-
Notifications
You must be signed in to change notification settings - Fork 16
/
Copy pathindex.html
1869 lines (1863 loc) · 105 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<meta name="description" content="">
<meta name="author" content="">
<title>MS MARCO</title>
<link rel="shortcut icon" type="image/x-icon" href="img/favicon.ico" />
<!-- Bootstrap core CSS -->
<link href="css/bootstrap.min.css" rel="stylesheet">
<!-- Custom styles for this template -->
<link href="css/scrolling-nav.css" rel="stylesheet">
<link href="css/main.css" rel="stylesheet">
<link href="css/fa-all.css" rel="stylesheet">
<link rel="stylesheet" href="css/academicons.min.css" />
<link rel="stylesheet" href="https://www.microsoft.com/onerfstatics/marketingsites-eus-prod/west-european/shell/_scrf/css/themes=default.device=uplevel_web_pc/31-9d5f3f/79-6aa410/c6-ce4cc7/f3-7d8ce1/18-6a72f8/26-12908c/88-de543b/18-e17dee?ver=2.0" type="text/css" media="all">
<script defer src="js/addDownloadLinks.js"></script>
</head>
<body id="page-top">
<nav class="navbar navbar-expand-xl navbar-dark fixed-top topnav" id="mainNav">
<div class="container">
<a class="navbar-brand js-scroll-trigger" href="#page-top">MS MARCO</a>
<button class="navbar-toggler" type="button" data-toggle="collapse" data-target="#navbarResponsive" aria-controls="navbarResponsive" aria-expanded="false" aria-label="Toggle navigation">
<span class="navbar-toggler-icon"></span>
</button>
<div class="collapse navbar-collapse" id="navbarResponsive">
<ul class="navbar-nav ml-auto">
<li class="nav-item">
<a class="nav-link js-scroll-trigger" href="#page-top">Home</a>
</li>
<li class="nav-item">
<a class="nav-link js-scroll-trigger" href="#docranking">Document Ranking</a>
</li>
<li class="nav-item">
<a class="nav-link js-scroll-trigger" href="#ranking">Passage Ranking</a>
</li>
<li class="nav-item">
<a class="nav-link js-scroll-trigger" href="#updates">Updates</a>
</li>
<li class="nav-item">
<a class="nav-link js-scroll-trigger" href="#contact">Submissions</a>
</li>
<li class="nav-item">
<a class="nav-link js-scroll-trigger" href="#about">About</a>
</li>
</ul>
</div>
</div>
</nav>
<section id="intro">
<div class="container">
<div class="col-12 d-flex justify-content-center align-items-center mb-5">
<img alt="MSMARCO Logo" src="img/MarcoLogo.small.png" />
<img alt="Microsoft Logo" src="img/MicrosoftLogo.png">
</div>
<div class="row">
<div class="col-12">
<a href="https://twitter.com/MSMarcoAI" class="twitter-follow-button" data-show-count="false">Follow @msmarco</a><script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
</div>
<div class="col-12">
<p>Starting with a paper released at <a href="https://arxiv.org/pdf/1611.09268.pdf"> NIPS 2016</a>, MS MARCO is a collection of datasets focused on deep learning in search.</p>
<p>The first dataset was a question answering dataset featuring 100,000 real Bing questions and a human generated answer. Since then we released a 1,000,000 question dataset, a natural langauge generation dataset, a passage ranking dataset, keyphrase extraction dataset, crawling dataset, and a conversational search.</p>
<div id ='currentstats'></div>
<p><strong>The NLGEN and QnA Leaderboard will close on 10/23/2020. see <a class="nav-link js-scroll-trigger" href="#retirement">DataSet Retirement</a> for details. If you would like to evaluate a model please submit before then</strong></p>
</div>
<div class="col-12">
<h3>Terms and Conditions</h3>
<p>The MS MARCO datasets are intended for non-commercial research purposes only to promote advancement in the field of artificial intelligence and related areas, and is made available free of charge without extending any license or other intellectual property rights. The dataset is provided “as is” without warranty and usage of the data has risks since we may not own the underlying rights in the documents. We are not be liable for any damages related to use of the dataset. Feedback is voluntarily given and can be used as we see fit. Upon violation of any of these terms, your rights to use the dataset will end automatically.</p><br /><br /> <p> Please contact us at <a href="mailto:ms-marco@microsoft.com?subject=MS Marco Issue">ms-marco@microsoft.com</a> if you own any of the documents made available but do not want them in this dataset. We will remove the data accordingly. If you have questions about use of the dataset or any research outputs in your products or services, we encourage you to undertake your own independent legal review. For other questions, please feel free to contact us.</p>
<div class="checkbox">
<label>
<input type="checkbox" name="termAgreed" id="TA" value="true" />
I agree to terms and conditions. Upon accepting links to dataset will become available.
</label>
</div>
</div>
<div class="col-12">
<button type="button" style="background-color:#000000;border-color:#000000;" id="agree-terms" class="btn btn-primary btn-default" >Submit</button>
</div>
</div>
</div>
</section>
<section id="docranking" class="bg-light">
<div class="container">
<div class="row">
<div class="col-md-12 mx-auto">
<h2>Document Retrieval:RETIRED(08/11/2020-01/01/2023)</h2>
<p>Based the questions in the <a href='#qna'>Question Answering Dataset</a> and the documents which answered the questions a document ranking task was formulated. There are 3.2 million documents and the goal is to rank based on their relevance.</p>
<p>Relevance labels are derived from what passages was marked as having the answer in the QnA dataset making this one of the largest relevance datasets ever.</p>
<p>This dataset is the focus of the <a href='https://trec.nist.gov/'>2020 and 2019 TREC Deep Learning Track</a> and has been used as a teaching aid for <a href='http://sigir.org/afirm2020/'>ACM SIGIR/SIGKDD AFIRM Summer School on Machine Learning for Data Mining and Search</a>.</p>
<p>In 2020 we release a set of cleaned and formated clicks for all documents in the collection. This collection of 20 million clicks is called ORCAS.</p>
<h4>Tasks</h4>
<ol>
<li>Document Re-Ranking:Given a candidate top 100 document as retrieved by BM25, re-rank documents by relevance.</li>
<li>Document Full Ranking:Given a corpus of 3.2m documents generate a candidate top 100 documents sorted by relevance.</li>
</ol>
<h4>Relevant Links</h4>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://github.com/microsoft/MSMARCO-Document-Ranking">MSMARCO Document Ranking Github</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://trec.nist.gov/data/deep2019.html">NIST Judgments for TREC 2019 Deep Learning Track</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://arxiv.org/abs/2003.07820">Overview of the TREC 2019 deep learning track</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://arxiv.org/abs/1611.09268">Paper</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://arxiv.org/pdf/2006.05324.pdf">ORCAS Dataset</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://github.com/microsoft/TREC-2020-Deep-Learning">TREC 2020 Deep Learning</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://github.com/microsoft/TREC-2019-Deep-Learning">TREC 2019 Deep Learning</a><br/>
<div id="docrankingdata"> </div>
<iframe src="https://microsoft.github.io/MSMARCO-Document-Ranking-Submissions/leaderboard/" title="Document Ranking Leaderboard" width="100%" height="512" border="0" frameBorder="0"></iframe>
<a href="https://microsoft.github.io/MSMARCO-Document-Ranking-Submissions/leaderboard/">Link to full leaderboard</a>
</div>
</div>
</div>
</div>
</section>
<section id="ranking" class="bg-light">
<div class="container">
<div class="row">
<div class="col-md-12 mx-auto">
<h2>Passage Retrieval:RETIRED(10/26/2018-01/01/2023)</h2>
<p>Based on the passages and questions in the <a href='#qna'>Question Answering Dataset</a>, a passage ranking task was formulated. There are 8.8 million passages and the goal is to rank based on their relevance.</p>
<p>Relevance labels are derived from what passages was marked as having the answer in the QnA dataset making this one of the largest relevance datasets ever.</p>
<p>This dataset is the focus of the <a href='https://trec.nist.gov/'>2020 and 2019 TREC Deep Learning Track</a> and has been used as a teaching aid for <a href='http://sigir.org/afirm2020/'>ACM SIGIR/SIGKDD AFIRM Summer School on Machine Learning for Data Mining and Search</a>.</p>
<p>In 2020 we release a set of cleaned and formated clicks for all documents in the collection. This collection of 20 million clicks is called ORCAS.</p>
<h4>Tasks</h4>
<ol>
<li>Passage Re-Ranking:Given a candidate top 1000 passages as retrieved by BM25, re-rank passage by relevance.</li>
<li>Passage Full Ranking:Given a corpus of 8.8m passages generate a candidate top 1000 passages sorted by relevance.</li>
</ol>
<h4>Relevant Links</h4>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://trec.nist.gov/data/deep2019.html">NIST Judgments for TREC 2019 Deep Learning Track</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://arxiv.org/abs/2003.07820">Overview of the TREC 2019 deep learning track</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://arxiv.org/abs/1611.09268">Paper</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://arxiv.org/pdf/2006.05324.pdf">ORCAS Dataset</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://github.com/microsoft/MSMARCO-Passage-Ranking">MSMARCO Passage Ranking Github</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://github.com/microsoft/TREC-2020-Deep-Learning">TREC 2020 Deep Learning</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://github.com/microsoft/TREC-2019-Deep-Learning">TREC 2019 Deep Learning</a>
<h3>Dataset Download links</h3>
<div id="rankingdata"> </div>
<iframe src="https://microsoft.github.io/MSMARCO-Passage-Ranking-Submissions/leaderboard/" title="Passage Ranking Leaderboard" width="100%" height="512" border="0" frameBorder="0"></iframe>
<a href="https://microsoft.github.io/MSMARCO-Passage-Ranking-Submissions/leaderboard/">Link to full leaderboard</a>
</div>
</div>
</div>
</section>
<section id="kp" class="bg-light">
<div class="container">
<div class="row">
<div class="col-md-12 mx-auto">
<h2>KeyPhrase Extraction:RETIRED(10/18/2019-10/30/2020)</h2>
<p>Keyphrase extraction on open domain document is an up and coming area that can be used for many NLP tasks like document ranking, Topic Clusetring, etc. To enable the research community to build performant KeyPhrase Extraction systems we have build OpenKP a human annotated extraction of Keyphrases on a wide variety of documents.</p>
<p>The dataset features 148,124 real world web documents along with a human annotation indicating the 1-3 most relevant keyphrases. More information about the dataset and our initial experiments can be found in the paper <a href='https://www.aclweb.org/anthology/D19-1521/'>Open Domain Web Keyphrase Extraction Beyond Language Modeling</a> which was an oral presentation at <a href='https://www.emnlp-ijcnlp2019.org/'>EMNLP-IJCNLP 2019</a>. It is part of the MSMARCO dataset family and research projects like this power the core document understanding pipeline that Bing uses.</p>
<h4>Tasks</h4>
<ol>
<li>Given a document produce the top 3 most salient keyphrases</li>
</ol>
<h4>Relevant Links</h4>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://github.com/microsoft/OpenKP">KeyPhrase Extraction Github Repo</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://arxiv.org/abs/1911.02671">Paper</a>
<h3>Dataset Download links</h3>
<div id="keyphrasedata"> </div>
<h4>KeyPhrase Extraction(10/18/2019) ranked by F1 @3 on Eval</h4>
<div class="table-responsive">
<table class="table table-striped" id='kptable'>
<thead>
<tr>
<th>Rank</th>
<th class="col-lg-6">Model</th>
<th class="col-md-1">Submission Date</th>
<th class="col-md-5">F1 @1,<strong>@3</strong>,@5</th>
</tr>
</thead>
<tbody>
<tr>
<td><strong>ETC-large</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">May31 st, 2020</a></td>
<td>0.393, <strong>0.420</strong>, 0.360</td>
</tr>
<tr>
<td><strong>RoBERTa-JointKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>] </td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.364, <strong>0.391</strong>, 0.338</td>
</tr>
<tr>
<td><strong>RoBERTa-RankKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.361, <strong>0.390</strong>, 0.337</td>
</tr>
<tr>
<td><strong>SpanBERT-JointKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.359, <strong>0.385</strong>, 0.335</td>
</tr>
<tr>
<td><strong>RoBERTa-TagKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.356, <strong>0.381</strong>, 0.332</td>
</tr>
<tr>
<td><strong>SpanBERT-RankKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.355, <strong>0.380</strong>, 0.331</td>
</tr>
<tr>
<td><strong>BERT-JointKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.349, <strong>0.376</strong>, 0.325</td>
</tr>
<tr>
<td><strong>SpanBERT-TagKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.351, <strong>0.374</strong>, 0.325</td>
</tr>
<tr>
<td><strong>BERT-RankKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.342, <strong>0.374</strong>, 0.325</td>
</tr>
<tr>
<td><strong>RoBERTa-ChunkKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.355, <strong>0.373</strong>, 0.324</td>
</tr>
<tr>
<td><strong>SpanBERT-ChunkKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.348, <strong>0.372</strong>, 0.324</td>
</tr>
<tr>
<td><strong>BERT-TagKPE (Base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.343, <strong>0.364</strong>, 0.318</td>
</tr>
<tr>
<td><strong>BERT (Base) Sequence Tagging Baseline</strong> Si Sun (Tsinghua University), Chenyan Xiong (MSR AI), Zhiyuan Liu (Tsinghua University) [<a href="https://github.com/thunlp/Bert2Tag">Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1191781706832867328">November 5th, 2019</a></td>
<td>0.321, <strong>0.361</strong>, 0.314</td>
</tr>
<tr>
<td><strong>BERT-ChunkKPE (base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.340, <strong>0.355</strong>, 0.311</td>
</tr>
<tr>
<td><strong>SpanBERT-SpanKPE (base)</strong>Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.329, <strong>0.351</strong>, 0.304</td>
</tr>
<tr>
<td><strong>RoBERTa-SpanKPE (base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.330, <strong>0.350</strong>, 0.305</td>
</tr>
<tr>
<td><strong>LLbeBack</strong> Rodrigo Nogueira (Epistemic AI), Jimmy Lin (University of Waterloo)</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1196862630738006018">November 19th, 2019</a></td>
<td>0.349, <strong>0.341</strong>, 0.246</td>
</tr>
<tr>
<td><strong>BERT-SpanKPE (base)</strong> Si Sun(1), Chenyan Xiong(2), Zhenghao Liu(3), Zhiyuan Liu(4), Jie Bao(5) - Tsinghua University(1,3,4,5), MSR AI(2)- <a href='https://arxiv.org/pdf/2004.13639.pdf'>[Sun et al '20]</a> and [<a href = 'https://github.com/thunlp/BERT-KPE'>Code</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 6th, 2020</a></td>
<td>0.317, <strong>0.332</strong>, 0.289</td>
</tr>
<tr>
<td><strong>Baseline finetuned on Bing Queries</strong> MSMARCO Team [<a href="https://www.aclweb.org/anthology/D19-1521/ ">Xiong, et al. '19</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1186755348188676096">October 19th, 2019</a></td>
<td>0.267, <strong>0.292</strong>, 0.209</td>
</tr>
<tr>
<td><strong>Baseline</strong> MSMARCO Team [<a href="https://www.aclweb.org/anthology/D19-1521/ ">Xiong, et al. '19</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1186755348188676096">October 19th, 2019</a></td>
<td>0.244, <strong>0.277</strong>, 0.198</td>
</tr>
</tbody>
</table>
</div>
</div>
</div>
</div>
</section>
<section id="qna" class="bg-light">
<div class="container">
<div class="row">
<div class="col-md-12 mx-auto">
<h2>Question Answering and Natural Langauge Generation: RETIRED(12/01/2016-10/30/2020)</h2>
<p>The original focus of MSMARCO was to provide a corpus for training and testing systems which given a real domain user query systems would then provide the most likley candidate answer and do so in language which was natural and conversational.</p>
<p>This data comes in three tasks/forms: Original QnA dataset(v1.1), Question Answering(v2.1), Natural Language Generation(v2.1). The original question answering datset featured 100,000 examples and was released in 2016. Leaderboard is now closed but data is availible below.</p>
<p>The current competitive tasks are Question Answering and Natural Language Generation. Question Answering features over 1,000,000 queries and is much like the original QnA dataset but bigger and with higher quality. The Natural Language Generation dataset features 180,000 examples and builds upon the QnA dataset to deliver answers that could be spoken by a smart speaker.</p>
<h4>Tasks</h4>
<ol>
<li>QnA(v1.1 now closed):Given a query and 10 candidate passages select the most relvant one and use it to answer the question.</li>
<li>QnA(v2.1):Given a query and 10 candidate passages select the most relvant one and use it to answer the question.</li>
<li>NLGEN(v2.1):Given a query and 10 candidate passages select the most relvant one and use it to answer the question. Provide your answer in a way in which it could be read from a smart speaker and make sense without any additional context.</li>
</ol>
<h4>Relevant Links</h4>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://arxiv.org/abs/1611.09268">Paper</a>
<a style="background-color:#000000;border-color:#000000;" class="btn btn-primary btn-md mb-4 mr-2" href="https://github.com/microsoft/MSMARCO-Question-Answering">Question Answering Github Repo</a>
<h3>Dataset Download links</h3>
<div id="qnadataset"> </div>
<h4>Question Answering Task: RETIRED(03/01/2018-10/30/2020) Leaderboard</h4>
<div class="table-responsive">
<table class="table table-striped" id='qnav2table'>
<thead>
<tr>
<th>Rank</th>
<th class="col-md-5">Model</th>
<th class="col-md-2">Submission Date</th>
<th class="col-md-1">Rouge-L</th>
<th class="col-md-1">Bleu-1</th>
</tr>
</thead>
<tbody>
<tr>
<td><strong>Multi-doc Enriched BERT </strong>Ming Yan of Alibaba Damo NLP</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1108072710259634176">June 20th, 2019</a></td>
<td>0.540</td>
<td>0.565</td>
</tr>
<tr>
<td><strong>Human Performance</strong></td>
<td>April 23th, 2018</td>
<td>0.539</td>
<td>0.485</td>
</tr>
<td><strong>BERT Encoded T-Net</strong> Y. Zhang, C. Wang, X.L. Chen</td>
<td><a href="https://twitter.com/MSMarcoAI">August 5th, 2019</a></td>
<td>0.526</td>
<td>0.539</td>
</tr>
<tr>
<td><strong>Selector+Combine-Content-Generator QA Model</strong> Shengjie Qian of Caiyun xiaoyi AI and BUPT</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1108072710259634176">March 19th, 2019</a></td>
<td>0.525</td>
<td>0.544</td>
</tr>
<tr>
<td><strong>LM+Generator</strong> Alibaba Damo NLP </td>
<td><a href="https://twitter.com/MSMarcoAI/">November 25th,2019</a> </td>
<td>0.522</td>
<td>0.516</td>
</tr>
<tr>
<td><strong>Masque Q&A Style</strong> NTT Media Intelligence Laboratories [<a href="https://arxiv.org/abs/1901.02262">Nishida et al. '19</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1037765491140452352">January 3rd, 2019</a></td>
<td>0.522</td>
<td>0.437</td>
</tr>
<tr>
<td><strong>Deep Cascade QA</strong> Ming Yan of Alibaba Damo NLP [<a href="https://arxiv.org/abs/1811.11374">Yan et al. '18</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1055570991458410497">December 12th, 2018</a> </td>
<td>0.520</td>
<td>0.546</td>
</tr>
<tr>
<td><strong>Unnamed</strong> anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/">December 9th,2019</a> </td>
<td>0.518</td>
<td>0.507</td>
</tr>
<tr>
<td><strong>PALM</strong> Alibaba Damo NLP </td>
<td><a href="https://twitter.com/MSMarcoAI/">December 9th,2019</a> </td>
<td>0.518</td>
<td>0.507</td>
</tr>
<tr>
<td><strong>VNET</strong> Baidu NLP [<a href="https://arxiv.org/pdf/1805.02220.pdf">Wang et al. '18</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1063472881940815874">November 8th, 2018</a></td>
<td>0.516</td>
<td>0.543</td>
</tr>
<tr>
<td><strong>LNET</strong> S.L. Liu of NEUKG </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1248087073153679360?s=20">April 8th, 2020</a></td>
<td>0.514</td>
<td>0.553</td>
</tr>
<tr>
<td><strong>MultiLM QnA Model</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">December 2nd, 2019</a></td>
<td>0.514</td>
<td>0.498</td>
</tr>
<tr>
<td><strong>LNET</strong>S.L. Liu of NEUKG</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1242155084340809731?s=20">March 23rd,2020</a></td>
<td>0.506</td>
<td>0.542</td>
</tr>
<tr>
<td><strong>BERT Encoded T-NET</strong> Y. Zhang, C. Wang, X.L. Chen</td>
<td><a href="https://twitter.com/MSMarcoAI">July 12th, 2019</a></td>
<td>0.506</td>
<td>0.525</td>
</tr>
<tr>
<td><strong>MultiLM QnA Model</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">December 5th, 2019</a></td>
<td>0.499</td>
<td>0.430</td>
</tr>
<tr>
<td><strong>BERT+ Multi-Pointer-Generator</strong> Tongjun Li of the ColorfulClouds Tech and BUPT </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1079864997323321344">June 11th, 2019</a> </td>
<td>0.498</td>
<td>0.525</td>
</tr>
<tr>
<td><strong>Selector+Combine-Content-Generator NL Model</strong> Shengjie Qian of Caiyun xiaoyi AI and BUPT</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1100265573768933376">March 11th, 2019</a></td>
<td>0.496</td>
<td>0.535</td>
</tr>
<tr>
<td><strong>REAG</strong> Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">March 27th, 2020</a></td>
<td>0.495</td>
<td>0.500</td>
</tr>
<tr>
<td><strong>CompLM</strong> Alibaba Damo NLP</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1201632732473769984">December 2nd, 2019</a></td>
<td>0.495</td>
<td>0.516</td>
</tr>
<tr>
<td><strong>LM+Generator</strong> anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/">November 21st,2019</a> </td>
<td>0.494</td>
<td>0.529</td>
</tr>
<tr>
<td><strong>PALM</strong> Alibaba Damo NLP </td>
<td><a href="https://twitter.com/MSMarcoAI/">December 9th,2019</a> </td>
<td>0.492</td>
<td>0.510</td>
</tr>
<tr>
<td><strong>anonymous</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">December 16th,2019</a> </td>
<td>0.492</td>
<td>0.499</td>
</tr>
<tr>
<td><strong>LNET</strong> S.L. Liu of the NEUKG </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1196863805554192384">Nov 19th, 2019</a> </td>
<td>0.491</td>
<td>0.530</td>
</tr>
<tr>
<td><strong>BERT+ Multi-Pointer-Generator</strong> Tongjun Li of the ColorfulClouds Tech and BUPT </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1079864997323321344">May 21st, 2019</a> </td>
<td>0.491</td>
<td>0.520</td>
</tr>
<tr>
<td><strong>MUSST-NLG</strong> Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">May 15th, 2020</a></td>
<td>0.490</td>
<td>0.516</td>
</tr>
<tr>
<td><strong>CompLM</strong> Alibaba Damo NLP</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1201632732473769984">December 3rd, 2019</a></td>
<td>0.490</td>
<td>0.502</td>
</tr>
<tr>
<td><strong>Masque NLGEN Style</strong> NTT Media Intelligence Laboratories [<a href="https://arxiv.org/abs/1901.02262">Nishida et al. '19</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1037765491140452352">January 3rd, 2019</a></td>
<td>0.489</td>
<td>0.488</td>
</tr>
<tr>
<td><strong>roberta_T_tlcd_18k </strong> Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">May 14th, 2020</a></td>
<td>0.483</td>
<td>0.516</td>
</tr>
<tr>
<td><strong>Communicating BERT </strong> Xuan Liang of RIDLL from the University of Technology Sydney</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1180233411245436928">October 4th, 2019</a></td>
<td>0.483</td>
<td>0.506</td>
</tr>
<tr>
<td><strong>MDC-Generator</strong> Ssk-nlp</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1253391936976650240?s=20">April 23rd, 2020</a></td>
<td>0.482</td>
<td>0.516</td>
</tr>
<tr>
<td><strong>MultiLM NLGen Model</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">December 2nd, 2019</a></td>
<td>0.482</td>
<td>0.514</td>
</tr>
<tr>
<td><strong>LM+Generator</strong> anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/">November 19th,2019</a> </td>
<td>0.478</td>
<td>0.481</td>
</tr>
<tr>
<td><strong>MultiLM NLGen Model</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">December 5th, 2019</a></td>
<td>0.475</td>
<td>0.479</td>
</tr>
<tr>
<td><strong>BERT + Transfer</strong> anonymous</td>
<td><a href="https://twitter.com/">October 16th, 2019</a></td>
<td>0.474</td>
<td>0.499</td>
</tr>
<tr>
<td><strong>Bert Based Multi-task</strong>ZhangY & WangC</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1144116547176169472">June 26th, 2019</a></td>
<td>0.471</td>
<td>0.512</td>
</tr>
<tr>
<td><strong>T-RoBERTa-wf-BERTbaseA-120k</strong> Anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/">February 13th, 2020</a></td>
<td>0.471</td>
<td>0.483</td>
</tr>
<tr>
<td><strong>BERT-SS-K1-100k </strong>Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">January 26th, 2020</a></td>
<td>0.470</td>
<td>0.493</td>
</tr>
<tr>
<td><strong>T-RoBERTa-wf-BERTbaseA-80k</strong> Anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/">February 21st, 2020</a></td>
<td>0.468</td>
<td>0.500</td>
</tr>
<tr>
<td><strong>Multi-passage QA Model </strong>SudaNLP</td>
<td><a href="https://twitter.com/MSMarcoAI/">October 21st, 2020</a></td>
<td>0.466</td>
<td>0.508</td>
</tr>
<tr>
<td><strong>BERT-SS-K1-100k </strong>Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">February 2nd, 2020</a></td>
<td>0.464</td>
<td>0.485</td>
</tr>
<tr>
<td><strong>BERT-RGLM </strong>Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">April 22nd, 2020</a></td>
<td>0.457</td>
<td>0.479</td>
</tr>
<tr>
<td><strong>REAG </strong>Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">May 28th, 2020</a></td>
<td>0.456</td>
<td>0.449</td>
</tr>
<tr>
<td><strong>SNET + CES2S</strong> Bo Shao of SYSU University</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1021861819592335360">July 24th, 2018</a></td>
<td>0.450</td>
<td>0.464</td>
</tr>
<tr>
<td><strong>ranking+nlg</strong> anonymous</td>
<td><a href="https://twitter.com/">October 9th, 2019</a></td>
<td>0.449</td>
<td>0.468</td>
</tr>
<tr>
<td><strong>ranker-reader</strong> RCZoo of UCAS</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1128738595215298560">May 15th, 2019</a></td>
<td>0.441</td>
<td>0.371</td>
</tr>
<tr>
<td><strong>Extraction-net</strong> zlsh80826</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1024382804380057601">October 20th, 2018</a></td>
<td>0.437</td>
<td>0.444</td>
</tr>
<tr>
<td><strong>SNET</strong> JY Zhao</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1001601302885879808">August 30th, 2018</a></td>
<td>0.436</td>
<td>0.463</td>
</tr>
<tr>
<td><strong>BIDAF+ELMo+SofterMax</strong> Wang Changbao</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1063472543372378112">November 16th, 2018</a></td>
<td>0.436</td>
<td>0.459</td>
</tr>
<tr>
<td><strong>ranking+nlg</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">August 12th, 2019</a></td>
<td>0.434</td>
<td>0.411</td>
</tr>
<tr>
<td><strong>DNET</strong> QA Geeks</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1004425343602577408">August 1st, 2018</a></td>
<td>0.432</td>
<td>0.479</td>
</tr>
<tr>
<td><strong>T-RoBERTa-wf-BERTbaseA-120k</strong> Anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/">February 13th, 2020</a></td>
<td>0.431</td>
<td>0.424</td>
</tr>
<tr>
<td><strong>KIGN-QA</strong> Chenliang Li</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1067188900894957568">April 22nd, 2019</a></td>
<td>0.429</td>
<td>0.404</td>
</tr>
<tr>
<td><strong>MaRCo-da-GAAMA</strong> IBM Research AI Multilingual NLP Group</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1247606676049416192?s=20">April 7th, 2020</a></td>
<td>0.426</td>
<td>0.462</td>
</tr>
<tr>
<td><strong>Reader-Writer</strong> Microsoft Business Applications Group AI Research </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1037764974410588160">September 16th, 2018</a></td>
<td>0.421</td>
<td>0.436</td>
</tr>
<tr>
<td><strong>Masque2 (single / NLG Style)</strong> NTT Media Intelligence Laboratories </td>
<td><a href="https://twitter.com/MSMarcoAI/status/">October 22nd, 2020</a></td>
<td>0.419</td>
<td>0.469</td>
</tr>
<tr>
<td><strong>BERT+Multi-Loss</strong> S.L. Liu of NEUKG </td>
<td><a href="https://twitter.com/MSMarcoAI">November 4th, 2019</a></td>
<td>0.413</td>
<td>0.422</td>
</tr>
<tr>
<td><strong>REAG(based on PALM)</strong>anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">June 1st,2020</a></td>
<td>0.410</td>
<td>0.430</td>
</tr>
<tr>
<td><strong>RGLM</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">May 5th, 2020</a></td>
<td>0.406</td>
<td>0.455</td>
</tr>
<tr>
<td><strong>SNET+seq2seq</strong> Yihan Ni of the CAS Key Lab of Web Data Science and Technology, ICT, CAS </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1001627093438902272">June 1st, 2018</a></td>
<td>0.398</td>
<td>0.423</td>
</tr>
<tr>
<td><strong>SSK3+BERTBaseAnswerGenerator</strong> anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/status/">Jan 21st, 2020</a></td>
<td>0.391</td>
<td>0.413</td>
</tr>
<tr>
<td><strong>MP-MRC BERT</strong> H.Y. Zhang </td>
<td><a href="https://twitter.com/MSMarcoAI/status/">Aug 27th, 2020</a></td>
<td>0.389</td>
<td>0.410</td>
</tr>
<tr>
<td><strong>MP-MRC BERT-base</strong> H.Y. Zhang </td>
<td><a href="https://twitter.com/MSMarcoAI/status/">Sep 4th, 2020</a></td>
<td>0.388</td>
<td>0.411</td>
</tr>
<tr>
<td><strong>MUSST</strong> anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/status/">March 31st, 2020</a></td>
<td>0.376</td>
<td>0.405</td>
</tr>
<tr>
<td><strong>Anonymous</strong> anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/status/">October 12th, 2020</a></td>
<td>0.359</td>
<td>0.409</td>
</tr>
<tr>
<td><strong>fj-net(single)</strong> yzm nlp group </td>
<td><a href="https://twitter.com/MSMarcoAI/status/">August 3rd, 2020</a></td>
<td>0.343</td>
<td>0.409</td>
</tr>
<tr>
<td><strong>MNet-Base(Single) NLGEN</strong> fuii of iDW </td>
<td><a href="https://twitter.com/MSMarcoAI/status/">July 8th, 2020</a></td>
<td>0.337</td>
<td>0.405</td>
</tr>
<tr>
<td><strong>fj-reader(single)</strong> yzm nlp group </td>
<td><a href="https://twitter.com/MSMarcoAI/status/">July 28th, 2020</a></td>
<td>0.336</td>
<td>0.404</td>
</tr>
<tr>
<td><strong>Generation with latent retrieval per answer</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">May 11th, 2020</a></td>
<td>0.335</td>
<td>0.290</td>
</tr>
<tr>
<td><strong>MDCG-Base</strong> ssk-nlp</td>
<td><a href="https://twitter.com/MSMarcoAI">June 8th, 2020</a></td>
<td>0.334</td>
<td>0.398</td>
</tr>
<tr>
<td><strong>MUSST-NLG</strong> Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">June 2nd, 2020</a></td>
<td>0.334</td>
<td>0.388</td>
</tr>
<tr>
<td><strong>MDCC-Base</strong> ssk-nlp</td>
<td><a href="https://twitter.com/MSMarcoAI">June 10th, 2020</a></td>
<td>0.333</td>
<td>0.400</td>
</tr>
<tr>
<td><strong>Generation with latent retrieval Baseline 2</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">May 11th, 2020</a></td>
<td>0.331</td>
<td>0.307</td>
</tr>
<tr>
<td><strong>MDCC</strong> ssk-nlp</td>
<td><a href="https://twitter.com/MSMarcoAI">June 10th, 2020</a></td>
<td>0.328</td>
<td>0.391</td>
</tr>
<tr>
<td><strong>Generation with latent retrieval Baseline 1</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">May 11th, 2020</a></td>
<td>0.305</td>
<td>0.275</td>
</tr>
<tr>
<td><strong>MultiTask+DataAug+Unlikelihood </strong> UvA</td>
<td><a href="https://twitter.com/MSMarcoAI/status/">June 3rd, 2020</a></td>
<td>0.300</td>
<td>0.332</td>
</tr>
<tr>
<td><strong>MUSST-QA </strong> Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/status/">June 1st, 2020</a></td>
<td>0.298</td>
<td>0.354</td>
</tr>
<tr>
<td><strong>lightNLP+BiDAF</strong> Enliple AI</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1091935023442219018">February 1st, 2019</a></td>
<td>0.298</td>
<td>0.156</td>
</tr>
<tr>
<td><strong>Pretrained seq2seq model</strong> BDEG</td>
<td><a href="https://twitter.com/MSMarcoAI/">September 10th, 2020</a></td>
<td>0.290</td>
<td>0.331</td>
</tr>
<tr>
<td><strong>roberta_T_tlx_90k</strong> Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">July 29th, 2020</a></td>
<td>0.286</td>
<td>0.327</td>
</tr>
<tr>
<td><strong>BIDAF+seq2seq</strong> Yihan Ni of the CAS Key Lab of Web Data Science and Technology, ICT, CAS </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1001627093438902272">May 29th, 2018</a></td>
<td>0.276</td>
<td>0.288</td>
</tr>
<tr>
<td><strong>BiDaF Baseline(Implemented By MSMARCO Team)</strong><br>Allen Institute for AI & University of Washington [<a href="https://arxiv.org/abs/1611.01603">Seo et al. '16</a>]</td>
<td>April 23th, 2018</td>
<td>0.240</td>
<td>0.106</td>
</tr>
<tr>
<td><strong>TrioNLP + BiDAF</strong> Trio.AI of the CCNU</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1044091698442326021">September 23rd, 2018</a></td>
<td>0.205</td>
<td>0.232</td>
</tr>
<tr>
<td><strong>BiDAF + LSTM</strong> Meefly</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1085088144083566592">January 15th,2019</a></td>
<td>0.153</td>
<td>0.120</td>
</tr>
</tbody>
</table>
</div>
<br>
<h4>Natural Language Generation Task:RETIRED(03/01/2018-10/30/2020)</h4>
<div class="table-responsive">
<table class="table table-striped" id='nlgentable'>
<thead>
<tr>
<th>Rank</th>
<th class="col-md-5">Model</th>
<th class="col-md-2">Submission Date</th>
<th class="col-md-1">Rouge-L</th>
<th class="col-md-1">Bleu-1</th>
</tr>
</thead>
<tbody>
<tr>
<td><strong>Human Performance</strong></td>
<td>April 23th, 2018</td>
<td>0.632</td>
<td>0.530</td>
</tr>
<tr>
<td><strong>PALM</strong> Alibaba Damo NLP</td>
<td><a href="https://twitter.com/MSMarcoAI/">December 16th,2019</a> </td>
<td>0.498</td>
<td>0.499</td>
</tr>
<tr>
<td><strong>REAG</strong> Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">March 27th, 2020</a></td>
<td>0.498</td>
<td>0.497</td>
</tr>
<tr>
<td><strong>Masque NLGEN Style</strong> NTT Media Intelligence Laboratories [<a href="https://arxiv.org/abs/1901.02262">Nishida et al. '19</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1037765491140452352">January 3rd, 2019</a></td>
<td>0.496</td>
<td>0.501</td>
</tr>
<tr>
<td><strong>CompLM</strong> Alibaba Damo NLP</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1201632732473769984">December 3rd, 2019</a></td>
<td>0.496</td>
<td>0.489</td>
</tr>
<tr>
<td><strong>PALM</strong> Alibaba Damo NLP </td>
<td><a href="https://twitter.com/MSMarcoAI/">December 9th,2019</a> </td>
<td>0.496</td>
<td>0.484</td>
</tr>
<tr>
<td><strong>BERT+ Multi-Pointer-Generator</strong> Tongjun Li of the ColorfulClouds Tech and BUPT </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1079864997323321344">June 11th,2019</a> </td>
<td>0.495</td>
<td>0.476</td>
</tr>
<tr>
<td><strong>CompLM</strong> Alibaba Damo NLP </td>
<td><a href="https://twitter.com/MSMarcoAI/">November 19th,2019</a> </td>
<td>0.495</td>
<td>0.470</td>
</tr>
<tr>
<td><strong>CompLM</strong> Alibaba Damo NLP</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1201632732473769984">December 2nd, 2019</a></td>
<td>0.493</td>
<td>0.475</td>
</tr>
<tr>
<td><strong>BERT+ Multi-Pointer-Generator</strong> Tongjun Li of the ColorfulClouds Tech and BUPT </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1079864997323321344">May 21st,2019</a> </td>
<td>0.491</td>
<td>0.474</td>
</tr>
<tr>
<td><strong>CompLM</strong> Alibaba Damo NLP</td>
<td><a href="https://twitter.com/MSMarcoAI/">November 19th,2019</a> </td>
<td>0.488</td>
<td>0.485</td>
</tr>
<tr>
<td><strong>roberta_T_tlcd_18k </strong> Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">May 14th, 2020</a></td>
<td>0.487</td>
<td>0.468</td>
</tr>
<tr>
<td><strong>BERT+ Multi-Pointer-Generator</strong> Tongjun Li of the ColorfulClouds Tech and BUPT </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1079864997323321344">March 26th,2019</a> </td>
<td>0.487</td>
<td>0.465</td>
</tr>
<tr>
<td><strong>Selector+Combine-Content-Generator NLGEN Model</strong> Shengjie Qian of Caiyun xiaoyi AI and BUPT</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1100265573768933376">March 11th, 2019</a></td>
<td>0.487</td>
<td>0.449</td>
</tr>
<tr>
<td><strong>VNET</strong> Baidu NLP [<a href="https://arxiv.org/pdf/1805.02220.pdf">Wang et al. '18</a>]</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1009163361773240320">November 8th, 2018</a></td>
<td>0.484</td>
<td>0.468</td>
</tr>
<tr>
<td><strong>BERT+ Multi-Pointer-Generator (Single) </strong> Tongjun Li of the ColorfulClouds Tech and BUPT </td>
<td><a href="https://twitter.com/MSMarcoAI/status/1079864997323321344">March 19th,2019</a> </td>
<td>0.484</td>
<td>0.459</td>
</tr>
<tr>
<td><strong>Communicating BERT </strong> Xuan Liang of RIDLL from the University of Technology Sydney</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1180233411245436928">October 4th, 2019</a></td>
<td>0.483</td>
<td>0.472</td>
</tr>
<tr>
<td><strong>MultiLM NLGen Model</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">December 2nd, 2019</a></td>
<td>0.483</td>
<td>0.461</td>
</tr>
<tr>
<td><strong>ranking+nlg</strong> anonymous</td>
<td><a href="https://twitter.com/">October 9th, 2019</a></td>
<td>0.481</td>
<td>0.468</td>
</tr>
<tr>
<td><strong>MUSST-NLG</strong> Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">May 15th, 2020</a></td>
<td>0.480</td>
<td>0.458</td>
</tr>
<tr>
<td><strong>MultiLM NLGen Model</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">December 5th, 2019</a></td>
<td>0.478</td>
<td>0.481</td>
</tr>
<tr>
<td><strong>BERT-RGLM </strong>Anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">April 22nd, 2020</a></td>
<td>0.470</td>
<td>0.452</td>
</tr>
<tr>
<td><strong>BERT-SS-K1-100k</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">January 26th, 2020</a></td>
<td>0.470</td>
<td>0.437</td>
</tr>
<tr>
<td><strong>MDC-Generator</strong> Ssk-nlp</td>
<td><a href="https://twitter.com/MSMarcoAI/status/1253391936976650240?s=20">April 23rd, 2020</a></td>
<td>0.466</td>
<td>0.446</td>
</tr>
<tr>
<td><strong>BERT-SS-K1-100k</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI">February 2nd, 2020</a></td>
<td>0.465</td>
<td>0.427</td>
</tr>
<tr>
<td><strong>T-RoBERTa-wf-BERTbaseA-120k</strong> Anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/">February 17th, 2020</a></td>
<td>0.464</td>
<td>0.420</td>
</tr>
<tr>
<td><strong>T-RoBERTa-wf-BERTbaseA-80k</strong> Anonymous </td>
<td><a href="https://twitter.com/MSMarcoAI/">February 21st, 2020</a></td>
<td>0.463</td>
<td>0.438</td>
</tr>
<tr>
<td><strong>ranking+nlg</strong> anonymous</td>
<td><a href="https://twitter.com/">October 9th, 2019</a></td>
<td>0.462</td>
<td>0.451</td>
</tr>
<tr>
<td><strong>PM-MUG-1</strong> anonymous</td>
<td><a href="https://twitter.com/MSMarcoAI/">May 20th, 2020</a></td>
<td>0.453</td>
<td>0.441</td>
</tr>
<tr>