Datasets:

Modalities:
Text
Languages:
English
ArXiv:
Libraries:
Datasets
License:
StefanH commited on
Commit
24925b9
1 Parent(s): dc7dfdb

Update: dataset zip file structure; loading script WIP

Browse files
Files changed (6) hide show
  1. .gitignore +1 -0
  2. config.json +1686 -0
  3. datasets.zip +2 -2
  4. datasets_joined.zip +3 -0
  5. dev.py +8 -0
  6. utcd.py +143 -0
.gitignore CHANGED
@@ -1,3 +1,4 @@
1
  .idea/
2
  .vscode/
3
  .DS_Store
 
 
1
  .idea/
2
  .vscode/
3
  .DS_Store
4
+ dev.py
config.json ADDED
@@ -0,0 +1,1686 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "go_emotion": {
3
+ "path": "UTCD/in-domain/go_emotion",
4
+ "aspect": "sentiment",
5
+ "eval_labels_same": true,
6
+ "domain": "in",
7
+ "name": "GoEmotions",
8
+ "name_compact": "GoEmotions",
9
+ "splits": {
10
+ "train": {
11
+ "labels": [
12
+ "admiration",
13
+ "amusement",
14
+ "anger",
15
+ "annoyance",
16
+ "approval",
17
+ "caring",
18
+ "confusion",
19
+ "curiosity",
20
+ "desire",
21
+ "disappointment",
22
+ "disapproval",
23
+ "disgust",
24
+ "embarrassment",
25
+ "excitement",
26
+ "fear",
27
+ "gratitude",
28
+ "grief",
29
+ "joy",
30
+ "love",
31
+ "nervousness",
32
+ "neutral",
33
+ "optimism",
34
+ "pride",
35
+ "realization",
36
+ "relief",
37
+ "remorse",
38
+ "sadness",
39
+ "surprise"
40
+ ],
41
+ "n_label": 28,
42
+ "n_text": 43227,
43
+ "n_pair": 50966,
44
+ "multi_label": true,
45
+ "n_multi_label": 7132
46
+ },
47
+ "test": {
48
+ "labels": [
49
+ "admiration",
50
+ "amusement",
51
+ "anger",
52
+ "annoyance",
53
+ "approval",
54
+ "caring",
55
+ "confusion",
56
+ "curiosity",
57
+ "desire",
58
+ "disappointment",
59
+ "disapproval",
60
+ "disgust",
61
+ "embarrassment",
62
+ "excitement",
63
+ "fear",
64
+ "gratitude",
65
+ "grief",
66
+ "joy",
67
+ "love",
68
+ "nervousness",
69
+ "neutral",
70
+ "optimism",
71
+ "pride",
72
+ "realization",
73
+ "relief",
74
+ "remorse",
75
+ "sadness",
76
+ "surprise"
77
+ ],
78
+ "n_label": 28,
79
+ "n_text": 5421,
80
+ "n_pair": 6324,
81
+ "multi_label": true,
82
+ "n_multi_label": 838
83
+ }
84
+ },
85
+ "re-txt_avg_tokens": 13.351977470810722,
86
+ "bert-txt_avg_tokens": 20.1944170366716,
87
+ "gpt2-txt_avg_tokens": 16.94573260976813,
88
+ "re-lb_avg_tokens": 1.0,
89
+ "bert-lb_avg_tokens": 3.0648455227788443,
90
+ "gpt2-lb_avg_tokens": 1.967551056030721
91
+ },
92
+ "sentiment_tweets_2020": {
93
+ "path": "UTCD/in-domain/sentiment_tweets_2020",
94
+ "aspect": "sentiment",
95
+ "eval_labels_same": true,
96
+ "domain": "in",
97
+ "name": "TweetEval",
98
+ "name_compact": "TweetEval",
99
+ "splits": {
100
+ "train": {
101
+ "labels": [
102
+ "negative",
103
+ "neutral",
104
+ "positive"
105
+ ],
106
+ "n_label": 3,
107
+ "n_text": 45586,
108
+ "n_pair": 45589,
109
+ "multi_label": true,
110
+ "n_multi_label": 3
111
+ },
112
+ "test": {
113
+ "labels": [
114
+ "negative",
115
+ "neutral",
116
+ "positive"
117
+ ],
118
+ "n_label": 3,
119
+ "n_text": 12284,
120
+ "n_pair": 12284,
121
+ "multi_label": false,
122
+ "n_multi_label": 0
123
+ }
124
+ },
125
+ "re-txt_avg_tokens": 18.91060998790392,
126
+ "bert-txt_avg_tokens": 30.66431657162606,
127
+ "gpt2-txt_avg_tokens": 26.351667530672195,
128
+ "re-lb_avg_tokens": 1.0,
129
+ "bert-lb_avg_tokens": 3.0,
130
+ "gpt2-lb_avg_tokens": 1.0
131
+ },
132
+ "emotion": {
133
+ "path": "UTCD/in-domain/emotion",
134
+ "aspect": "sentiment",
135
+ "eval_labels_same": true,
136
+ "domain": "in",
137
+ "name": "Emotion",
138
+ "name_compact": "Emotion",
139
+ "splits": {
140
+ "train": {
141
+ "labels": [
142
+ "anger",
143
+ "fear",
144
+ "joy",
145
+ "love",
146
+ "sadness",
147
+ "surprise"
148
+ ],
149
+ "n_label": 6,
150
+ "n_text": 15969,
151
+ "n_pair": 15999,
152
+ "multi_label": true,
153
+ "n_multi_label": 30
154
+ },
155
+ "test": {
156
+ "labels": [
157
+ "anger",
158
+ "fear",
159
+ "joy",
160
+ "love",
161
+ "sadness",
162
+ "surprise"
163
+ ],
164
+ "n_label": 6,
165
+ "n_text": 2000,
166
+ "n_pair": 2000,
167
+ "multi_label": false,
168
+ "n_multi_label": 0
169
+ }
170
+ },
171
+ "re-txt_avg_tokens": 19.169736768879737,
172
+ "bert-txt_avg_tokens": 22.773610106294175,
173
+ "gpt2-txt_avg_tokens": 20.000278256998165,
174
+ "re-lb_avg_tokens": 1.0,
175
+ "bert-lb_avg_tokens": 3.0,
176
+ "gpt2-lb_avg_tokens": 1.7385410300572255
177
+ },
178
+ "sgd": {
179
+ "path": "UTCD/in-domain/sgd",
180
+ "aspect": "intent",
181
+ "eval_labels_same": false,
182
+ "domain": "in",
183
+ "name": "Schema-Guided Dialogue",
184
+ "name_compact": "SGD",
185
+ "splits": {
186
+ "train": {
187
+ "labels": [
188
+ "Book Appointment",
189
+ "Book House",
190
+ "Buy Bus Ticket",
191
+ "Buy Event Tickets",
192
+ "Check Balance",
193
+ "Find Apartment",
194
+ "Find Attractions",
195
+ "Find Bus",
196
+ "Find Events",
197
+ "Find Movies",
198
+ "Find Provider",
199
+ "Find Restaurants",
200
+ "Get Available Time",
201
+ "Get Cars Available",
202
+ "Get Event Dates",
203
+ "Get Events",
204
+ "Get Ride",
205
+ "Get Times For Movie",
206
+ "Get Weather",
207
+ "Lookup Music",
208
+ "Lookup Song",
209
+ "Play Media",
210
+ "Play Movie",
211
+ "Play Song",
212
+ "Reserve Car",
213
+ "Reserve Hotel",
214
+ "Reserve One way Flight",
215
+ "Reserve Restaurant",
216
+ "Reserve Round trip Flights",
217
+ "Search Hotel",
218
+ "Search House",
219
+ "Search One way Flight",
220
+ "Search Round trip Flights",
221
+ "Transfer Money"
222
+ ],
223
+ "n_label": 34,
224
+ "n_text": 16142,
225
+ "n_pair": 16142,
226
+ "multi_label": false,
227
+ "n_multi_label": 0
228
+ },
229
+ "test": {
230
+ "labels": [
231
+ "Add Alarm",
232
+ "Buy Bus Ticket",
233
+ "Buy Event Tickets",
234
+ "Buy Movie Tickets",
235
+ "Find Attractions",
236
+ "Find Bus",
237
+ "Find Events",
238
+ "Find Home By Area",
239
+ "Find Movies",
240
+ "Find Provider",
241
+ "Find Restaurants",
242
+ "Find Trains",
243
+ "Get Alarms",
244
+ "Get Cars Available",
245
+ "Get Ride",
246
+ "Get Weather",
247
+ "Lookup Music",
248
+ "Make Payment",
249
+ "Play Movie",
250
+ "Request Payment",
251
+ "Reserve Restaurant",
252
+ "Schedule Visit",
253
+ "Search Hotel",
254
+ "Search House",
255
+ "Search One way Flight",
256
+ "Search Round trip Flights"
257
+ ],
258
+ "n_label": 26,
259
+ "n_text": 4201,
260
+ "n_pair": 4201,
261
+ "multi_label": false,
262
+ "n_multi_label": 0
263
+ }
264
+ },
265
+ "re-txt_avg_tokens": 226.98967703878483,
266
+ "bert-txt_avg_tokens": 337.6333382490291,
267
+ "gpt2-txt_avg_tokens": 331.08071572531094,
268
+ "re-lb_avg_tokens": 2.404168510052598,
269
+ "bert-lb_avg_tokens": 5.232709039964607,
270
+ "gpt2-lb_avg_tokens": 2.700290026053188
271
+ },
272
+ "clinc_150": {
273
+ "path": "UTCD/in-domain/clinc_150",
274
+ "aspect": "intent",
275
+ "eval_labels_same": true,
276
+ "domain": "in",
277
+ "name": "Clinc-150",
278
+ "name_compact": "Clinc-150",
279
+ "splits": {
280
+ "train": {
281
+ "labels": [
282
+ "accept reservations",
283
+ "account blocked",
284
+ "alarm",
285
+ "application status",
286
+ "apr",
287
+ "are you a bot",
288
+ "balance",
289
+ "bill balance",
290
+ "bill due",
291
+ "book flight",
292
+ "book hotel",
293
+ "calculator",
294
+ "calendar",
295
+ "calendar update",
296
+ "calories",
297
+ "cancel",
298
+ "cancel reservation",
299
+ "car rental",
300
+ "card declined",
301
+ "carry on",
302
+ "change accent",
303
+ "change ai name",
304
+ "change language",
305
+ "change speed",
306
+ "change user name",
307
+ "change volume",
308
+ "confirm reservation",
309
+ "cook time",
310
+ "credit limit",
311
+ "credit limit change",
312
+ "credit score",
313
+ "current location",
314
+ "damaged card",
315
+ "date",
316
+ "definition",
317
+ "direct deposit",
318
+ "directions",
319
+ "distance",
320
+ "do you have pets",
321
+ "exchange rate",
322
+ "expiration date",
323
+ "find phone",
324
+ "flight status",
325
+ "flip coin",
326
+ "food last",
327
+ "freeze account",
328
+ "fun fact",
329
+ "gas",
330
+ "gas type",
331
+ "goodbye",
332
+ "greeting",
333
+ "how busy",
334
+ "how old are you",
335
+ "improve credit score",
336
+ "income",
337
+ "ingredient substitution",
338
+ "ingredients list",
339
+ "insurance",
340
+ "insurance change",
341
+ "interest rate",
342
+ "international fees",
343
+ "international visa",
344
+ "jump start",
345
+ "last maintenance",
346
+ "lost luggage",
347
+ "make call",
348
+ "maybe",
349
+ "meal suggestion",
350
+ "meaning of life",
351
+ "measurement conversion",
352
+ "meeting schedule",
353
+ "min payment",
354
+ "mpg",
355
+ "new card",
356
+ "next holiday",
357
+ "next song",
358
+ "no",
359
+ "nutrition info",
360
+ "oil change how",
361
+ "oil change when",
362
+ "order",
363
+ "order checks",
364
+ "order status",
365
+ "paid time off request status",
366
+ "paid time off used",
367
+ "pay bill",
368
+ "payday",
369
+ "pin change",
370
+ "play music",
371
+ "plug type",
372
+ "pto balance",
373
+ "pto request",
374
+ "recipe",
375
+ "redeem rewards",
376
+ "reminder",
377
+ "reminder update",
378
+ "repeat",
379
+ "replacement card duration",
380
+ "report fraud",
381
+ "report lost card",
382
+ "reset settings",
383
+ "restaurant reservation",
384
+ "restaurant reviews",
385
+ "restaurant suggestion",
386
+ "rewards balance",
387
+ "roll dice",
388
+ "rollover 401k",
389
+ "routing",
390
+ "schedule maintenance",
391
+ "schedule meeting",
392
+ "share location",
393
+ "shopping list",
394
+ "shopping list update",
395
+ "smart home",
396
+ "spelling",
397
+ "spending history",
398
+ "sync device",
399
+ "taxes",
400
+ "tell joke",
401
+ "text",
402
+ "thank you",
403
+ "time",
404
+ "timer",
405
+ "timezone",
406
+ "tire change",
407
+ "tire pressure",
408
+ "todo list",
409
+ "todo list update",
410
+ "traffic",
411
+ "transactions",
412
+ "transfer",
413
+ "translate",
414
+ "travel alert",
415
+ "travel notification",
416
+ "travel suggestion",
417
+ "uber",
418
+ "update playlist",
419
+ "user name",
420
+ "vaccines",
421
+ "w2 wage and tax statement",
422
+ "weather",
423
+ "what are your hobbies",
424
+ "what can i ask you",
425
+ "what is your name",
426
+ "what song",
427
+ "where are you from",
428
+ "whisper mode",
429
+ "who do you work for",
430
+ "who made you",
431
+ "yes"
432
+ ],
433
+ "n_label": 150,
434
+ "n_text": 15000,
435
+ "n_pair": 15000,
436
+ "multi_label": false,
437
+ "n_multi_label": 0
438
+ },
439
+ "test": {
440
+ "labels": [
441
+ "accept reservations",
442
+ "account blocked",
443
+ "alarm",
444
+ "application status",
445
+ "apr",
446
+ "are you a bot",
447
+ "balance",
448
+ "bill balance",
449
+ "bill due",
450
+ "book flight",
451
+ "book hotel",
452
+ "calculator",
453
+ "calendar",
454
+ "calendar update",
455
+ "calories",
456
+ "cancel",
457
+ "cancel reservation",
458
+ "car rental",
459
+ "card declined",
460
+ "carry on",
461
+ "change accent",
462
+ "change ai name",
463
+ "change language",
464
+ "change speed",
465
+ "change user name",
466
+ "change volume",
467
+ "confirm reservation",
468
+ "cook time",
469
+ "credit limit",
470
+ "credit limit change",
471
+ "credit score",
472
+ "current location",
473
+ "damaged card",
474
+ "date",
475
+ "definition",
476
+ "direct deposit",
477
+ "directions",
478
+ "distance",
479
+ "do you have pets",
480
+ "exchange rate",
481
+ "expiration date",
482
+ "find phone",
483
+ "flight status",
484
+ "flip coin",
485
+ "food last",
486
+ "freeze account",
487
+ "fun fact",
488
+ "gas",
489
+ "gas type",
490
+ "goodbye",
491
+ "greeting",
492
+ "how busy",
493
+ "how old are you",
494
+ "improve credit score",
495
+ "income",
496
+ "ingredient substitution",
497
+ "ingredients list",
498
+ "insurance",
499
+ "insurance change",
500
+ "interest rate",
501
+ "international fees",
502
+ "international visa",
503
+ "jump start",
504
+ "last maintenance",
505
+ "lost luggage",
506
+ "make call",
507
+ "maybe",
508
+ "meal suggestion",
509
+ "meaning of life",
510
+ "measurement conversion",
511
+ "meeting schedule",
512
+ "min payment",
513
+ "mpg",
514
+ "new card",
515
+ "next holiday",
516
+ "next song",
517
+ "no",
518
+ "nutrition info",
519
+ "oil change how",
520
+ "oil change when",
521
+ "order",
522
+ "order checks",
523
+ "order status",
524
+ "paid time off request status",
525
+ "paid time off used",
526
+ "pay bill",
527
+ "payday",
528
+ "pin change",
529
+ "play music",
530
+ "plug type",
531
+ "pto balance",
532
+ "pto request",
533
+ "recipe",
534
+ "redeem rewards",
535
+ "reminder",
536
+ "reminder update",
537
+ "repeat",
538
+ "replacement card duration",
539
+ "report fraud",
540
+ "report lost card",
541
+ "reset settings",
542
+ "restaurant reservation",
543
+ "restaurant reviews",
544
+ "restaurant suggestion",
545
+ "rewards balance",
546
+ "roll dice",
547
+ "rollover 401k",
548
+ "routing",
549
+ "schedule maintenance",
550
+ "schedule meeting",
551
+ "share location",
552
+ "shopping list",
553
+ "shopping list update",
554
+ "smart home",
555
+ "spelling",
556
+ "spending history",
557
+ "sync device",
558
+ "taxes",
559
+ "tell joke",
560
+ "text",
561
+ "thank you",
562
+ "time",
563
+ "timer",
564
+ "timezone",
565
+ "tire change",
566
+ "tire pressure",
567
+ "todo list",
568
+ "todo list update",
569
+ "traffic",
570
+ "transactions",
571
+ "transfer",
572
+ "translate",
573
+ "travel alert",
574
+ "travel notification",
575
+ "travel suggestion",
576
+ "uber",
577
+ "update playlist",
578
+ "user name",
579
+ "vaccines",
580
+ "w2 wage and tax statement",
581
+ "weather",
582
+ "what are your hobbies",
583
+ "what can i ask you",
584
+ "what is your name",
585
+ "what song",
586
+ "where are you from",
587
+ "whisper mode",
588
+ "who do you work for",
589
+ "who made you",
590
+ "yes"
591
+ ],
592
+ "n_label": 150,
593
+ "n_text": 4500,
594
+ "n_pair": 4500,
595
+ "multi_label": false,
596
+ "n_multi_label": 0
597
+ }
598
+ },
599
+ "re-txt_avg_tokens": 8.451641025641026,
600
+ "bert-txt_avg_tokens": 11.499794871794872,
601
+ "gpt2-txt_avg_tokens": 9.075589743589743,
602
+ "re-lb_avg_tokens": 1.9933333333333334,
603
+ "bert-lb_avg_tokens": 4.206666666666667,
604
+ "gpt2-lb_avg_tokens": 2.433333333333333
605
+ },
606
+ "slurp": {
607
+ "path": "UTCD/in-domain/slurp",
608
+ "aspect": "intent",
609
+ "eval_labels_same": false,
610
+ "domain": "in",
611
+ "name": "SLURP",
612
+ "name_compact": "SLURP",
613
+ "splits": {
614
+ "train": {
615
+ "labels": [
616
+ "add contact",
617
+ "alarm query",
618
+ "alarm remove",
619
+ "alarm set",
620
+ "audio volume down",
621
+ "audio volume mute",
622
+ "audio volume other",
623
+ "audio volume up",
624
+ "calendar query",
625
+ "calendar remove",
626
+ "calendar set",
627
+ "cleaning",
628
+ "coffee",
629
+ "convert",
630
+ "cooking query",
631
+ "cooking recipe",
632
+ "create or add",
633
+ "currency",
634
+ "date time convert",
635
+ "date time query",
636
+ "definition",
637
+ "email add contact",
638
+ "email query",
639
+ "email query contact",
640
+ "email send email",
641
+ "events",
642
+ "factoid",
643
+ "game",
644
+ "general greet",
645
+ "general joke",
646
+ "general quirky",
647
+ "greet",
648
+ "hue light dim",
649
+ "hue light off",
650
+ "hue light up",
651
+ "iot cleaning",
652
+ "iot coffee",
653
+ "iot hue light change",
654
+ "iot hue light dim",
655
+ "iot hue light off",
656
+ "iot hue light on",
657
+ "iot hue light up",
658
+ "iot wemo on",
659
+ "iot wemo plug off",
660
+ "joke",
661
+ "lists create or add",
662
+ "lists query",
663
+ "lists remove",
664
+ "music",
665
+ "music dislike ness",
666
+ "music likeness",
667
+ "music query",
668
+ "music settings",
669
+ "news query",
670
+ "play audiobook",
671
+ "play game",
672
+ "play music",
673
+ "play podcasts",
674
+ "play radio",
675
+ "podcasts",
676
+ "post",
677
+ "qa currency",
678
+ "qa definition",
679
+ "qa factoid",
680
+ "qa maths",
681
+ "qa stock",
682
+ "query",
683
+ "query contact",
684
+ "quirky",
685
+ "radio",
686
+ "recommendation events",
687
+ "recommendation locations",
688
+ "recommendation movies",
689
+ "remove",
690
+ "send email",
691
+ "set",
692
+ "settings",
693
+ "social post",
694
+ "social query",
695
+ "take away order",
696
+ "take away query",
697
+ "ticket",
698
+ "traffic",
699
+ "transport query",
700
+ "transport taxi",
701
+ "transport ticket",
702
+ "transport traffic",
703
+ "volume other",
704
+ "weather query",
705
+ "wemo off",
706
+ "wemo plug on"
707
+ ],
708
+ "n_label": 91,
709
+ "n_text": 11502,
710
+ "n_pair": 11503,
711
+ "multi_label": true,
712
+ "n_multi_label": 1
713
+ },
714
+ "test": {
715
+ "labels": [
716
+ "alarm query",
717
+ "alarm remove",
718
+ "alarm set",
719
+ "audio volume down",
720
+ "audio volume mute",
721
+ "audio volume other",
722
+ "audio volume up",
723
+ "calendar query",
724
+ "calendar remove",
725
+ "calendar set",
726
+ "cleaning",
727
+ "cooking recipe",
728
+ "create or add",
729
+ "date time convert",
730
+ "date time query",
731
+ "email add contact",
732
+ "email query",
733
+ "email query contact",
734
+ "email send email",
735
+ "game",
736
+ "general greet",
737
+ "general joke",
738
+ "general quirky",
739
+ "hue light dim",
740
+ "hue light off",
741
+ "hue light up",
742
+ "iot cleaning",
743
+ "iot coffee",
744
+ "iot hue light change",
745
+ "iot hue light dim",
746
+ "iot hue light off",
747
+ "iot hue light on",
748
+ "iot hue light up",
749
+ "iot wemo on",
750
+ "iot wemo plug off",
751
+ "joke",
752
+ "lists create or add",
753
+ "lists query",
754
+ "lists remove",
755
+ "music",
756
+ "music dislike ness",
757
+ "music likeness",
758
+ "music query",
759
+ "music settings",
760
+ "news query",
761
+ "play audiobook",
762
+ "play game",
763
+ "play music",
764
+ "play podcasts",
765
+ "play radio",
766
+ "podcasts",
767
+ "qa currency",
768
+ "qa definition",
769
+ "qa factoid",
770
+ "qa maths",
771
+ "qa stock",
772
+ "query",
773
+ "quirky",
774
+ "radio",
775
+ "recommendation events",
776
+ "recommendation locations",
777
+ "recommendation movies",
778
+ "remove",
779
+ "send email",
780
+ "set",
781
+ "social post",
782
+ "social query",
783
+ "take away order",
784
+ "take away query",
785
+ "transport query",
786
+ "transport taxi",
787
+ "transport ticket",
788
+ "transport traffic",
789
+ "weather query",
790
+ "wemo off"
791
+ ],
792
+ "n_label": 75,
793
+ "n_text": 2592,
794
+ "n_pair": 2592,
795
+ "multi_label": false,
796
+ "n_multi_label": 0
797
+ }
798
+ },
799
+ "re-txt_avg_tokens": 6.932311621966794,
800
+ "bert-txt_avg_tokens": 9.88250319284802,
801
+ "gpt2-txt_avg_tokens": 7.519369944657301,
802
+ "re-lb_avg_tokens": 2.2376019865200427,
803
+ "bert-lb_avg_tokens": 4.951472153245832,
804
+ "gpt2-lb_avg_tokens": 2.6293721177722595
805
+ },
806
+ "ag_news": {
807
+ "path": "UTCD/in-domain/ag_news",
808
+ "aspect": "topic",
809
+ "eval_labels_same": true,
810
+ "domain": "in",
811
+ "name": "AG News",
812
+ "name_compact": "AG News",
813
+ "splits": {
814
+ "train": {
815
+ "labels": [
816
+ "Business",
817
+ "Science & Technology",
818
+ "Sports",
819
+ "World News"
820
+ ],
821
+ "n_label": 4,
822
+ "n_text": 120000,
823
+ "n_pair": 120000,
824
+ "multi_label": false,
825
+ "n_multi_label": 0
826
+ },
827
+ "test": {
828
+ "labels": [
829
+ "Business",
830
+ "Science & Technology",
831
+ "Sports",
832
+ "World News"
833
+ ],
834
+ "n_label": 4,
835
+ "n_text": 7600,
836
+ "n_pair": 7600,
837
+ "multi_label": false,
838
+ "n_multi_label": 0
839
+ }
840
+ },
841
+ "re-txt_avg_tokens": 39.64653605015674,
842
+ "bert-txt_avg_tokens": 57.16152821316614,
843
+ "gpt2-txt_avg_tokens": 52.08076018808777,
844
+ "re-lb_avg_tokens": 1.5,
845
+ "bert-lb_avg_tokens": 3.75,
846
+ "gpt2-lb_avg_tokens": 1.75
847
+ },
848
+ "dbpedia": {
849
+ "path": "UTCD/in-domain/dbpedia",
850
+ "aspect": "topic",
851
+ "eval_labels_same": true,
852
+ "domain": "in",
853
+ "name": "DBpedia",
854
+ "name_compact": "DBpedia",
855
+ "splits": {
856
+ "train": {
857
+ "labels": [
858
+ "Album",
859
+ "Animal",
860
+ "Artist",
861
+ "Athlete",
862
+ "Building",
863
+ "Company",
864
+ "Educational Institution",
865
+ "Film",
866
+ "Mean Of Transportation",
867
+ "Natural Place",
868
+ "Office Holder",
869
+ "Plant",
870
+ "Village",
871
+ "Written Work"
872
+ ],
873
+ "n_label": 14,
874
+ "n_text": 559879,
875
+ "n_pair": 559879,
876
+ "multi_label": false,
877
+ "n_multi_label": 0
878
+ },
879
+ "test": {
880
+ "labels": [
881
+ "Album",
882
+ "Animal",
883
+ "Artist",
884
+ "Athlete",
885
+ "Building",
886
+ "Company",
887
+ "Educational Institution",
888
+ "Film",
889
+ "Mean Of Transportation",
890
+ "Natural Place",
891
+ "Office Holder",
892
+ "Plant",
893
+ "Village",
894
+ "Written Work"
895
+ ],
896
+ "n_label": 14,
897
+ "n_text": 70000,
898
+ "n_pair": 70000,
899
+ "multi_label": false,
900
+ "n_multi_label": 0
901
+ }
902
+ },
903
+ "re-txt_avg_tokens": 47.514625824960035,
904
+ "bert-txt_avg_tokens": 66.15378350445086,
905
+ "gpt2-txt_avg_tokens": 65.47365763900686,
906
+ "re-lb_avg_tokens": 1.4284902338385626,
907
+ "bert-lb_avg_tokens": 3.5713716443951933,
908
+ "gpt2-lb_avg_tokens": 1.9285561195086676
909
+ },
910
+ "yahoo": {
911
+ "path": "UTCD/in-domain/yahoo",
912
+ "aspect": "topic",
913
+ "eval_labels_same": true,
914
+ "domain": "in",
915
+ "name": "Yahoo Answer Topics",
916
+ "name_compact": "Yahoo",
917
+ "splits": {
918
+ "train": {
919
+ "labels": [
920
+ "Business & Finance",
921
+ "Computers & Internet",
922
+ "Education & Reference",
923
+ "Entertainment & Music",
924
+ "Family & Relationships",
925
+ "Health",
926
+ "Politics & Government",
927
+ "Science & Mathematics",
928
+ "Society & Culture",
929
+ "Sports"
930
+ ],
931
+ "n_label": 10,
932
+ "n_text": 1365398,
933
+ "n_pair": 1374775,
934
+ "multi_label": true,
935
+ "n_multi_label": 7059
936
+ },
937
+ "test": {
938
+ "labels": [
939
+ "Business & Finance",
940
+ "Computers & Internet",
941
+ "Education & Reference",
942
+ "Entertainment & Music",
943
+ "Family & Relationships",
944
+ "Health",
945
+ "Politics & Government",
946
+ "Science & Mathematics",
947
+ "Society & Culture",
948
+ "Sports"
949
+ ],
950
+ "n_label": 10,
951
+ "n_text": 59767,
952
+ "n_pair": 59867,
953
+ "multi_label": true,
954
+ "n_multi_label": 73
955
+ }
956
+ },
957
+ "re-txt_avg_tokens": 11.046990348485965,
958
+ "bert-txt_avg_tokens": 17.002040465489962,
959
+ "gpt2-txt_avg_tokens": 13.989866436517877,
960
+ "re-lb_avg_tokens": 1.7997632858929267,
961
+ "bert-lb_avg_tokens": 4.7987714008094,
962
+ "gpt2-lb_avg_tokens": 2.9990227527146147
963
+ },
964
+ "amazon_polarity": {
965
+ "path": "UTCD/out-of-domain/amazon_polarity",
966
+ "aspect": "sentiment",
967
+ "eval_labels_same": true,
968
+ "domain": "out",
969
+ "name": "Amazon Review Polarity",
970
+ "name_compact": "Amazon Polarity",
971
+ "splits": {
972
+ "train": {
973
+ "labels": [
974
+ "negative",
975
+ "positive"
976
+ ],
977
+ "n_label": 2,
978
+ "n_text": 3594783,
979
+ "n_pair": 3594783,
980
+ "multi_label": false,
981
+ "n_multi_label": 0
982
+ },
983
+ "test": {
984
+ "labels": [
985
+ "negative",
986
+ "positive"
987
+ ],
988
+ "n_label": 2,
989
+ "n_text": 399917,
990
+ "n_pair": 399917,
991
+ "multi_label": false,
992
+ "n_multi_label": 0
993
+ }
994
+ },
995
+ "re-txt_avg_tokens": 76.45573706544108,
996
+ "bert-txt_avg_tokens": 99.28606685887321,
997
+ "gpt2-txt_avg_tokens": 91.57986282153547,
998
+ "re-lb_avg_tokens": 1.0,
999
+ "bert-lb_avg_tokens": 3.0,
1000
+ "gpt2-lb_avg_tokens": 1.0
1001
+ },
1002
+ "finance_sentiment": {
1003
+ "path": "UTCD/out-of-domain/finance_sentiment",
1004
+ "aspect": "sentiment",
1005
+ "eval_labels_same": true,
1006
+ "domain": "out",
1007
+ "name": "Financial Phrase Bank",
1008
+ "name_compact": "Fin. Phrase Bank",
1009
+ "splits": {
1010
+ "train": {
1011
+ "labels": [
1012
+ "negative",
1013
+ "neutral",
1014
+ "positive"
1015
+ ],
1016
+ "n_label": 3,
1017
+ "n_text": 1808,
1018
+ "n_pair": 1808,
1019
+ "multi_label": false,
1020
+ "n_multi_label": 0
1021
+ },
1022
+ "test": {
1023
+ "labels": [
1024
+ "negative",
1025
+ "neutral",
1026
+ "positive"
1027
+ ],
1028
+ "n_label": 3,
1029
+ "n_text": 453,
1030
+ "n_pair": 453,
1031
+ "multi_label": false,
1032
+ "n_multi_label": 0
1033
+ }
1034
+ },
1035
+ "re-txt_avg_tokens": 20.50110375275938,
1036
+ "bert-txt_avg_tokens": 30.507726269315672,
1037
+ "gpt2-txt_avg_tokens": 27.368653421633553,
1038
+ "re-lb_avg_tokens": 1.0,
1039
+ "bert-lb_avg_tokens": 3.0,
1040
+ "gpt2-lb_avg_tokens": 1.0
1041
+ },
1042
+ "yelp": {
1043
+ "path": "UTCD/out-of-domain/yelp",
1044
+ "aspect": "sentiment",
1045
+ "eval_labels_same": true,
1046
+ "domain": "out",
1047
+ "name": "Yelp Review",
1048
+ "name_compact": "Yelp",
1049
+ "splits": {
1050
+ "train": {
1051
+ "labels": [
1052
+ "negative",
1053
+ "neutral",
1054
+ "positive"
1055
+ ],
1056
+ "n_label": 3,
1057
+ "n_text": 650000,
1058
+ "n_pair": 650000,
1059
+ "multi_label": false,
1060
+ "n_multi_label": 0
1061
+ },
1062
+ "test": {
1063
+ "labels": [
1064
+ "negative",
1065
+ "neutral",
1066
+ "positive"
1067
+ ],
1068
+ "n_label": 3,
1069
+ "n_text": 50000,
1070
+ "n_pair": 50000,
1071
+ "multi_label": false,
1072
+ "n_multi_label": 0
1073
+ }
1074
+ },
1075
+ "re-txt_avg_tokens": 140.22632,
1076
+ "bert-txt_avg_tokens": 188.35528,
1077
+ "gpt2-txt_avg_tokens": 175.77972,
1078
+ "re-lb_avg_tokens": 1.0,
1079
+ "bert-lb_avg_tokens": 3.0,
1080
+ "gpt2-lb_avg_tokens": 1.0
1081
+ },
1082
+ "banking77": {
1083
+ "path": "UTCD/out-of-domain/banking77",
1084
+ "aspect": "intent",
1085
+ "eval_labels_same": true,
1086
+ "domain": "out",
1087
+ "name": "Banking77",
1088
+ "name_compact": "Banking77",
1089
+ "splits": {
1090
+ "train": {
1091
+ "labels": [
1092
+ "Refund not showing up",
1093
+ "activate my card",
1094
+ "age limit",
1095
+ "apple pay or google pay",
1096
+ "atm support",
1097
+ "automatic top up",
1098
+ "balance not updated after bank transfer",
1099
+ "balance not updated after cheque or cash deposit",
1100
+ "beneficiary not allowed",
1101
+ "cancel transfer",
1102
+ "card about to expire",
1103
+ "card acceptance",
1104
+ "card arrival",
1105
+ "card delivery estimate",
1106
+ "card linking",
1107
+ "card not working",
1108
+ "card payment fee charged",
1109
+ "card payment not recognised",
1110
+ "card payment wrong exchange rate",
1111
+ "card swallowed",
1112
+ "cash withdrawal charge",
1113
+ "cash withdrawal not recognised",
1114
+ "change pin",
1115
+ "compromised card",
1116
+ "contactless not working",
1117
+ "country support",
1118
+ "declined card payment",
1119
+ "declined cash withdrawal",
1120
+ "declined transfer",
1121
+ "direct debit payment not recognised",
1122
+ "disposable card limits",
1123
+ "edit personal details",
1124
+ "exchange charge",
1125
+ "exchange rate",
1126
+ "exchange via app",
1127
+ "extra charge on statement",
1128
+ "failed transfer",
1129
+ "fiat currency support",
1130
+ "get disposable virtual card",
1131
+ "get physical card",
1132
+ "getting spare card",
1133
+ "getting virtual card",
1134
+ "lost or stolen card",
1135
+ "lost or stolen phone",
1136
+ "order physical card",
1137
+ "passcode forgotten",
1138
+ "pending card payment",
1139
+ "pending cash withdrawal",
1140
+ "pending top up",
1141
+ "pending transfer",
1142
+ "pin blocked",
1143
+ "receiving money",
1144
+ "request refund",
1145
+ "reverted card payment?",
1146
+ "supported cards and currencies",
1147
+ "terminate account",
1148
+ "top up by bank transfer charge",
1149
+ "top up by card charge",
1150
+ "top up by cash or cheque",
1151
+ "top up failed",
1152
+ "top up limits",
1153
+ "top up reverted",
1154
+ "topping up by card",
1155
+ "transaction charged twice",
1156
+ "transfer fee charged",
1157
+ "transfer into account",
1158
+ "transfer not received by recipient",
1159
+ "transfer timing",
1160
+ "unable to verify identity",
1161
+ "verify my identity",
1162
+ "verify source of funds",
1163
+ "verify top up",
1164
+ "virtual card not working",
1165
+ "visa or mastercard",
1166
+ "why verify identity",
1167
+ "wrong amount of cash received",
1168
+ "wrong exchange rate for cash withdrawal"
1169
+ ],
1170
+ "n_label": 77,
1171
+ "n_text": 10003,
1172
+ "n_pair": 10003,
1173
+ "multi_label": false,
1174
+ "n_multi_label": 0
1175
+ },
1176
+ "test": {
1177
+ "labels": [
1178
+ "Refund not showing up",
1179
+ "activate my card",
1180
+ "age limit",
1181
+ "apple pay or google pay",
1182
+ "atm support",
1183
+ "automatic top up",
1184
+ "balance not updated after bank transfer",
1185
+ "balance not updated after cheque or cash deposit",
1186
+ "beneficiary not allowed",
1187
+ "cancel transfer",
1188
+ "card about to expire",
1189
+ "card acceptance",
1190
+ "card arrival",
1191
+ "card delivery estimate",
1192
+ "card linking",
1193
+ "card not working",
1194
+ "card payment fee charged",
1195
+ "card payment not recognised",
1196
+ "card payment wrong exchange rate",
1197
+ "card swallowed",
1198
+ "cash withdrawal charge",
1199
+ "cash withdrawal not recognised",
1200
+ "change pin",
1201
+ "compromised card",
1202
+ "contactless not working",
1203
+ "country support",
1204
+ "declined card payment",
1205
+ "declined cash withdrawal",
1206
+ "declined transfer",
1207
+ "direct debit payment not recognised",
1208
+ "disposable card limits",
1209
+ "edit personal details",
1210
+ "exchange charge",
1211
+ "exchange rate",
1212
+ "exchange via app",
1213
+ "extra charge on statement",
1214
+ "failed transfer",
1215
+ "fiat currency support",
1216
+ "get disposable virtual card",
1217
+ "get physical card",
1218
+ "getting spare card",
1219
+ "getting virtual card",
1220
+ "lost or stolen card",
1221
+ "lost or stolen phone",
1222
+ "order physical card",
1223
+ "passcode forgotten",
1224
+ "pending card payment",
1225
+ "pending cash withdrawal",
1226
+ "pending top up",
1227
+ "pending transfer",
1228
+ "pin blocked",
1229
+ "receiving money",
1230
+ "request refund",
1231
+ "reverted card payment?",
1232
+ "supported cards and currencies",
1233
+ "terminate account",
1234
+ "top up by bank transfer charge",
1235
+ "top up by card charge",
1236
+ "top up by cash or cheque",
1237
+ "top up failed",
1238
+ "top up limits",
1239
+ "top up reverted",
1240
+ "topping up by card",
1241
+ "transaction charged twice",
1242
+ "transfer fee charged",
1243
+ "transfer into account",
1244
+ "transfer not received by recipient",
1245
+ "transfer timing",
1246
+ "unable to verify identity",
1247
+ "verify my identity",
1248
+ "verify source of funds",
1249
+ "verify top up",
1250
+ "virtual card not working",
1251
+ "visa or mastercard",
1252
+ "why verify identity",
1253
+ "wrong amount of cash received",
1254
+ "wrong exchange rate for cash withdrawal"
1255
+ ],
1256
+ "n_label": 77,
1257
+ "n_text": 3080,
1258
+ "n_pair": 3080,
1259
+ "multi_label": false,
1260
+ "n_multi_label": 0
1261
+ }
1262
+ },
1263
+ "re-txt_avg_tokens": 11.219805194805195,
1264
+ "bert-txt_avg_tokens": 15.266883116883117,
1265
+ "gpt2-txt_avg_tokens": 12.672727272727272,
1266
+ "re-lb_avg_tokens": 3.2857142857142856,
1267
+ "bert-lb_avg_tokens": 5.662337662337662,
1268
+ "gpt2-lb_avg_tokens": 3.7532467532467533
1269
+ },
1270
+ "snips": {
1271
+ "path": "UTCD/out-of-domain/snips",
1272
+ "aspect": "intent",
1273
+ "eval_labels_same": true,
1274
+ "domain": "out",
1275
+ "name": "SNIPS",
1276
+ "name_compact": "SNIPS",
1277
+ "splits": {
1278
+ "train": {
1279
+ "labels": [
1280
+ "Add To Playlist",
1281
+ "Book Restaurant",
1282
+ "Get Weather",
1283
+ "Play Music",
1284
+ "Rate Book",
1285
+ "Search Creative Work",
1286
+ "Search Screening Event"
1287
+ ],
1288
+ "n_label": 7,
1289
+ "n_text": 13615,
1290
+ "n_pair": 13615,
1291
+ "multi_label": false,
1292
+ "n_multi_label": 0
1293
+ },
1294
+ "test": {
1295
+ "labels": [
1296
+ "Add To Playlist",
1297
+ "Book Restaurant",
1298
+ "Get Weather",
1299
+ "Play Music",
1300
+ "Rate Book",
1301
+ "Search Creative Work",
1302
+ "Search Screening Event"
1303
+ ],
1304
+ "n_label": 7,
1305
+ "n_text": 697,
1306
+ "n_pair": 697,
1307
+ "multi_label": false,
1308
+ "n_multi_label": 0
1309
+ }
1310
+ },
1311
+ "re-txt_avg_tokens": 9.154949784791965,
1312
+ "bert-txt_avg_tokens": 13.558106169296988,
1313
+ "gpt2-txt_avg_tokens": 11.057388809182209,
1314
+ "re-lb_avg_tokens": 2.430416068866571,
1315
+ "bert-lb_avg_tokens": 4.860832137733142,
1316
+ "gpt2-lb_avg_tokens": 2.7173601147776183
1317
+ },
1318
+ "nlu_evaluation": {
1319
+ "path": "UTCD/out-of-domain/nlu_evaluation",
1320
+ "aspect": "intent",
1321
+ "eval_labels_same": true,
1322
+ "domain": "out",
1323
+ "name": "NLU Evaluation",
1324
+ "name_compact": "NLU Eval",
1325
+ "splits": {
1326
+ "train": {
1327
+ "labels": [
1328
+ "alarm query",
1329
+ "alarm remove",
1330
+ "alarm set",
1331
+ "audio volume down",
1332
+ "audio volume mute",
1333
+ "audio volume other",
1334
+ "audio volume up",
1335
+ "calendar query",
1336
+ "calendar remove",
1337
+ "calendar set",
1338
+ "cooking query",
1339
+ "cooking recipe",
1340
+ "datetime convert",
1341
+ "datetime query",
1342
+ "email addcontact",
1343
+ "email query",
1344
+ "email querycontact",
1345
+ "email sendemail",
1346
+ "general affirm",
1347
+ "general commandstop",
1348
+ "general confirm",
1349
+ "general dontcare",
1350
+ "general explain",
1351
+ "general greet",
1352
+ "general joke",
1353
+ "general negate",
1354
+ "general praise",
1355
+ "general quirky",
1356
+ "general repeat",
1357
+ "iot cleaning",
1358
+ "iot coffee",
1359
+ "iot hue lightchange",
1360
+ "iot hue lightdim",
1361
+ "iot hue lightoff",
1362
+ "iot hue lighton",
1363
+ "iot hue lightup",
1364
+ "iot wemo off",
1365
+ "iot wemo on",
1366
+ "lists createoradd",
1367
+ "lists query",
1368
+ "lists remove",
1369
+ "music dislikeness",
1370
+ "music likeness",
1371
+ "music query",
1372
+ "music settings",
1373
+ "news query",
1374
+ "play audiobook",
1375
+ "play game",
1376
+ "play music",
1377
+ "play podcasts",
1378
+ "play radio",
1379
+ "qa currency",
1380
+ "qa definition",
1381
+ "qa factoid",
1382
+ "qa maths",
1383
+ "qa stock",
1384
+ "recommendation events",
1385
+ "recommendation locations",
1386
+ "recommendation movies",
1387
+ "social post",
1388
+ "social query",
1389
+ "takeaway order",
1390
+ "takeaway query",
1391
+ "transport query",
1392
+ "transport taxi",
1393
+ "transport ticket",
1394
+ "transport traffic",
1395
+ "weather query"
1396
+ ],
1397
+ "n_label": 68,
1398
+ "n_text": 20524,
1399
+ "n_pair": 20530,
1400
+ "multi_label": true,
1401
+ "n_multi_label": 6
1402
+ },
1403
+ "test": {
1404
+ "labels": [
1405
+ "alarm query",
1406
+ "alarm remove",
1407
+ "alarm set",
1408
+ "audio volume down",
1409
+ "audio volume mute",
1410
+ "audio volume other",
1411
+ "audio volume up",
1412
+ "calendar query",
1413
+ "calendar remove",
1414
+ "calendar set",
1415
+ "cooking query",
1416
+ "cooking recipe",
1417
+ "datetime convert",
1418
+ "datetime query",
1419
+ "email addcontact",
1420
+ "email query",
1421
+ "email querycontact",
1422
+ "email sendemail",
1423
+ "general affirm",
1424
+ "general commandstop",
1425
+ "general confirm",
1426
+ "general dontcare",
1427
+ "general explain",
1428
+ "general greet",
1429
+ "general joke",
1430
+ "general negate",
1431
+ "general praise",
1432
+ "general quirky",
1433
+ "general repeat",
1434
+ "iot cleaning",
1435
+ "iot coffee",
1436
+ "iot hue lightchange",
1437
+ "iot hue lightdim",
1438
+ "iot hue lightoff",
1439
+ "iot hue lighton",
1440
+ "iot hue lightup",
1441
+ "iot wemo off",
1442
+ "iot wemo on",
1443
+ "lists createoradd",
1444
+ "lists query",
1445
+ "lists remove",
1446
+ "music dislikeness",
1447
+ "music likeness",
1448
+ "music query",
1449
+ "music settings",
1450
+ "news query",
1451
+ "play audiobook",
1452
+ "play game",
1453
+ "play music",
1454
+ "play podcasts",
1455
+ "play radio",
1456
+ "qa currency",
1457
+ "qa definition",
1458
+ "qa factoid",
1459
+ "qa maths",
1460
+ "qa stock",
1461
+ "recommendation events",
1462
+ "recommendation locations",
1463
+ "recommendation movies",
1464
+ "social post",
1465
+ "social query",
1466
+ "takeaway order",
1467
+ "takeaway query",
1468
+ "transport query",
1469
+ "transport taxi",
1470
+ "transport ticket",
1471
+ "transport traffic",
1472
+ "weather query"
1473
+ ],
1474
+ "n_label": 68,
1475
+ "n_text": 5165,
1476
+ "n_pair": 5165,
1477
+ "multi_label": false,
1478
+ "n_multi_label": 0
1479
+ }
1480
+ },
1481
+ "re-txt_avg_tokens": 6.8321393998063895,
1482
+ "bert-txt_avg_tokens": 9.963213939980639,
1483
+ "gpt2-txt_avg_tokens": 7.631558567279767,
1484
+ "re-lb_avg_tokens": 2.0536302032913842,
1485
+ "bert-lb_avg_tokens": 4.991868344627299,
1486
+ "gpt2-lb_avg_tokens": 2.530300096805421
1487
+ },
1488
+ "multi_eurlex": {
1489
+ "path": "UTCD/out-of-domain/multi_eurlex",
1490
+ "aspect": "topic",
1491
+ "eval_labels_same": true,
1492
+ "domain": "out",
1493
+ "name": "MultiEURLEX",
1494
+ "name_compact": "MultiEURLEX",
1495
+ "splits": {
1496
+ "train": {
1497
+ "labels": [
1498
+ "EUROPEAN UNION",
1499
+ "agri-foodstuffs",
1500
+ "agriculture, forestry and fisheries",
1501
+ "business and competition",
1502
+ "economics",
1503
+ "education and communications",
1504
+ "employment and working conditions",
1505
+ "energy",
1506
+ "environment",
1507
+ "finance",
1508
+ "geography",
1509
+ "industry",
1510
+ "international organisations",
1511
+ "international relations",
1512
+ "law",
1513
+ "politics",
1514
+ "production, technology and research",
1515
+ "science",
1516
+ "social questions",
1517
+ "trade",
1518
+ "transport"
1519
+ ],
1520
+ "n_label": 21,
1521
+ "n_text": 54986,
1522
+ "n_pair": 177719,
1523
+ "multi_label": true,
1524
+ "n_multi_label": 53925
1525
+ },
1526
+ "test": {
1527
+ "labels": [
1528
+ "EUROPEAN UNION",
1529
+ "agri-foodstuffs",
1530
+ "agriculture, forestry and fisheries",
1531
+ "business and competition",
1532
+ "economics",
1533
+ "education and communications",
1534
+ "employment and working conditions",
1535
+ "energy",
1536
+ "environment",
1537
+ "finance",
1538
+ "geography",
1539
+ "industry",
1540
+ "international organisations",
1541
+ "international relations",
1542
+ "law",
1543
+ "politics",
1544
+ "production, technology and research",
1545
+ "science",
1546
+ "social questions",
1547
+ "trade",
1548
+ "transport"
1549
+ ],
1550
+ "n_label": 21,
1551
+ "n_text": 5000,
1552
+ "n_pair": 17969,
1553
+ "multi_label": true,
1554
+ "n_multi_label": 4939
1555
+ }
1556
+ },
1557
+ "re-txt_avg_tokens": 1889.4374,
1558
+ "bert-txt_avg_tokens": 2436.3836,
1559
+ "gpt2-txt_avg_tokens": 2426.6724,
1560
+ "re-lb_avg_tokens": 1.834882297289777,
1561
+ "bert-lb_avg_tokens": 4.810006121653959,
1562
+ "gpt2-lb_avg_tokens": 3.133396404919584
1563
+ },
1564
+ "patent": {
1565
+ "path": "UTCD/out-of-domain/patent",
1566
+ "aspect": "topic",
1567
+ "eval_labels_same": true,
1568
+ "domain": "out",
1569
+ "name": "Big Patent",
1570
+ "name_compact": "Patent",
1571
+ "splits": {
1572
+ "train": {
1573
+ "labels": [
1574
+ "Chemistry; Metallurgy",
1575
+ "Electricity",
1576
+ "Fixed Constructions",
1577
+ "General tagging of new or cross-sectional technology",
1578
+ "Human Necessities",
1579
+ "Mechanical Engineering; Lightning; Heating; Weapons; Blasting",
1580
+ "Performing Operations; Transporting",
1581
+ "Physics",
1582
+ "Textiles; Paper"
1583
+ ],
1584
+ "n_label": 9,
1585
+ "n_text": 24950,
1586
+ "n_pair": 24956,
1587
+ "multi_label": true,
1588
+ "n_multi_label": 6
1589
+ },
1590
+ "test": {
1591
+ "labels": [
1592
+ "Chemistry; Metallurgy",
1593
+ "Electricity",
1594
+ "Fixed Constructions",
1595
+ "General tagging of new or cross-sectional technology",
1596
+ "Human Necessities",
1597
+ "Mechanical Engineering; Lightning; Heating; Weapons; Blasting",
1598
+ "Performing Operations; Transporting",
1599
+ "Physics",
1600
+ "Textiles; Paper"
1601
+ ],
1602
+ "n_label": 9,
1603
+ "n_text": 4997,
1604
+ "n_pair": 4997,
1605
+ "multi_label": false,
1606
+ "n_multi_label": 0
1607
+ }
1608
+ },
1609
+ "re-txt_avg_tokens": 3130.2157294376625,
1610
+ "bert-txt_avg_tokens": 4231.3251951170705,
1611
+ "gpt2-txt_avg_tokens": 3963.504902941765,
1612
+ "re-lb_avg_tokens": 2.5919551731038624,
1613
+ "bert-lb_avg_tokens": 6.424054432659596,
1614
+ "gpt2-lb_avg_tokens": 4.598559135481289
1615
+ },
1616
+ "consumer_finance": {
1617
+ "path": "UTCD/out-of-domain/consumer_finance",
1618
+ "aspect": "topic",
1619
+ "eval_labels_same": true,
1620
+ "domain": "out",
1621
+ "name": "Consumer Finance Complaints",
1622
+ "name_compact": "Consumer Finance",
1623
+ "splits": {
1624
+ "train": {
1625
+ "labels": [
1626
+ "Bank account or service",
1627
+ "Checking or savings account",
1628
+ "Consumer Loan",
1629
+ "Credit card",
1630
+ "Credit card or prepaid card",
1631
+ "Credit reporting",
1632
+ "Credit reporting, credit repair services, or other personal consumer reports",
1633
+ "Debt collection",
1634
+ "Money transfer, virtual currency, or money service",
1635
+ "Money transfers",
1636
+ "Mortgage",
1637
+ "Other financial service",
1638
+ "Payday loan",
1639
+ "Payday loan, title loan, or personal loan",
1640
+ "Prepaid card",
1641
+ "Student loan",
1642
+ "Vehicle loan or lease",
1643
+ "Virtual currency"
1644
+ ],
1645
+ "n_label": 18,
1646
+ "n_text": 626004,
1647
+ "n_pair": 626489,
1648
+ "multi_label": true,
1649
+ "n_multi_label": 399
1650
+ },
1651
+ "test": {
1652
+ "labels": [
1653
+ "Bank account or service",
1654
+ "Checking or savings account",
1655
+ "Consumer Loan",
1656
+ "Credit card",
1657
+ "Credit card or prepaid card",
1658
+ "Credit reporting",
1659
+ "Credit reporting, credit repair services, or other personal consumer reports",
1660
+ "Debt collection",
1661
+ "Money transfer, virtual currency, or money service",
1662
+ "Money transfers",
1663
+ "Mortgage",
1664
+ "Other financial service",
1665
+ "Payday loan",
1666
+ "Payday loan, title loan, or personal loan",
1667
+ "Prepaid card",
1668
+ "Student loan",
1669
+ "Vehicle loan or lease",
1670
+ "Virtual currency"
1671
+ ],
1672
+ "n_label": 18,
1673
+ "n_text": 156602,
1674
+ "n_pair": 156635,
1675
+ "multi_label": true,
1676
+ "n_multi_label": 30
1677
+ }
1678
+ },
1679
+ "re-txt_avg_tokens": 202.49348028760807,
1680
+ "bert-txt_avg_tokens": 293.5908736797742,
1681
+ "gpt2-txt_avg_tokens": 254.61107137839875,
1682
+ "re-lb_avg_tokens": 5.312324831614901,
1683
+ "bert-lb_avg_tokens": 8.884297889999042,
1684
+ "gpt2-lb_avg_tokens": 6.611644906949277
1685
+ }
1686
+ }
datasets.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2cee07002b203df72a56ac5b09a56ac625b741ba053b136ecdbceca1036c7193
3
- size 1744154943
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20d9d447b2c283880abe2a28fbf1f3617cdceda0f3788a1f545ae8df884e7847
3
+ size 1744258165
datasets_joined.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2cee07002b203df72a56ac5b09a56ac625b741ba053b136ecdbceca1036c7193
3
+ size 1744154943
dev.py ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ from stefutil import *
2
+
3
+
4
+ if __name__ == '__main__':
5
+ from datasets import load_dataset
6
+
7
+ dnm = 'claritylab/utcd'
8
+ mic(load_dataset(dnm))
utcd.py CHANGED
@@ -0,0 +1,143 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """Universal Text Classification Dataset (UTCD)"""
2
+
3
+
4
+ import os
5
+ import json
6
+ from os.path import join as os_join
7
+ from typing import List
8
+
9
+ import datasets
10
+
11
+ from stefutil import *
12
+
13
+
14
+ _DESCRIPTION = """
15
+ UTCD is a compilation of 18 classification datasets spanning 3 categories of Sentiment,
16
+ Intent/Dialogue and Topic classification. UTCD focuses on the task of zero-shot text classification where the
17
+ candidate labels are descriptive of the text being classified. UTCD consists of ~ 6M/800K train/test examples.
18
+ """
19
+
20
+ # TODO: citation
21
+
22
+ _URL = "https://github.com/ChrisIsKing/zero-shot-text-classification/tree/master"
23
+ _URL_ZIP = "https://huggingface.co/datasets/claritylab/UTCD/raw/main/datasets.zip"
24
+
25
+ _VERSION = datasets.Version('0.0.1')
26
+
27
+
28
+ class UtcdConfig(datasets.BuilderConfig):
29
+ """BuilderConfig for SuperGLUE."""
30
+
31
+ def __init__(self, domain: str, normalize_aspect: bool = False, **kwargs):
32
+ """BuilderConfig for UTCD.
33
+ Args:
34
+ domain: `string`, dataset domain, one of [`in`, `out`].
35
+ normalize_aspect: `bool`, if True, an aspect-normalized version of the dataset is returned.
36
+ **kwargs: keyword arguments forwarded to super.
37
+ """
38
+ # Version history:
39
+ # 0.0.1: Initial version.
40
+ super(UtcdConfig, self).__init__(version=_VERSION, **kwargs)
41
+ ca.check_mismatch('Dataset Domain', domain, ['in', 'out'])
42
+ self.domain = domain
43
+ self.normalize_aspect = normalize_aspect
44
+
45
+ def to_dir_name(self):
46
+ """
47
+ :return: directory name for the dataset files for this config stored on hub
48
+ """
49
+ domain_str = 'in-domain' if self.domain == 'in' else 'out-of-domain'
50
+ prefix = 'aspect-normalized-' if self.normalize_aspect else ''
51
+ return f'{prefix}{domain_str}'
52
+
53
+
54
+ config = StefConfig('config.json')
55
+ # mic(config('go_emotion'))
56
+
57
+
58
+ class Utcd(datasets.GeneratorBasedBuilder):
59
+ """UTCD: Universal Text Classification Dataset. Version 0.0."""
60
+
61
+ # _config = dict(
62
+ # go_emotion=dict(aspect='sentiment', domain='in', name='GoEmotions'),
63
+ # sentiment_tweets_2020=dict(aspect='sentiment', domain='in', name='TweetEval'),
64
+ # emotion=dict(aspect='sentiment', domain='in', name='Emotion'),
65
+ # sgd=dict(aspect='intent', domain='in', name='Schema-Guided Dialogue'),
66
+ # clinc_150=dict(aspect='intent', domain='in', name='Clinc-150'),
67
+ # slurp=dict(aspect='intent', domain='in', name='SLURP'),
68
+ # ag_news=dict(aspect='topic', domain='in', name='AG News'),
69
+ # dbpedia=dict(aspect='topic', domain='in', name='DBpedia'),
70
+ # yahoo=dict(aspect='topic', domain='in', name='Yahoo Answer Topics'),
71
+ #
72
+ # amazon_polarity=dict(aspect='sentiment', domain='out', name='Amazon Review Polarity'),
73
+ # finance_sentiment=dict( aspect='sentiment', domain='out', name='Financial Phrase Bank'),
74
+ # yelp=dict(aspect='sentiment', domain='out', name='Yelp Review'),
75
+ # banking77=dict(aspect='intent', domain='out', name='Banking77'),
76
+ # snips=dict(aspect='intent', domain='out', name='SNIPS'),
77
+ # nlu_evaluation=dict(aspect='intent', domain='out', name='NLU Evaluation'),
78
+ # multi_eurlex=dict(aspect='topic', domain='out', name='MultiEURLEX'),
79
+ # patent=dict(aspect='topic', domain='out', name='Big Patent'),
80
+ # consumer_finance=dict(aspect='topic', domain='out', name='Consumer Finance Complaints')
81
+ # )
82
+
83
+ VERSION = _VERSION
84
+
85
+ BUILDER_CONFIGS = [
86
+ UtcdConfig(
87
+ name='in-domain',
88
+ description='All in-domain datasets.',
89
+ domain='in',
90
+ normalize_aspect=False
91
+ ),
92
+ UtcdConfig(
93
+ name='aspect-normalized-in-domain',
94
+ description='Aspect-normalized version of all in-domain datasets.',
95
+ domain='in',
96
+ normalize_aspect=True
97
+ ),
98
+ UtcdConfig(
99
+ name='out-of-domain',
100
+ description='All out-of-domain datasets.',
101
+ domain='out',
102
+ normalize_aspect=False
103
+ ),
104
+ UtcdConfig(
105
+ name='aspect-normalized-out-of-domain',
106
+ description='Aspect-normalized version of all out-of-domain datasets.',
107
+ domain='out',
108
+ normalize_aspect=True
109
+ )
110
+ ]
111
+
112
+ def _get_dataset_names(self):
113
+ return [dnm for dnm, d_dset in config().items() if d_dset['domain'] == self.config.domain]
114
+
115
+ def _info(self):
116
+ dnms = self._get_dataset_names()
117
+ labels = [config(f'{dnm}.splits.{split}.labels') for dnm in dnms for split in ['train', 'test']]
118
+ mic(dnms, labels)
119
+ return datasets.DatasetInfo(
120
+ description=_DESCRIPTION,
121
+ features=datasets.Features(
122
+ text=datasets.Value(dtype='string'), labels=labels, dataset_name=datasets.ClassLabel(names=dnms)
123
+ ),
124
+ homepage=_URL
125
+ # TODO: citation
126
+ )
127
+
128
+ def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
129
+ # for aspect-normalized versions of the dataset, we include a validation set
130
+ splits = ['train', 'eval', 'test'] if self.config.normalize_aspect else ['train', 'test']
131
+ dnms = self._get_dataset_names()
132
+ dir_nm = self.config.to_dir_name()
133
+ split2paths = {s: [os_join(dir_nm, dnm, s) for dnm in dnms] for s in splits}
134
+ mic(split2paths)
135
+
136
+ downloaded_files = dl_manager.download_and_extract('datasets.zip')
137
+ mic(downloaded_files)
138
+ raise NotImplementedError
139
+
140
+ return [
141
+ datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files["train"]}),
142
+ datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files["dev"]}),
143
+ ]