ivykopal commited on
Commit
8b393c0
1 Parent(s): 7f11275

Upload templates.json

Browse files
Files changed (1) hide show
  1. templates.json +1063 -0
templates.json ADDED
@@ -0,0 +1,1063 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": [
3
+ {
4
+ "dataset": "aeslc",
5
+ "model": "T5",
6
+ "name": "aeslc-prompt-t5",
7
+ "input": "summarize: {{email_body}}",
8
+ "target": "{{subject_line}}",
9
+ "metadata": {
10
+ "languages": [
11
+ "en"
12
+ ],
13
+ "metrics": [
14
+ "rouge"
15
+ ]
16
+ },
17
+ "languages": [
18
+ "en"
19
+ ],
20
+ "metrics": [
21
+ "rouge"
22
+ ]
23
+ },
24
+ {
25
+ "dataset": "anli",
26
+ "model": "T5",
27
+ "name": "anli-prompt-t5",
28
+ "input": "premise: {{premise}} hypothesis: {{hypothesis}}",
29
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
30
+ "metadata": {
31
+ "languages": [
32
+ "en"
33
+ ],
34
+ "metrics": [
35
+ "accuracy"
36
+ ]
37
+ },
38
+ "languages": [
39
+ "en"
40
+ ],
41
+ "metrics": [
42
+ "accuracy"
43
+ ]
44
+ },
45
+ {
46
+ "dataset": "billsum",
47
+ "model": "T5",
48
+ "name": "billsum-prompt-t5",
49
+ "input": "summarize: {{text}}",
50
+ "target": "{{summary}}",
51
+ "metadata": {
52
+ "languages": [
53
+ "en"
54
+ ],
55
+ "metrics": [
56
+ "rouge"
57
+ ]
58
+ },
59
+ "languages": [
60
+ "en"
61
+ ],
62
+ "metrics": [
63
+ "rouge"
64
+ ]
65
+ },
66
+ {
67
+ "dataset": "boolq",
68
+ "model": "T5",
69
+ "name": "boolq-prompt-t5",
70
+ "input": "boolq question: {{question}} passage: {{passage}}",
71
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
72
+ "metadata": {
73
+ "languages": [
74
+ "en"
75
+ ],
76
+ "metrics": [
77
+ "accuracy"
78
+ ]
79
+ },
80
+ "languages": [
81
+ "en"
82
+ ],
83
+ "metrics": [
84
+ "accuracy"
85
+ ]
86
+ },
87
+ {
88
+ "dataset": "c4",
89
+ "model": "T5",
90
+ "name": "c4-prompt-t5",
91
+ "input": "",
92
+ "target": "{{text}}",
93
+ "metadata": {
94
+ "languages": [
95
+ "en"
96
+ ],
97
+ "metrics": [
98
+ "rouge"
99
+ ]
100
+ },
101
+ "languages": [
102
+ "en"
103
+ ],
104
+ "metrics": [
105
+ "rouge"
106
+ ]
107
+ },
108
+ {
109
+ "dataset": "cb",
110
+ "model": "T5",
111
+ "name": "cb-prompt-t5",
112
+ "input": "cb premise: {{premise}} hypothesis: {{hypothesis}}",
113
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
114
+ "metadata": {
115
+ "languages": [
116
+ "en"
117
+ ],
118
+ "metrics": [
119
+ "accuracy",
120
+ "multiclass_f1"
121
+ ]
122
+ },
123
+ "languages": [
124
+ "en"
125
+ ],
126
+ "metrics": [
127
+ "accuracy",
128
+ "multiclass_f1"
129
+ ]
130
+ },
131
+ {
132
+ "dataset": "cnn_dailymail",
133
+ "model": "T5",
134
+ "name": "cnn_dailymail-prompt-t5",
135
+ "input": "summarize: {{article}}",
136
+ "target": "{{highlights}}",
137
+ "metadata": {
138
+ "languages": [
139
+ "en"
140
+ ],
141
+ "metrics": [
142
+ "rouge"
143
+ ]
144
+ },
145
+ "languages": [
146
+ "en"
147
+ ],
148
+ "metrics": [
149
+ "rouge"
150
+ ]
151
+ },
152
+ {
153
+ "dataset": "cola",
154
+ "model": "T5",
155
+ "name": "cola-prompt-t5",
156
+ "input": "cola sentence: {{sentence}}",
157
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
158
+ "metadata": {
159
+ "languages": [
160
+ "en"
161
+ ],
162
+ "metrics": [
163
+ "matthews_corrcoef"
164
+ ]
165
+ },
166
+ "languages": [
167
+ "en"
168
+ ],
169
+ "metrics": [
170
+ "matthews_corrcoef"
171
+ ]
172
+ },
173
+ {
174
+ "dataset": "common_gen",
175
+ "model": "T5",
176
+ "name": "common_gen-prompt-t5",
177
+ "input": "generate: {{lambda concepts: \" \".join(concepts)}}",
178
+ "target": "{{target}}",
179
+ "metadata": {
180
+ "languages": [
181
+ "en"
182
+ ],
183
+ "metrics": [
184
+ "rouge"
185
+ ]
186
+ },
187
+ "languages": [
188
+ "en"
189
+ ],
190
+ "metrics": [
191
+ "rouge"
192
+ ]
193
+ },
194
+ {
195
+ "dataset": "copa",
196
+ "model": "T5",
197
+ "name": "copa-prompt-t5",
198
+ "input": "copa premise: {{premise}} choice1: {{choice1}} choice2: {{choice2}}",
199
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
200
+ "metadata": {
201
+ "languages": [
202
+ "en"
203
+ ],
204
+ "metrics": [
205
+ "accuracy"
206
+ ]
207
+ },
208
+ "languages": [
209
+ "en"
210
+ ],
211
+ "metrics": [
212
+ "accuracy"
213
+ ]
214
+ },
215
+ {
216
+ "dataset": "cosmos_qa",
217
+ "model": "T5",
218
+ "name": "cosmos_qa-prompt-t5",
219
+ "input": "question: {{question}} context: {{context}} choice0: {{answer0}} choice1: {{answer1}} choice2: {{answer2}} choice3: {{answer3}}",
220
+ "target": "{{lambda label: str(label)}}",
221
+ "metadata": {
222
+ "languages": [
223
+ "en"
224
+ ],
225
+ "metrics": [
226
+ "squad"
227
+ ]
228
+ },
229
+ "languages": [
230
+ "en"
231
+ ],
232
+ "metrics": [
233
+ "squad"
234
+ ]
235
+ },
236
+ {
237
+ "dataset": "cxc",
238
+ "model": "T5",
239
+ "name": "cxc-prompt-t5",
240
+ "input": "sentence1: {{sentence1}} sentence2: {{sentence2}}",
241
+ "target": "{{lambda score: np.round((score * 5) / 5, decimals=1)}}",
242
+ "metadata": {
243
+ "languages": [
244
+ "en"
245
+ ],
246
+ "metrics": [
247
+ "pearson_corrcoef",
248
+ "spearman_corrcoef"
249
+ ]
250
+ },
251
+ "languages": [
252
+ "en"
253
+ ],
254
+ "metrics": [
255
+ "pearson_corrcoef",
256
+ "spearman_corrcoef"
257
+ ]
258
+ },
259
+ {
260
+ "dataset": "doc_nli",
261
+ "model": "T5",
262
+ "name": "doc_nli-prompt-t5",
263
+ "input": "premise: {{premise}} hypothesis: {{hypothesis}}",
264
+ "target": "{{label}}",
265
+ "metadata": {
266
+ "languages": [
267
+ "en"
268
+ ],
269
+ "metrics": [
270
+ "accuracy"
271
+ ]
272
+ },
273
+ "languages": [
274
+ "en"
275
+ ],
276
+ "metrics": [
277
+ "accuracy"
278
+ ]
279
+ },
280
+ {
281
+ "dataset": "drop",
282
+ "model": "T5",
283
+ "name": "drop-prompt-t5",
284
+ "input": "question: {{question}} context: {{passage}}",
285
+ "target": "{{answers_spans[\"spans\"][0]}}",
286
+ "metadata": {
287
+ "languages": [
288
+ "en"
289
+ ],
290
+ "metrics": [
291
+ "squad"
292
+ ],
293
+ "preprocessing": [
294
+ "pad_punctuation"
295
+ ]
296
+ },
297
+ "languages": [
298
+ "en"
299
+ ],
300
+ "metrics": [
301
+ "squad"
302
+ ]
303
+ },
304
+ {
305
+ "dataset": "gigaword",
306
+ "model": "T5",
307
+ "name": "gigaword-prompt-t5",
308
+ "input": "summarize: {{document}}",
309
+ "target": "{{summary}}",
310
+ "metadata": {
311
+ "languages": [
312
+ "en"
313
+ ],
314
+ "metrics": [
315
+ "rouge"
316
+ ]
317
+ },
318
+ "languages": [
319
+ "en"
320
+ ],
321
+ "metrics": [
322
+ "rouge"
323
+ ]
324
+ },
325
+ {
326
+ "dataset": "hellaswag",
327
+ "model": "T5",
328
+ "name": "hellaswag-prompt-t5",
329
+ "input": "context: {{ctx}} ending0: {{endings[0]}} ending1: {{endings[1]}} ending2: {{endings[2]}} ending3: {{endings[3]}}",
330
+ "target": "{{lambda label: str(label)}}",
331
+ "metadata": {
332
+ "languages": [
333
+ "en"
334
+ ],
335
+ "metrics": [
336
+ "accuracy"
337
+ ]
338
+ },
339
+ "languages": [
340
+ "en"
341
+ ],
342
+ "metrics": [
343
+ "accuracy"
344
+ ]
345
+ },
346
+ {
347
+ "dataset": "hotpot_qa",
348
+ "model": "T5",
349
+ "name": "hotpot_qa-prompt-t5",
350
+ "input": "question: {{question}} context: {{\"\".join(context[\"sentences\"][0])}}",
351
+ "target": "{{answer}}",
352
+ "metadata": {
353
+ "languages": [
354
+ "en"
355
+ ],
356
+ "metrics": [
357
+ "squad"
358
+ ],
359
+ "preprocessing": [
360
+ "pad_punctuation"
361
+ ]
362
+ },
363
+ "languages": [
364
+ "en"
365
+ ],
366
+ "metrics": [
367
+ "squad"
368
+ ]
369
+ },
370
+ {
371
+ "dataset": "mnli",
372
+ "model": "T5",
373
+ "name": "mnli-prompt-t5",
374
+ "input": "mnli premise: {{premise}} hypothesis: {{hypothesis}}",
375
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
376
+ "metadata": {
377
+ "languages": [
378
+ "en"
379
+ ],
380
+ "metrics": [
381
+ "accuracy"
382
+ ]
383
+ },
384
+ "languages": [
385
+ "en"
386
+ ],
387
+ "metrics": [
388
+ "accuracy"
389
+ ]
390
+ },
391
+ {
392
+ "dataset": "mrpc",
393
+ "model": "T5",
394
+ "name": "mrpc-prompt-t5",
395
+ "input": "mrpc sentence1: {{sentence1}} sentence2: {{sentence2}}",
396
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
397
+ "metadata": {
398
+ "languages": [
399
+ "en"
400
+ ],
401
+ "metrics": [
402
+ "accuracy",
403
+ "f1_invalid"
404
+ ]
405
+ },
406
+ "languages": [
407
+ "en"
408
+ ],
409
+ "metrics": [
410
+ "accuracy",
411
+ "f1_invalid"
412
+ ]
413
+ },
414
+ {
415
+ "dataset": "mrqa",
416
+ "model": "T5",
417
+ "name": "mrqa-prompt-t5",
418
+ "input": "question: {{question}} context: {{context}}",
419
+ "target": "{{answers[0]}}",
420
+ "metadata": {
421
+ "languages": [
422
+ "en"
423
+ ],
424
+ "metrics": [
425
+ "squad"
426
+ ],
427
+ "preprocessing": [
428
+ "pad_punctuation"
429
+ ]
430
+ },
431
+ "languages": [
432
+ "en"
433
+ ],
434
+ "metrics": [
435
+ "squad"
436
+ ]
437
+ },
438
+ {
439
+ "dataset": "multirc",
440
+ "model": "T5",
441
+ "name": "multirc-prompt-t5",
442
+ "input": "multirc question: {{question}} answer: {{answer}} paragraph: {{paragraph}}",
443
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
444
+ "metadata": {
445
+ "languages": [
446
+ "en"
447
+ ],
448
+ "metrics": [
449
+ "exact_match",
450
+ "multirc_f1"
451
+ ],
452
+ "preprocessing": [
453
+ "remove_markup"
454
+ ]
455
+ },
456
+ "languages": [
457
+ "en"
458
+ ],
459
+ "metrics": [
460
+ "exact_match",
461
+ "multirc_f1"
462
+ ]
463
+ },
464
+ {
465
+ "dataset": "multi_news",
466
+ "model": "T5",
467
+ "name": "multi_news-prompt-t5",
468
+ "input": "summarize: {{document}}",
469
+ "target": "{{summary}}",
470
+ "metadata": {
471
+ "languages": [
472
+ "en"
473
+ ],
474
+ "metrics": [
475
+ "rouge"
476
+ ]
477
+ },
478
+ "languages": [
479
+ "en"
480
+ ],
481
+ "metrics": [
482
+ "rouge"
483
+ ]
484
+ },
485
+ {
486
+ "dataset": "multi_nli",
487
+ "model": "T5",
488
+ "name": "multi_nli-prompt-t5",
489
+ "input": "premise: {{premise}} hypothesis: {{hypothesis}}",
490
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
491
+ "metadata": {
492
+ "languages": [
493
+ "en"
494
+ ],
495
+ "metrics": [
496
+ "accuracy"
497
+ ]
498
+ },
499
+ "languages": [
500
+ "en"
501
+ ],
502
+ "metrics": [
503
+ "accuracy"
504
+ ]
505
+ },
506
+ {
507
+ "dataset": "newsqa",
508
+ "model": "T5",
509
+ "name": "newsqa-prompt-t5",
510
+ "input": "question: {{question}} context: {{context}}",
511
+ "target": "{{answer[0]}}",
512
+ "metadata": {
513
+ "languages": [
514
+ "en"
515
+ ],
516
+ "metrics": [
517
+ "rouge"
518
+ ],
519
+ "preprocessing": [
520
+ "pad_punctuation"
521
+ ]
522
+ },
523
+ "languages": [
524
+ "en"
525
+ ],
526
+ "metrics": [
527
+ "rouge"
528
+ ]
529
+ },
530
+ {
531
+ "dataset": "newsqa",
532
+ "model": "T5",
533
+ "name": "newsqa-prompt-t5-without-context",
534
+ "input": "question: {{question}}",
535
+ "target": "{{answer[0]}}",
536
+ "metadata": {
537
+ "languages": [
538
+ "en"
539
+ ],
540
+ "metrics": [
541
+ "rouge"
542
+ ],
543
+ "preprocessing": [
544
+ "pad_punctuation"
545
+ ]
546
+ },
547
+ "languages": [
548
+ "en"
549
+ ],
550
+ "metrics": [
551
+ "rouge"
552
+ ]
553
+ },
554
+ {
555
+ "dataset": "newsroom",
556
+ "model": "T5",
557
+ "name": "newsroom-prompt-t5",
558
+ "input": "summarize: {{text}}",
559
+ "target": "{{summary}}",
560
+ "metadata": {
561
+ "languages": [
562
+ "en"
563
+ ],
564
+ "metrics": [
565
+ "rouge"
566
+ ]
567
+ },
568
+ "languages": [
569
+ "en"
570
+ ],
571
+ "metrics": [
572
+ "rouge"
573
+ ]
574
+ },
575
+ {
576
+ "dataset": "nq_open",
577
+ "model": "T5",
578
+ "name": "nq_open-prompt-t5",
579
+ "input": "nq question: {{question}}",
580
+ "target": "{{answer[0]}}",
581
+ "metadata": {
582
+ "languages": [
583
+ "en"
584
+ ],
585
+ "metrics": [
586
+ "squad"
587
+ ],
588
+ "preprocessing": [
589
+ "pad_punctuation"
590
+ ]
591
+ },
592
+ "languages": [
593
+ "en"
594
+ ],
595
+ "metrics": [
596
+ "squad"
597
+ ]
598
+ },
599
+ {
600
+ "dataset": "piqa",
601
+ "model": "T5",
602
+ "name": "piqa-prompt-t5",
603
+ "input": "question: {{goal}} choice1: {{sol1}} choice2: {{sol2}}",
604
+ "target": "{{lambda label: str(label)}}",
605
+ "metadata": {
606
+ "languages": [
607
+ "en"
608
+ ],
609
+ "metrics": [
610
+ "accuracy"
611
+ ]
612
+ },
613
+ "languages": [
614
+ "en"
615
+ ],
616
+ "metrics": [
617
+ "accuracy"
618
+ ]
619
+ },
620
+ {
621
+ "dataset": "qnli",
622
+ "model": "T5",
623
+ "name": "qnli-prompt-t5",
624
+ "input": "qnli question: {{question}} sentence: {{sentence}}",
625
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
626
+ "metadata": {
627
+ "languages": [
628
+ "en"
629
+ ],
630
+ "metrics": [
631
+ "accuracy"
632
+ ]
633
+ },
634
+ "languages": [
635
+ "en"
636
+ ],
637
+ "metrics": [
638
+ "accuracy"
639
+ ]
640
+ },
641
+ {
642
+ "dataset": "qqp",
643
+ "model": "T5",
644
+ "name": "qqp-prompt-t5",
645
+ "input": "qqp question1: {{question1}} question2: {{question2}}",
646
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
647
+ "metadata": {
648
+ "languages": [
649
+ "en"
650
+ ],
651
+ "metrics": [
652
+ "accuracy",
653
+ "f1_invalid"
654
+ ]
655
+ },
656
+ "languages": [
657
+ "en"
658
+ ],
659
+ "metrics": [
660
+ "accuracy",
661
+ "f1_invalid"
662
+ ]
663
+ },
664
+ {
665
+ "dataset": "race",
666
+ "model": "T5",
667
+ "name": "race-prompt-t5",
668
+ "input": "question: {{question}} context: {{article}} choice0: {{options[0]}} choice1: {{options[1]}} choice2: {{options[2]}} choice3: {{options[3]}}",
669
+ "target": "{{lambda answer: str(ord(answer) - ord(\"A\"))}}",
670
+ "metadata": {
671
+ "languages": [
672
+ "en"
673
+ ],
674
+ "metrics": [
675
+ "accuracy"
676
+ ]
677
+ },
678
+ "languages": [
679
+ "en"
680
+ ],
681
+ "metrics": [
682
+ "accuracy"
683
+ ]
684
+ },
685
+ {
686
+ "dataset": "record",
687
+ "model": "T5",
688
+ "name": "record-prompt-t5",
689
+ "input": "record query: {{query}} entities: {{\", \".join(entities)}} passage: {{passage}}",
690
+ "target": "{{answers[0]}}",
691
+ "metadata": {
692
+ "languages": [
693
+ "en"
694
+ ],
695
+ "metrics": [
696
+ "squad"
697
+ ],
698
+ "preprocessing": [
699
+ "record_preprocess"
700
+ ]
701
+ },
702
+ "languages": [
703
+ "en"
704
+ ],
705
+ "metrics": [
706
+ "squad"
707
+ ]
708
+ },
709
+ {
710
+ "dataset": "rte",
711
+ "model": "T5",
712
+ "name": "rte-prompt-t5",
713
+ "input": "rte sentence1: {{sentence1}} sentence2: {{sentence2}}",
714
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
715
+ "metadata": {
716
+ "languages": [
717
+ "en"
718
+ ],
719
+ "metrics": [
720
+ "accuracy"
721
+ ]
722
+ },
723
+ "languages": [
724
+ "en"
725
+ ],
726
+ "metrics": [
727
+ "accuracy"
728
+ ]
729
+ },
730
+ {
731
+ "dataset": "samsum",
732
+ "model": "T5",
733
+ "name": "samsum-prompt-t5",
734
+ "input": "summarize: {{dialogue}}",
735
+ "target": "{{summary}}",
736
+ "metadata": {
737
+ "languages": [
738
+ "en"
739
+ ],
740
+ "metrics": [
741
+ "rouge"
742
+ ]
743
+ },
744
+ "languages": [
745
+ "en"
746
+ ],
747
+ "metrics": [
748
+ "rouge"
749
+ ]
750
+ },
751
+ {
752
+ "dataset": "search_qa",
753
+ "model": "T5",
754
+ "name": "search_qa-prompt-t5",
755
+ "input": "question: {{question}}",
756
+ "target": "{{answer}}",
757
+ "metadata": {
758
+ "languages": [
759
+ "en"
760
+ ],
761
+ "metrics": [
762
+ "squad"
763
+ ],
764
+ "preprocessing": [
765
+ "pad_punctuation"
766
+ ]
767
+ },
768
+ "languages": [
769
+ "en"
770
+ ],
771
+ "metrics": [
772
+ "squad"
773
+ ]
774
+ },
775
+ {
776
+ "dataset": "snli",
777
+ "model": "T5",
778
+ "name": "snli-prompt-t5",
779
+ "input": "premise: {{premise}} hypothesis: {{hypothesis}}",
780
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
781
+ "metadata": {
782
+ "languages": [
783
+ "en"
784
+ ],
785
+ "metrics": [
786
+ "accuracy"
787
+ ]
788
+ },
789
+ "languages": [
790
+ "en"
791
+ ],
792
+ "metrics": [
793
+ "accuracy"
794
+ ]
795
+ },
796
+ {
797
+ "dataset": "social_i_qa",
798
+ "model": "T5",
799
+ "name": "social_i_qa-prompt-t5",
800
+ "input": "question: {{question}} context: {{context}} || choice0: {{answerA}} || choice1: {{answerB}} || choice2: {{answerC}}",
801
+ "target": "{{lambda label: str(int(label) - 1)}}",
802
+ "metadata": {
803
+ "languages": [
804
+ "en"
805
+ ],
806
+ "metrics": [
807
+ "accuracy"
808
+ ]
809
+ },
810
+ "languages": [
811
+ "en"
812
+ ],
813
+ "metrics": [
814
+ "accuracy"
815
+ ]
816
+ },
817
+ {
818
+ "dataset": "squad",
819
+ "model": "T5",
820
+ "name": "squad-prompt-t5",
821
+ "input": "question: {{question}} context: {{context}}",
822
+ "target": "{{answers[\"text\"][0]}}",
823
+ "metadata": {
824
+ "languages": [
825
+ "en"
826
+ ],
827
+ "metrics": [
828
+ "squad"
829
+ ],
830
+ "preprocessing": [
831
+ "pad_punctuation"
832
+ ]
833
+ },
834
+ "languages": [
835
+ "en"
836
+ ],
837
+ "metrics": [
838
+ "squad"
839
+ ]
840
+ },
841
+ {
842
+ "dataset": "squad",
843
+ "model": "T5",
844
+ "name": "squad-prompt-trivia-t5",
845
+ "input": "squad trivia question: {{question}}",
846
+ "target": "{{answers[\"text\"][0]}}",
847
+ "metadata": {
848
+ "languages": [
849
+ "en"
850
+ ],
851
+ "metrics": [
852
+ "squad"
853
+ ],
854
+ "preprocessing": [
855
+ "pad_punctuation"
856
+ ]
857
+ },
858
+ "languages": [
859
+ "en"
860
+ ],
861
+ "metrics": [
862
+ "squad"
863
+ ]
864
+ },
865
+ {
866
+ "dataset": "sst2",
867
+ "model": "T5",
868
+ "name": "sst2-prompt-t5",
869
+ "input": "sst2 sentence: {{sentence}}",
870
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
871
+ "metadata": {
872
+ "languages": [
873
+ "en"
874
+ ],
875
+ "metrics": [
876
+ "accuracy"
877
+ ],
878
+ "preprocessing": [
879
+ "pad_punctuation"
880
+ ]
881
+ },
882
+ "languages": [
883
+ "en"
884
+ ],
885
+ "metrics": [
886
+ "accuracy"
887
+ ]
888
+ },
889
+ {
890
+ "dataset": "stsb",
891
+ "model": "T5",
892
+ "name": "stsb-prompt-t5",
893
+ "input": "stsb sentence1: {{sentence1}} sentence2: {{sentence2}}",
894
+ "target": "{{lambda label: np.round((label * 5) / 5, decimals=1)}}",
895
+ "metadata": {
896
+ "languages": [
897
+ "en"
898
+ ],
899
+ "metrics": [
900
+ "pearson_corrcoef",
901
+ "spearman_corrcoef"
902
+ ]
903
+ },
904
+ "languages": [
905
+ "en"
906
+ ],
907
+ "metrics": [
908
+ "pearson_corrcoef",
909
+ "spearman_corrcoef"
910
+ ]
911
+ },
912
+ {
913
+ "dataset": "wic",
914
+ "model": "T5",
915
+ "name": "wic-prompt-t5",
916
+ "input": "wic sentence1: {{sentence1}} sentence2: {{sentence2}} word: {{word}}",
917
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
918
+ "metadata": {
919
+ "languages": [
920
+ "en"
921
+ ],
922
+ "metrics": [
923
+ "accuracy"
924
+ ]
925
+ },
926
+ "languages": [
927
+ "en"
928
+ ],
929
+ "metrics": [
930
+ "accuracy"
931
+ ]
932
+ },
933
+ {
934
+ "dataset": "wiki_auto",
935
+ "model": "T5",
936
+ "name": "wiki_auto-prompt-t5",
937
+ "input": "{{source}}",
938
+ "target": "{{target}}",
939
+ "metadata": {
940
+ "languages": [
941
+ "en"
942
+ ],
943
+ "metrics": [
944
+ "bleu"
945
+ ]
946
+ },
947
+ "languages": [
948
+ "en"
949
+ ],
950
+ "metrics": [
951
+ "bleu"
952
+ ]
953
+ },
954
+ {
955
+ "dataset": "wiki_lingua",
956
+ "model": "T5",
957
+ "name": "wiki_lingua-prompt-t5",
958
+ "input": "{{source_aligned[\"en\"]}}",
959
+ "target": "{{target_aligned[\"en\"]}}",
960
+ "metadata": {
961
+ "languages": [
962
+ "en"
963
+ ],
964
+ "metrics": [
965
+ "rouge"
966
+ ]
967
+ },
968
+ "languages": [
969
+ "en"
970
+ ],
971
+ "metrics": [
972
+ "rouge"
973
+ ]
974
+ },
975
+ {
976
+ "dataset": "winogrande",
977
+ "model": "T5",
978
+ "name": "winogrande-prompt-t5",
979
+ "input": "sentence: {{sentence}} option0: {{option1}} option1: {{option2}}",
980
+ "target": "{{lambda answer: str(int(answer) - 1)}}",
981
+ "metadata": {
982
+ "languages": [
983
+ "en"
984
+ ],
985
+ "metrics": [
986
+ "accuracy"
987
+ ]
988
+ },
989
+ "languages": [
990
+ "en"
991
+ ],
992
+ "metrics": [
993
+ "accuracy"
994
+ ]
995
+ },
996
+ {
997
+ "dataset": "wnli",
998
+ "model": "T5",
999
+ "name": "wnli-prompt-t5",
1000
+ "input": "wnli sentence1: {{sentence1}} sentence2: {{sentence2}}",
1001
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
1002
+ "metadata": {
1003
+ "languages": [
1004
+ "en"
1005
+ ],
1006
+ "metrics": [
1007
+ "accuracy"
1008
+ ]
1009
+ },
1010
+ "languages": [
1011
+ "en"
1012
+ ],
1013
+ "metrics": [
1014
+ "accuracy"
1015
+ ]
1016
+ },
1017
+ {
1018
+ "dataset": "wsc",
1019
+ "model": "T5",
1020
+ "name": "wsc-prompt-t5",
1021
+ "input": "wsc text: {{text}}",
1022
+ "target": "{{lambda label: \"<unk>\" if label == -1 else choices[label]}}",
1023
+ "metadata": {
1024
+ "languages": [
1025
+ "en"
1026
+ ],
1027
+ "metrics": [
1028
+ "accuracy"
1029
+ ],
1030
+ "preprocessing": [
1031
+ "wsc_preprocess"
1032
+ ]
1033
+ },
1034
+ "languages": [
1035
+ "en"
1036
+ ],
1037
+ "metrics": [
1038
+ "accuracy"
1039
+ ]
1040
+ },
1041
+ {
1042
+ "dataset": "xsum",
1043
+ "model": "T5",
1044
+ "name": "xsum-prompt-t5",
1045
+ "input": "summarize: {{document}}",
1046
+ "target": "{{target}}",
1047
+ "metadata": {
1048
+ "languages": [
1049
+ "en"
1050
+ ],
1051
+ "metrics": [
1052
+ "rouge"
1053
+ ]
1054
+ },
1055
+ "languages": [
1056
+ "en"
1057
+ ],
1058
+ "metrics": [
1059
+ "rouge"
1060
+ ]
1061
+ }
1062
+ ]
1063
+ }