zeroshot commited on
Commit
6ca6180
1 Parent(s): 5def9b9

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +535 -0
README.md CHANGED
@@ -50,6 +50,62 @@ model-index:
50
  value: 45.494
51
  - type: f1
52
  value: 44.917953161904805
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
53
  - task:
54
  type: Classification
55
  dataset:
@@ -63,6 +119,28 @@ model-index:
63
  value: 84.29545454545455
64
  - type: f1
65
  value: 84.26780483160312
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
66
  - task:
67
  type: Classification
68
  dataset:
@@ -143,6 +221,342 @@ model-index:
143
  value: 77.5353059852051
144
  - type: f1
145
  value: 77.42427561340143
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
146
  - task:
147
  type: Classification
148
  dataset:
@@ -171,6 +585,127 @@ model-index:
171
  value: 59.49349179400113
172
  - type: f1
173
  value: 59.815392064510775
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
174
  ---
175
  This is the sparsified ONNX variant of the [bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5) embeddings model created with [DeepSparse Optimum](https://github.com/neuralmagic/optimum-deepsparse) for ONNX export/inference pipeline and Neural Magic's [Sparsify](https://github.com/neuralmagic/sparsify) for one-shot quantization (INT8) and unstructured pruning (50%).
176
 
 
50
  value: 45.494
51
  - type: f1
52
  value: 44.917953161904805
53
+ - task:
54
+ type: Clustering
55
+ dataset:
56
+ type: mteb/arxiv-clustering-p2p
57
+ name: MTEB ArxivClusteringP2P
58
+ config: default
59
+ split: test
60
+ revision: a122ad7f3f0291bf49cc6f4d32aa80929df69d5d
61
+ metrics:
62
+ - type: v_measure
63
+ value: 46.50495921726095
64
+ - task:
65
+ type: Clustering
66
+ dataset:
67
+ type: mteb/arxiv-clustering-s2s
68
+ name: MTEB ArxivClusteringS2S
69
+ config: default
70
+ split: test
71
+ revision: f910caf1a6075f7329cdf8c1a6135696f37dbd53
72
+ metrics:
73
+ - type: v_measure
74
+ value: 40.080055890804836
75
+ - task:
76
+ type: Reranking
77
+ dataset:
78
+ type: mteb/askubuntudupquestions-reranking
79
+ name: MTEB AskUbuntuDupQuestions
80
+ config: default
81
+ split: test
82
+ revision: 2000358ca161889fa9c082cb41daa8dcfb161a54
83
+ metrics:
84
+ - type: map
85
+ value: 60.22880715757138
86
+ - type: mrr
87
+ value: 73.11227630479708
88
+ - task:
89
+ type: STS
90
+ dataset:
91
+ type: mteb/biosses-sts
92
+ name: MTEB BIOSSES
93
+ config: default
94
+ split: test
95
+ revision: d3fb88f8f02e40887cd149695127462bbcf29b4a
96
+ metrics:
97
+ - type: cos_sim_pearson
98
+ value: 86.9542549153515
99
+ - type: cos_sim_spearman
100
+ value: 83.93865958725257
101
+ - type: euclidean_pearson
102
+ value: 86.00372707912037
103
+ - type: euclidean_spearman
104
+ value: 84.97302050526537
105
+ - type: manhattan_pearson
106
+ value: 85.63207676453459
107
+ - type: manhattan_spearman
108
+ value: 84.82542678079645
109
  - task:
110
  type: Classification
111
  dataset:
 
119
  value: 84.29545454545455
120
  - type: f1
121
  value: 84.26780483160312
122
+ - task:
123
+ type: Clustering
124
+ dataset:
125
+ type: mteb/biorxiv-clustering-p2p
126
+ name: MTEB BiorxivClusteringP2P
127
+ config: default
128
+ split: test
129
+ revision: 65b79d1d13f80053f67aca9498d9402c2d9f1f40
130
+ metrics:
131
+ - type: v_measure
132
+ value: 36.78678386185847
133
+ - task:
134
+ type: Clustering
135
+ dataset:
136
+ type: mteb/biorxiv-clustering-s2s
137
+ name: MTEB BiorxivClusteringS2S
138
+ config: default
139
+ split: test
140
+ revision: 258694dd0231531bc1fd9de6ceb52a0853c6d908
141
+ metrics:
142
+ - type: v_measure
143
+ value: 34.42462869304013
144
  - task:
145
  type: Classification
146
  dataset:
 
221
  value: 77.5353059852051
222
  - type: f1
223
  value: 77.42427561340143
224
+ - task:
225
+ type: Clustering
226
+ dataset:
227
+ type: mteb/medrxiv-clustering-p2p
228
+ name: MTEB MedrxivClusteringP2P
229
+ config: default
230
+ split: test
231
+ revision: e7a26af6f3ae46b30dde8737f02c07b1505bcc73
232
+ metrics:
233
+ - type: v_measure
234
+ value: 32.00163251745748
235
+ - task:
236
+ type: Clustering
237
+ dataset:
238
+ type: mteb/medrxiv-clustering-s2s
239
+ name: MTEB MedrxivClusteringS2S
240
+ config: default
241
+ split: test
242
+ revision: 35191c8c0dca72d8ff3efcd72aa802307d469663
243
+ metrics:
244
+ - type: v_measure
245
+ value: 30.37879992380756
246
+ - task:
247
+ type: Clustering
248
+ dataset:
249
+ type: mteb/reddit-clustering
250
+ name: MTEB RedditClustering
251
+ config: default
252
+ split: test
253
+ revision: 24640382cdbf8abc73003fb0fa6d111a705499eb
254
+ metrics:
255
+ - type: v_measure
256
+ value: 50.99679402527969
257
+ - task:
258
+ type: Clustering
259
+ dataset:
260
+ type: mteb/reddit-clustering-p2p
261
+ name: MTEB RedditClusteringP2P
262
+ config: default
263
+ split: test
264
+ revision: 282350215ef01743dc01b456c7f5241fa8937f16
265
+ metrics:
266
+ - type: v_measure
267
+ value: 59.28024721612242
268
+ - task:
269
+ type: STS
270
+ dataset:
271
+ type: mteb/sickr-sts
272
+ name: MTEB SICK-R
273
+ config: default
274
+ split: test
275
+ revision: a6ea5a8cab320b040a23452cc28066d9beae2cee
276
+ metrics:
277
+ - type: cos_sim_pearson
278
+ value: 84.54645068673153
279
+ - type: cos_sim_spearman
280
+ value: 78.64401947043316
281
+ - type: euclidean_pearson
282
+ value: 82.36873285307261
283
+ - type: euclidean_spearman
284
+ value: 78.57406974337181
285
+ - type: manhattan_pearson
286
+ value: 82.33000263843067
287
+ - type: manhattan_spearman
288
+ value: 78.51127629983256
289
+ - task:
290
+ type: STS
291
+ dataset:
292
+ type: mteb/sts12-sts
293
+ name: MTEB STS12
294
+ config: default
295
+ split: test
296
+ revision: a0d554a64d88156834ff5ae9920b964011b16384
297
+ metrics:
298
+ - type: cos_sim_pearson
299
+ value: 83.3001843293691
300
+ - type: cos_sim_spearman
301
+ value: 74.87989254109124
302
+ - type: euclidean_pearson
303
+ value: 80.88523322810525
304
+ - type: euclidean_spearman
305
+ value: 75.6469299496058
306
+ - type: manhattan_pearson
307
+ value: 80.8921104008781
308
+ - type: manhattan_spearman
309
+ value: 75.65942956132456
310
+ - task:
311
+ type: STS
312
+ dataset:
313
+ type: mteb/sts13-sts
314
+ name: MTEB STS13
315
+ config: default
316
+ split: test
317
+ revision: 7e90230a92c190f1bf69ae9002b8cea547a64cca
318
+ metrics:
319
+ - type: cos_sim_pearson
320
+ value: 82.40319855455617
321
+ - type: cos_sim_spearman
322
+ value: 83.63807375781141
323
+ - type: euclidean_pearson
324
+ value: 83.28557187260904
325
+ - type: euclidean_spearman
326
+ value: 83.65223617817439
327
+ - type: manhattan_pearson
328
+ value: 83.30411918680012
329
+ - type: manhattan_spearman
330
+ value: 83.69204806663276
331
+ - task:
332
+ type: STS
333
+ dataset:
334
+ type: mteb/sts14-sts
335
+ name: MTEB STS14
336
+ config: default
337
+ split: test
338
+ revision: 6031580fec1f6af667f0bd2da0a551cf4f0b2375
339
+ metrics:
340
+ - type: cos_sim_pearson
341
+ value: 83.08942420708404
342
+ - type: cos_sim_spearman
343
+ value: 80.39991846857053
344
+ - type: euclidean_pearson
345
+ value: 82.68275416568997
346
+ - type: euclidean_spearman
347
+ value: 80.49626214786178
348
+ - type: manhattan_pearson
349
+ value: 82.62993414444689
350
+ - type: manhattan_spearman
351
+ value: 80.44148684748403
352
+ - task:
353
+ type: STS
354
+ dataset:
355
+ type: mteb/sts15-sts
356
+ name: MTEB STS15
357
+ config: default
358
+ split: test
359
+ revision: ae752c7c21bf194d8b67fd573edf7ae58183cbe3
360
+ metrics:
361
+ - type: cos_sim_pearson
362
+ value: 86.70365000096972
363
+ - type: cos_sim_spearman
364
+ value: 88.00515486253518
365
+ - type: euclidean_pearson
366
+ value: 87.65142168651604
367
+ - type: euclidean_spearman
368
+ value: 88.05834854642737
369
+ - type: manhattan_pearson
370
+ value: 87.59548659661925
371
+ - type: manhattan_spearman
372
+ value: 88.00573237576926
373
+ - task:
374
+ type: STS
375
+ dataset:
376
+ type: mteb/sts16-sts
377
+ name: MTEB STS16
378
+ config: default
379
+ split: test
380
+ revision: 4d8694f8f0e0100860b497b999b3dbed754a0513
381
+ metrics:
382
+ - type: cos_sim_pearson
383
+ value: 82.47886818876728
384
+ - type: cos_sim_spearman
385
+ value: 84.30874770680975
386
+ - type: euclidean_pearson
387
+ value: 83.74580951498133
388
+ - type: euclidean_spearman
389
+ value: 84.60595431454789
390
+ - type: manhattan_pearson
391
+ value: 83.74122023121615
392
+ - type: manhattan_spearman
393
+ value: 84.60549899361064
394
+ - task:
395
+ type: STS
396
+ dataset:
397
+ type: mteb/sts17-crosslingual-sts
398
+ name: MTEB STS17 (en-en)
399
+ config: en-en
400
+ split: test
401
+ revision: af5e6fb845001ecf41f4c1e033ce921939a2a68d
402
+ metrics:
403
+ - type: cos_sim_pearson
404
+ value: 87.60257252565631
405
+ - type: cos_sim_spearman
406
+ value: 88.29577246271319
407
+ - type: euclidean_pearson
408
+ value: 88.25434138634807
409
+ - type: euclidean_spearman
410
+ value: 88.06678743723845
411
+ - type: manhattan_pearson
412
+ value: 88.3651048848073
413
+ - type: manhattan_spearman
414
+ value: 88.23688291108866
415
+ - task:
416
+ type: STS
417
+ dataset:
418
+ type: mteb/sts22-crosslingual-sts
419
+ name: MTEB STS22 (en)
420
+ config: en
421
+ split: test
422
+ revision: 6d1ba47164174a496b7fa5d3569dae26a6813b80
423
+ metrics:
424
+ - type: cos_sim_pearson
425
+ value: 61.666254720687206
426
+ - type: cos_sim_spearman
427
+ value: 63.83700525419119
428
+ - type: euclidean_pearson
429
+ value: 64.36325040161177
430
+ - type: euclidean_spearman
431
+ value: 63.99833771224718
432
+ - type: manhattan_pearson
433
+ value: 64.01356576965371
434
+ - type: manhattan_spearman
435
+ value: 63.7201674202641
436
+ - task:
437
+ type: STS
438
+ dataset:
439
+ type: mteb/stsbenchmark-sts
440
+ name: MTEB STSBenchmark
441
+ config: default
442
+ split: test
443
+ revision: b0fddb56ed78048fa8b90373c8a3cfc37b684831
444
+ metrics:
445
+ - type: cos_sim_pearson
446
+ value: 85.14584232139909
447
+ - type: cos_sim_spearman
448
+ value: 85.92570762612142
449
+ - type: euclidean_pearson
450
+ value: 86.34291503630607
451
+ - type: euclidean_spearman
452
+ value: 86.12670269109282
453
+ - type: manhattan_pearson
454
+ value: 86.26109450032494
455
+ - type: manhattan_spearman
456
+ value: 86.07665628498633
457
+ - task:
458
+ type: Reranking
459
+ dataset:
460
+ type: mteb/scidocs-reranking
461
+ name: MTEB SciDocsRR
462
+ config: default
463
+ split: test
464
+ revision: d3c5e1fc0b855ab6097bf1cda04dd73947d7caab
465
+ metrics:
466
+ - type: map
467
+ value: 84.46430478723548
468
+ - type: mrr
469
+ value: 95.63907044299201
470
+ - task:
471
+ type: PairClassification
472
+ dataset:
473
+ type: mteb/sprintduplicatequestions-pairclassification
474
+ name: MTEB SprintDuplicateQuestions
475
+ config: default
476
+ split: test
477
+ revision: d66bd1f72af766a5cc4b0ca5e00c162f89e8cc46
478
+ metrics:
479
+ - type: cos_sim_accuracy
480
+ value: 99.82178217821782
481
+ - type: cos_sim_ap
482
+ value: 95.49612561375889
483
+ - type: cos_sim_f1
484
+ value: 91.02691924227318
485
+ - type: cos_sim_precision
486
+ value: 90.75546719681908
487
+ - type: cos_sim_recall
488
+ value: 91.3
489
+ - type: dot_accuracy
490
+ value: 99.67821782178218
491
+ - type: dot_ap
492
+ value: 90.55740832326241
493
+ - type: dot_f1
494
+ value: 83.30765279917823
495
+ - type: dot_precision
496
+ value: 85.6388595564942
497
+ - type: dot_recall
498
+ value: 81.10000000000001
499
+ - type: euclidean_accuracy
500
+ value: 99.82475247524752
501
+ - type: euclidean_ap
502
+ value: 95.4739426775874
503
+ - type: euclidean_f1
504
+ value: 91.07413010590017
505
+ - type: euclidean_precision
506
+ value: 91.8616480162767
507
+ - type: euclidean_recall
508
+ value: 90.3
509
+ - type: manhattan_accuracy
510
+ value: 99.82376237623762
511
+ - type: manhattan_ap
512
+ value: 95.48506891694475
513
+ - type: manhattan_f1
514
+ value: 91.02822580645163
515
+ - type: manhattan_precision
516
+ value: 91.76829268292683
517
+ - type: manhattan_recall
518
+ value: 90.3
519
+ - type: max_accuracy
520
+ value: 99.82475247524752
521
+ - type: max_ap
522
+ value: 95.49612561375889
523
+ - type: max_f1
524
+ value: 91.07413010590017
525
+ - task:
526
+ type: Clustering
527
+ dataset:
528
+ type: mteb/stackexchange-clustering
529
+ name: MTEB StackExchangeClustering
530
+ config: default
531
+ split: test
532
+ revision: 6cbc1f7b2bc0622f2e39d2c77fa502909748c259
533
+ metrics:
534
+ - type: v_measure
535
+ value: 60.92486258951404
536
+ - task:
537
+ type: Clustering
538
+ dataset:
539
+ type: mteb/stackexchange-clustering-p2p
540
+ name: MTEB StackExchangeClusteringP2P
541
+ config: default
542
+ split: test
543
+ revision: 815ca46b2622cec33ccafc3735d572c266efdb44
544
+ metrics:
545
+ - type: v_measure
546
+ value: 32.97511013092965
547
+ - task:
548
+ type: Reranking
549
+ dataset:
550
+ type: mteb/stackoverflowdupquestions-reranking
551
+ name: MTEB StackOverflowDupQuestions
552
+ config: default
553
+ split: test
554
+ revision: e185fbe320c72810689fc5848eb6114e1ef5ec69
555
+ metrics:
556
+ - type: map
557
+ value: 52.31647363355174
558
+ - type: mrr
559
+ value: 53.26469792462439
560
  - task:
561
  type: Classification
562
  dataset:
 
585
  value: 59.49349179400113
586
  - type: f1
587
  value: 59.815392064510775
588
+ - task:
589
+ type: Clustering
590
+ dataset:
591
+ type: mteb/twentynewsgroups-clustering
592
+ name: MTEB TwentyNewsgroupsClustering
593
+ config: default
594
+ split: test
595
+ revision: 6125ec4e24fa026cec8a478383ee943acfbd5449
596
+ metrics:
597
+ - type: v_measure
598
+ value: 47.29662657485732
599
+ - task:
600
+ type: PairClassification
601
+ dataset:
602
+ type: mteb/twittersemeval2015-pairclassification
603
+ name: MTEB TwitterSemEval2015
604
+ config: default
605
+ split: test
606
+ revision: 70970daeab8776df92f5ea462b6173c0b46fd2d1
607
+ metrics:
608
+ - type: cos_sim_accuracy
609
+ value: 85.74834594981225
610
+ - type: cos_sim_ap
611
+ value: 72.92449226447182
612
+ - type: cos_sim_f1
613
+ value: 68.14611644433363
614
+ - type: cos_sim_precision
615
+ value: 64.59465847317419
616
+ - type: cos_sim_recall
617
+ value: 72.1108179419525
618
+ - type: dot_accuracy
619
+ value: 82.73827263515527
620
+ - type: dot_ap
621
+ value: 63.27505594570806
622
+ - type: dot_f1
623
+ value: 61.717543651265
624
+ - type: dot_precision
625
+ value: 56.12443292287751
626
+ - type: dot_recall
627
+ value: 68.54881266490766
628
+ - type: euclidean_accuracy
629
+ value: 85.90332002145796
630
+ - type: euclidean_ap
631
+ value: 73.08299660990401
632
+ - type: euclidean_f1
633
+ value: 67.9050313691721
634
+ - type: euclidean_precision
635
+ value: 63.6091265268495
636
+ - type: euclidean_recall
637
+ value: 72.82321899736148
638
+ - type: manhattan_accuracy
639
+ value: 85.87351731537224
640
+ - type: manhattan_ap
641
+ value: 73.02205874497865
642
+ - type: manhattan_f1
643
+ value: 67.87532596547871
644
+ - type: manhattan_precision
645
+ value: 64.109781843772
646
+ - type: manhattan_recall
647
+ value: 72.1108179419525
648
+ - type: max_accuracy
649
+ value: 85.90332002145796
650
+ - type: max_ap
651
+ value: 73.08299660990401
652
+ - type: max_f1
653
+ value: 68.14611644433363
654
+ - task:
655
+ type: PairClassification
656
+ dataset:
657
+ type: mteb/twitterurlcorpus-pairclassification
658
+ name: MTEB TwitterURLCorpus
659
+ config: default
660
+ split: test
661
+ revision: 8b6510b0b1fa4e4c4f879467980e9be563ec1cdf
662
+ metrics:
663
+ - type: cos_sim_accuracy
664
+ value: 88.84231769317343
665
+ - type: cos_sim_ap
666
+ value: 85.65683184516553
667
+ - type: cos_sim_f1
668
+ value: 77.60567077973222
669
+ - type: cos_sim_precision
670
+ value: 75.6563071297989
671
+ - type: cos_sim_recall
672
+ value: 79.65814598090545
673
+ - type: dot_accuracy
674
+ value: 86.85333954282609
675
+ - type: dot_ap
676
+ value: 80.79899186896125
677
+ - type: dot_f1
678
+ value: 74.15220098146928
679
+ - type: dot_precision
680
+ value: 70.70819946919961
681
+ - type: dot_recall
682
+ value: 77.94887588543271
683
+ - type: euclidean_accuracy
684
+ value: 88.77634183257655
685
+ - type: euclidean_ap
686
+ value: 85.67411484805298
687
+ - type: euclidean_f1
688
+ value: 77.61566374357423
689
+ - type: euclidean_precision
690
+ value: 76.23255123255123
691
+ - type: euclidean_recall
692
+ value: 79.04989220819218
693
+ - type: manhattan_accuracy
694
+ value: 88.79962743043428
695
+ - type: manhattan_ap
696
+ value: 85.6494795781639
697
+ - type: manhattan_f1
698
+ value: 77.54222877224805
699
+ - type: manhattan_precision
700
+ value: 76.14100185528757
701
+ - type: manhattan_recall
702
+ value: 78.99599630428088
703
+ - type: max_accuracy
704
+ value: 88.84231769317343
705
+ - type: max_ap
706
+ value: 85.67411484805298
707
+ - type: max_f1
708
+ value: 77.61566374357423
709
  ---
710
  This is the sparsified ONNX variant of the [bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5) embeddings model created with [DeepSparse Optimum](https://github.com/neuralmagic/optimum-deepsparse) for ONNX export/inference pipeline and Neural Magic's [Sparsify](https://github.com/neuralmagic/sparsify) for one-shot quantization (INT8) and unstructured pruning (50%).
711