bb0daba8c0ecf5b3efba1f10e8f410c3dfa360b50c8b9c685"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/feminist/test_text.txt": {"num_bytes": 30455, "checksum": "1bfdbdc2af64fd62dcc775d1288e192ac8ff805ef27ccf3aaac54a98616eefda"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/feminist/test_labels.txt": {"num_bytes": 570, "checksum": "ddbde6d253ee47c5d5ef8bc5386270fde45cf088d3be70bba9c382b8a024897a"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/feminist/val_text.txt": {"num_bytes": 7365, "checksum": "3518b2ddcf696626a7243d7cea720a975718c7a52a5a086931be87897c1de58b"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/feminist/val_labels.txt": {"num_bytes": 134, "checksum": "399e0d468d0e4ead7a445f69efdf35876c835acf4cefc00a16f451a5d42e5c13"}}, "download_size": 104257, "post_processing_size": null, "dataset_size": 111861, "size_in_bytes": 216118}, "stance_hillary": {"description": "TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.\n", "citation": "@inproceedings{barbieri2020tweeteval,\n title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},\n author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},\n booktitle={Proceedings of Findings of EMNLP},\n year={2020}\n}\n", "homepage": "https://github.com/cardiffnlp/tweeteval", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 3, "names": ["none", "against", "favor"], "names_file": null, "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "builder_name": " tweet_eval", "config_name": "stance_hillary", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 69600, "num_examples": 620, "dataset_name": " tweet_eval"}, "test": {"name": "test", "num_bytes": 34491, "num_examples": 295, "dataset_name": " tweet_eval"}, "validation": {"name": "validation", "num_bytes": 7536, "num_examples": 69, "dataset_name": " tweet_eval"}}, "download_checksums": {"https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/hillary/train_text.txt": {"num_bytes": 63398, "checksum": "0bd735de895cb74d63c224e64e3d955cac99be97aa225f803fe4d2f5978a2c99"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/hillary/train_labels.txt": {"num_bytes": 1240, "checksum": "0ea5753d13a717a9e91581d1d89c0b5206c8f905f0a717b2b27d02dbf419250d"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/hillary/test_text.txt": {"num_bytes": 31537, "checksum": "5c4e020285a62cfd88f264849e1db242ded356c171b1a68dd0050b76635053aa"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/hillary/test_labels.txt": {"num_bytes": 590, "checksum": "068468f6a72b85dfb65bf10e45f2453fa082d1ea9d7a40e7f560d5b6d75027f3"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/hillary/val_text.txt": {"num_bytes": 6842, "checksum": "9714b7dcc8617e095433d7b63df8aa155eb84216b9ac9195105ab83d85cd248d"}, "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/stance/hillary/val_labels.txt": {"num_bytes": 138, "checksum": "e5d44c771b7349a4a74309f56ca072fdf8f1c015068d519ca2ed3a931c833606"}}, "download_size": 103745, "post_processing_size": null, "dataset_size": 111627, "size_in_bytes": 215372}}
dummy/emoji/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02eb645424bd72fd3861b0d33dc615fb59a4589f3012077cacbc2ad9a72e4b1e
3
+ size 1649
dummy/emotion/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:768f0fcecb7049f4177436c0d658fda6ed8aeefd111f09bc9484a1de14ffc4d6
3
+ size 1919
dummy/hate/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f919b81149903e2cfd5ce1a35334c4f79c0f7faa04c3c553d8ad8612dd0ec2ca
3
+ size 2060
dummy/irony/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fe0442f2a443fd59ac5616c595fc097ea903cc3c12eb9a2b76ba9bc12046c2c
3
+ size 1757
dummy/offensive/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc349976f11f1817c0571de0227e9122838794952a991dc2cc063e54d423cba5
3
+ size 2233
dummy/sentiment/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22183ce313f65c9c0bb729342db484265daef7f560cc51469316465619ba347e
3
+ size 1941
dummy/stance_abortion/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f2166e5b997c35ebc0b4480d5568fcf0022bb94270d7e10302e17cc354040df
3
+ size 1927
dummy/stance_atheism/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51f174d94c00b71409e7a07db8a145d93725398639193c94b05af89b273c0430
3
+ size 1891
dummy/stance_climate/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5000c4fe274879381bbd31e4c56b987116709f566d8f058d1397f65c31ee85fc
3
+ size 1887
dummy/stance_feminist/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48e5cd270737a6515d8cab65de9c46195dd5d0108efe403b28dfc2c26b720ce9
3
+ size 1904
dummy/stance_hillary/1.1.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80cc08eb14a3d8e81d14c087f960b5423ed4e25b4ef6318603161fe63762cca5
3
+ size 1874
tweet_eval.py ADDED
@@ -0,0 +1,250 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
3
+ #
4
+ # Licensed under the Apache License, Version 2.0 (the "License");
5
+ # you may not use this file except in compliance with the License.
6
+ # You may obtain a copy of the License at
7
+ #
8
+ # http://www.apache.org/licenses/LICENSE-2.0
9
+ #
10
+ # Unless required by applicable law or agreed to in writing, software
11
+ # distributed under the License is distributed on an "AS IS" BASIS,
12
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13
+ # See the License for the specific language governing permissions and
14
+ # limitations under the License.
15
+ """The Tweet Eval Datasets"""
16
+
17
+ from __future__ import absolute_import, division, print_function
18
+
19
+ import datasets
20
+
21
+
22
+ _CITATION = """\
23
+ @inproceedings{barbieri2020tweeteval,
24
+ title={{TweetEval:Unified Benchmark and Comparative Evaluation for Tweet Classification}},
25
+ author={Barbieri, Francesco and Camacho-Collados, Jose and Espinosa-Anke, Luis and Neves, Leonardo},
26
+ booktitle={Proceedings of Findings of EMNLP},
27
+ year={2020}
28
+ }
29
+ """
30
+
31
+ _DESCRIPTION = """\
32
+ TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
33
+ """
34
+
35
+ _HOMEPAGE = "https://github.com/cardiffnlp/tweeteval"
36
+
37
+ _LICENSE = ""
38
+
39
+ URL = "https://raw.githubusercontent.com/cardiffnlp/tweeteval/main/datasets/"
40
+
41
+ _URLs = {
42
+ "emoji": {
43
+ "train_text": URL + "emoji/train_text.txt",
44
+ "train_labels": URL + "emoji/train_labels.txt",
45
+ "test_text": URL + "emoji/test_text.txt",
46
+ "test_labels": URL + "emoji/test_labels.txt",
47
+ "val_text": URL + "emoji/val_text.txt",
48
+ "val_labels": URL + "emoji/val_labels.txt",
49
+ },
50
+ "emotion": {
51
+ "train_text": URL + "emotion/train_text.txt",
52
+ "train_labels": URL + "emotion/train_labels.txt",
53
+ "test_text": URL + "emotion/test_text.txt",
54
+ "test_labels": URL + "emotion/test_labels.txt",
55
+ "val_text": URL + "emotion/val_text.txt",
56
+ "val_labels": URL + "emotion/val_labels.txt",
57
+ },
58
+ "hate": {
59
+ "train_text": URL + "hate/train_text.txt",
60
+ "train_labels": URL + "hate/train_labels.txt",
61
+ "test_text": URL + "hate/test_text.txt",
62
+ "test_labels": URL + "hate/test_labels.txt",
63
+ "val_text": URL + "hate/val_text.txt",
64
+ "val_labels": URL + "hate/val_labels.txt",
65
+ },
66
+ "irony": {
67
+ "train_text": URL + "irony/train_text.txt",
68
+ "train_labels": URL + "irony/train_labels.txt",
69
+ "test_text": URL + "irony/test_text.txt",
70
+ "test_labels": URL + "irony/test_labels.txt",
71
+ "val_text": URL + "irony/val_text.txt",
72
+ "val_labels": URL + "irony/val_labels.txt",
73
+ },
74
+ "offensive": {
75
+ "train_text": URL + "offensive/train_text.txt",
76
+ "train_labels": URL + "offensive/train_labels.txt",
77
+ "test_text": URL + "offensive/test_text.txt",
78
+ "test_labels": URL + "offensive/test_labels.txt",
79
+ "val_text": URL + "offensive/val_text.txt",
80
+ "val_labels": URL + "offensive/val_labels.txt",
81
+ },
82
+ "sentiment": {
83
+ "train_text": URL + "sentiment/train_text.txt",
84
+ "train_labels": URL + "sentiment/train_labels.txt",
85
+ "test_text": URL + "sentiment/test_text.txt",
86
+ "test_labels": URL + "sentiment/test_labels.txt",
87
+ "val_text": URL + "sentiment/val_text.txt",
88
+ "val_labels": URL + "sentiment/val_labels.txt",
89
+ },
90
+ "stance": {
91
+ "abortion": {
92
+ "train_text": URL + "stance/abortion/train_text.txt",
93
+ "train_labels": URL + "stance/abortion/train_labels.txt",
94
+ "test_text": URL + "stance/abortion/test_text.txt",
95
+ "test_labels": URL + "stance/abortion/test_labels.txt",
96
+ "val_text": URL + "stance/abortion/val_text.txt",
97
+ "val_labels": URL + "stance/abortion/val_labels.txt",
98
+ },
99
+ "atheism": {
100
+ "train_text": URL + "stance/atheism/train_text.txt",
101
+ "train_labels": URL + "stance/atheism/train_labels.txt",
102
+ "test_text": URL + "stance/atheism/test_text.txt",
103
+ "test_labels": URL + "stance/atheism/test_labels.txt",
104
+ "val_text": URL + "stance/atheism/val_text.txt",
105
+ "val_labels": URL + "stance/atheism/val_labels.txt",
106
+ },
107
+ "climate": {
108
+ "train_text": URL + "stance/climate/train_text.txt",
109
+ "train_labels": URL + "stance/climate/train_labels.txt",
110
+ "test_text": URL + "stance/climate/test_text.txt",
111
+ "test_labels": URL + "stance/climate/test_labels.txt",
112
+ "val_text": URL + "stance/climate/val_text.txt",
113
+ "val_labels": URL + "stance/climate/val_labels.txt",
114
+ },
115
+ "feminist": {
116
+ "train_text": URL + "stance/feminist/train_text.txt",
117
+ "train_labels": URL + "stance/feminist/train_labels.txt",
118
+ "test_text": URL + "stance/feminist/test_text.txt",
119
+ "test_labels": URL + "stance/feminist/test_labels.txt",
120
+ "val_text": URL + "stance/feminist/val_text.txt",
121
+ "val_labels": URL + "stance/feminist/val_labels.txt",
122
+ },
123
+ "hillary": {
124
+ "train_text": URL + "stance/hillary/train_text.txt",
125
+ "train_labels": URL + "stance/hillary/train_labels.txt",
126
+ "test_text": URL + "stance/hillary/test_text.txt",
127
+ "test_labels": URL + "stance/hillary/test_labels.txt",
128
+ "val_text": URL + "stance/hillary/val_text.txt",
129
+ "val_labels": URL + "stance/hillary/val_labels.txt",
130
+ },
131
+ },
132
+ }
133
+
134
+
135
+ class TweetEvalConfig(datasets.BuilderConfig):
136
+ def __init__(self, *args, type=None, sub_type=None, **kwargs):
137
+ super().__init__(
138
+ *args,
139
+ name=f"{type}" if type != "stance" else f"{type}_{sub_type}",
140
+ **kwargs,
141
+ )
142
+ self.type = type
143
+ self.sub_type = sub_type
144
+
145
+
146
+ class TweetEval(datasets.GeneratorBasedBuilder):
147
+ """TweetEval Dataset."""
148
+
149
+ BUILDER_CONFIGS = [
150
+ TweetEvalConfig(
151
+ type=key,
152
+ sub_type=None,
153
+ version=datasets.Version("1.1.0"),
154
+ description=f"This part of my dataset covers {key} part of TweetEval Dataset.",
155
+ )
156
+ for key in list(_URLs.keys())
157
+ if key != "stance"
158
+ ] + [
159
+ TweetEvalConfig(
160
+ type="stance",
161
+ sub_type=key,
162
+ version=datasets.Version("1.1.0"),
163
+ description=f"This part of my dataset covers stance_{key} part of TweetEval Dataset.",
164
+ )
165
+ for key in list(_URLs["stance"].keys())
166
+ ]
167
+
168
+ def _info(self):
169
+ if self.config.type == "stance":
170
+ names = ["none", "against", "favor"]
171
+ elif self.config.type == "sentiment":
172
+ names = ["negative", "neutral", "positive"]
173
+ elif self.config.type == "offensive":
174
+ names = ["non-offensive", "offensive"]
175
+ elif self.config.type == "irony":
176
+ names = ["non_irony", "irony"]
177
+ elif self.config.type == "hate":
178
+ names = ["non-hate", "hate"]
179
+ elif self.config.type == "emoji":
180
+ names = [
181
+ "❀",
182
+ "😍",
183
+ "πŸ˜‚",
184
+ "πŸ’•",
185
+ "πŸ”₯",
186
+ "😊",
187
+ "😎",
188
+ "✨",
189
+ "πŸ’™",
190
+ "😘",
191
+ "πŸ“·",
192
+ "πŸ‡ΊπŸ‡Έ",
193
+ "β˜€",
194
+ "πŸ’œ",
195
+ "πŸ˜‰",
196
+ "πŸ’―",
197
+ "😁",
198
+ "πŸŽ„",
199
+ "πŸ“Έ",
200
+ "😜",
201
+ ]
202
+
203
+ else:
204
+ names = ["anger", "joy", "optimism", "sadness"]
205
+
206
+ return datasets.DatasetInfo(
207
+ description=_DESCRIPTION,
208
+ features=datasets.Features(
209
+ {"text": datasets.Value("string"), "label": datasets.features.ClassLabel(names=names)}
210
+ ),
211
+ supervised_keys=None,
212
+ homepage=_HOMEPAGE,
213
+ license=_LICENSE,
214
+ citation=_CITATION,
215
+ )
216
+
217
+ def _split_generators(self, dl_manager):
218
+ """Returns SplitGenerators."""
219
+ if self.config.type != "stance":
220
+ my_urls = _URLs[self.config.type]
221
+ else:
222
+ my_urls = _URLs[self.config.type][self.config.sub_type]
223
+ data_dir = dl_manager.download_and_extract(my_urls)
224
+ return [
225
+ datasets.SplitGenerator(
226
+ name=datasets.Split.TRAIN,
227
+ # These kwargs will be passed to _generate_examples
228
+ gen_kwargs={"text_path": data_dir["train_text"], "labels_path": data_dir["train_labels"]},
229
+ ),
230
+ datasets.SplitGenerator(
231
+ name=datasets.Split.TEST,
232
+ # These kwargs will be passed to _generate_examples
233
+ gen_kwargs={"text_path": data_dir["test_text"], "labels_path": data_dir["test_labels"]},
234
+ ),
235
+ datasets.SplitGenerator(
236
+ name=datasets.Split.VALIDATION,
237
+ # These kwargs will be passed to _generate_examples
238
+ gen_kwargs={"text_path": data_dir["val_text"], "labels_path": data_dir["val_labels"]},
239
+ ),
240
+ ]
241
+
242
+ def _generate_examples(self, text_path, labels_path):
243
+ """ Yields examples. """
244
+
245
+ with open(text_path, encoding="utf-8") as f:
246
+ texts = f.readlines()
247
+ with open(labels_path, encoding="utf-8") as f:
248
+ labels = f.readlines()
249
+ for i, text in enumerate(texts):
250
+ yield i, {"text": text.strip(), "label": int(labels[i].strip())}