File size: 25,265 Bytes
ef6db93
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
language:
- en
- ar
- cs
- de
- es
- fr
- it
- ja
- ko
- nl
- pt
- zh
license: apache-2.0  
library_name: transformers
tags:
- language
- granite
- embeddings
- multilingual
model-index:
- name: ibm-granite/granite-embedding-107m-multilingual
  results:
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (en)
      config: en
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.41176
    - type: ndcg_at_10
      value: 0.46682
    - type: ndcg_at_100
      value: 0.54326
    - type: ndcg_at_1000
      value: 0.56567
    - type: ndcg_at_20
      value: 0.50157
    - type: ndcg_at_3
      value: 0.41197
    - type: ndcg_at_5
      value: 0.42086
    - type: recall_at_1
      value: 0.19322
    - type: recall_at_10
      value: 0.57721
    - type: recall_at_100
      value: 0.83256
    - type: recall_at_1000
      value: 0.95511
    - type: recall_at_20
      value: 0.6757
    - type: recall_at_3
      value: 0.37171
    - type: recall_at_5
      value: 0.44695
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (ar)
      config: ar
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.55559
    - type: ndcg_at_10
      value: 0.62541
    - type: ndcg_at_100
      value: 0.67101
    - type: ndcg_at_1000
      value: 0.6805
    - type: ndcg_at_20
      value: 0.64739
    - type: ndcg_at_3
      value: 0.56439
    - type: ndcg_at_5
      value: 0.59347
    - type: recall_at_1
      value: 0.37009
    - type: recall_at_10
      value: 0.73317
    - type: recall_at_100
      value: 0.90066
    - type: recall_at_1000
      value: 0.96272
    - type: recall_at_20
      value: 0.80205
    - type: recall_at_3
      value: 0.56903
    - type: recall_at_5
      value: 0.6518
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (bn)
      config: bn
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.56691
    - type: ndcg_at_10
      value: 0.65484
    - type: ndcg_at_100
      value: 0.70142
    - type: ndcg_at_1000
      value: 0.70994
    - type: ndcg_at_20
      value: 0.67838
    - type: ndcg_at_3
      value: 0.5988
    - type: ndcg_at_5
      value: 0.62718
    - type: recall_at_1
      value: 0.3605
    - type: recall_at_10
      value: 0.76854
    - type: recall_at_100
      value: 0.9285
    - type: recall_at_1000
      value: 0.97928
    - type: recall_at_20
      value: 0.83667
    - type: recall_at_3
      value: 0.61596
    - type: recall_at_5
      value: 0.69766
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (de)
      config: de
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.41967
    - type: ndcg_at_10
      value: 0.45141
    - type: ndcg_at_100
      value: 0.53461
    - type: ndcg_at_1000
      value: 0.55463
    - type: ndcg_at_20
      value: 0.49012
    - type: ndcg_at_3
      value: 0.39486
    - type: ndcg_at_5
      value: 0.41496
    - type: recall_at_1
      value: 0.19494
    - type: recall_at_10
      value: 0.53774
    - type: recall_at_100
      value: 0.83314
    - type: recall_at_1000
      value: 0.95045
    - type: recall_at_20
      value: 0.65659
    - type: recall_at_3
      value: 0.3556
    - type: recall_at_5
      value: 0.44448
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (es)
      config: es
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.54475
    - type: ndcg_at_10
      value: 0.46593
    - type: ndcg_at_100
      value: 0.58079
    - type: ndcg_at_1000
      value: 0.60656
    - type: ndcg_at_20
      value: 0.51858
    - type: ndcg_at_3
      value: 0.4578
    - type: ndcg_at_5
      value: 0.44321
    - type: recall_at_1
      value: 0.15966
    - type: recall_at_10
      value: 0.49343
    - type: recall_at_100
      value: 0.82684
    - type: recall_at_1000
      value: 0.95299
    - type: recall_at_20
      value: 0.62367
    - type: recall_at_3
      value: 0.2949
    - type: recall_at_5
      value: 0.37983
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (fa)
      config: fa
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.36709
    - type: ndcg_at_10
      value: 0.46961
    - type: ndcg_at_100
      value: 0.53262
    - type: ndcg_at_1000
      value: 0.55024
    - type: ndcg_at_20
      value: 0.49892
    - type: ndcg_at_3
      value: 0.40235
    - type: ndcg_at_5
      value: 0.42866
    - type: recall_at_1
      value: 0.22735
    - type: recall_at_10
      value: 0.59949
    - type: recall_at_100
      value: 0.83867
    - type: recall_at_1000
      value: 0.95007
    - type: recall_at_20
      value: 0.68947
    - type: recall_at_3
      value: 0.41781
    - type: recall_at_5
      value: 0.49374
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (fi)
      config: fi
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.59245
    - type: ndcg_at_10
      value: 0.65551
    - type: ndcg_at_100
      value: 0.6967
    - type: ndcg_at_1000
      value: 0.70521
    - type: ndcg_at_20
      value: 0.67552
    - type: ndcg_at_3
      value: 0.58876
    - type: ndcg_at_5
      value: 0.61779
    - type: recall_at_1
      value: 0.37669
    - type: recall_at_10
      value: 0.76529
    - type: recall_at_100
      value: 0.9156
    - type: recall_at_1000
      value: 0.96977
    - type: recall_at_20
      value: 0.82685
    - type: recall_at_3
      value: 0.60234
    - type: recall_at_5
      value: 0.67135
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (fr)
      config: fr
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.38776
    - type: ndcg_at_10
      value: 0.47589
    - type: ndcg_at_100
      value: 0.54641
    - type: ndcg_at_1000
      value: 0.5629
    - type: ndcg_at_20
      value: 0.51203
    - type: ndcg_at_3
      value: 0.38924
    - type: ndcg_at_5
      value: 0.42572
    - type: recall_at_1
      value: 0.22082
    - type: recall_at_10
      value: 0.61619
    - type: recall_at_100
      value: 0.87237
    - type: recall_at_1000
      value: 0.97449
    - type: recall_at_20
      value: 0.72689
    - type: recall_at_3
      value: 0.39527
    - type: recall_at_5
      value: 0.48983
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (hi)
      config: hi
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.33143
    - type: ndcg_at_10
      value: 0.42084
    - type: ndcg_at_100
      value: 0.48647
    - type: ndcg_at_1000
      value: 0.50712
    - type: ndcg_at_20
      value: 0.45399
    - type: ndcg_at_3
      value: 0.34988
    - type: ndcg_at_5
      value: 0.37938
    - type: recall_at_1
      value: 0.17852
    - type: recall_at_10
      value: 0.55217
    - type: recall_at_100
      value: 0.79929
    - type: recall_at_1000
      value: 0.93434
    - type: recall_at_20
      value: 0.65231
    - type: recall_at_3
      value: 0.33765
    - type: recall_at_5
      value: 0.43828
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (id)
      config: id
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.43854
    - type: ndcg_at_10
      value: 0.45459
    - type: ndcg_at_100
      value: 0.53643
    - type: ndcg_at_1000
      value: 0.56052
    - type: ndcg_at_20
      value: 0.48795
    - type: ndcg_at_3
      value: 0.41041
    - type: ndcg_at_5
      value: 0.42235
    - type: recall_at_1
      value: 0.19193
    - type: recall_at_10
      value: 0.5289
    - type: recall_at_100
      value: 0.79649
    - type: recall_at_1000
      value: 0.92937
    - type: recall_at_20
      value: 0.61813
    - type: recall_at_3
      value: 0.35431
    - type: recall_at_5
      value: 0.43348
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (ja)
      config: ja
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.53256
    - type: ndcg_at_10
      value: 0.59922
    - type: ndcg_at_100
      value: 0.65407
    - type: ndcg_at_1000
      value: 0.66484
    - type: ndcg_at_20
      value: 0.62596
    - type: ndcg_at_3
      value: 0.53717
    - type: ndcg_at_5
      value: 0.56523
    - type: recall_at_1
      value: 0.34555
    - type: recall_at_10
      value: 0.71476
    - type: recall_at_100
      value: 0.91152
    - type: recall_at_1000
      value: 0.97728
    - type: recall_at_20
      value: 0.79811
    - type: recall_at_3
      value: 0.53482
    - type: recall_at_5
      value: 0.62327
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (ko)
      config: ko
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.5493
    - type: ndcg_at_10
      value: 0.58413
    - type: ndcg_at_100
      value: 0.64374
    - type: ndcg_at_1000
      value: 0.65655
    - type: ndcg_at_20
      value: 0.61732
    - type: ndcg_at_3
      value: 0.53068
    - type: ndcg_at_5
      value: 0.55202
    - type: recall_at_1
      value: 0.32602
    - type: recall_at_10
      value: 0.68647
    - type: recall_at_100
      value: 0.87746
    - type: recall_at_1000
      value: 0.95524
    - type: recall_at_20
      value: 0.78089
    - type: recall_at_3
      value: 0.49173
    - type: recall_at_5
      value: 0.5827
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (ru)
      config: ru
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.43131
    - type: ndcg_at_10
      value: 0.48262
    - type: ndcg_at_100
      value: 0.56158
    - type: ndcg_at_1000
      value: 0.57929
    - type: ndcg_at_20
      value: 0.52023
    - type: ndcg_at_3
      value: 0.42808
    - type: ndcg_at_5
      value: 0.44373
    - type: recall_at_1
      value: 0.22018
    - type: recall_at_10
      value: 0.58034
    - type: recall_at_100
      value: 0.84074
    - type: recall_at_1000
      value: 0.93938
    - type: recall_at_20
      value: 0.68603
    - type: recall_at_3
      value: 0.39307
    - type: recall_at_5
      value: 0.47077
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (sw)
      config: sw
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.50415
    - type: ndcg_at_10
      value: 0.59111
    - type: ndcg_at_100
      value: 0.64312
    - type: ndcg_at_1000
      value: 0.65089
    - type: ndcg_at_20
      value: 0.61651
    - type: ndcg_at_3
      value: 0.5304
    - type: ndcg_at_5
      value: 0.56139
    - type: recall_at_1
      value: 0.33267
    - type: recall_at_10
      value: 0.72082
    - type: recall_at_100
      value: 0.91377
    - type: recall_at_1000
      value: 0.96152
    - type: recall_at_20
      value: 0.79943
    - type: recall_at_3
      value: 0.5548
    - type: recall_at_5
      value: 0.64302
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (te)
      config: te
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.64372
    - type: ndcg_at_10
      value: 0.78175
    - type: ndcg_at_100
      value: 0.79523
    - type: ndcg_at_1000
      value: 0.79774
    - type: ndcg_at_20
      value: 0.78826
    - type: ndcg_at_3
      value: 0.74856
    - type: ndcg_at_5
      value: 0.77128
    - type: recall_at_1
      value: 0.63688
    - type: recall_at_10
      value: 0.90358
    - type: recall_at_100
      value: 0.96558
    - type: recall_at_1000
      value: 0.9847
    - type: recall_at_20
      value: 0.92834
    - type: recall_at_3
      value: 0.81804
    - type: recall_at_5
      value: 0.87198
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (th)
      config: th
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.65484
    - type: ndcg_at_10
      value: 0.71774
    - type: ndcg_at_100
      value: 0.75362
    - type: ndcg_at_1000
      value: 0.75898
    - type: ndcg_at_20
      value: 0.73709
    - type: ndcg_at_3
      value: 0.66199
    - type: ndcg_at_5
      value: 0.68451
    - type: recall_at_1
      value: 0.45911
    - type: recall_at_10
      value: 0.82619
    - type: recall_at_100
      value: 0.95515
    - type: recall_at_1000
      value: 0.98854
    - type: recall_at_20
      value: 0.88447
    - type: recall_at_3
      value: 0.67437
    - type: recall_at_5
      value: 0.73786
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (yo)
      config: yo
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.46218
    - type: ndcg_at_10
      value: 0.64685
    - type: ndcg_at_100
      value: 0.66941
    - type: ndcg_at_1000
      value: 0.67361
    - type: ndcg_at_20
      value: 0.65548
    - type: ndcg_at_3
      value: 0.57609
    - type: ndcg_at_5
      value: 0.62021
    - type: recall_at_1
      value: 0.42787
    - type: recall_at_10
      value: 0.82913
    - type: recall_at_100
      value: 0.93277
    - type: recall_at_1000
      value: 0.96499
    - type: recall_at_20
      value: 0.85994
    - type: recall_at_3
      value: 0.65406
    - type: recall_at_5
      value: 0.7542
  - dataset:
      type: miracl/mmteb-miracl
      name: Miracl (zh)
      config: zh
      split: dev
    task:
      type: Retrieval
    metrics:
    - type: ndcg_at_1
      value: 0.41985
    - type: ndcg_at_10
      value: 0.4837
    - type: ndcg_at_100
      value: 0.55961
    - type: ndcg_at_1000
      value: 0.5762
    - type: ndcg_at_20
      value: 0.51595
    - type: ndcg_at_3
      value: 0.42094
    - type: ndcg_at_5
      value: 0.44273
    - type: recall_at_1
      value: 0.21446
    - type: recall_at_10
      value: 0.59695
    - type: recall_at_100
      value: 0.87388
    - type: recall_at_1000
      value: 0.96833
    - type: recall_at_20
      value: 0.69252
    - type: recall_at_3
      value: 0.40377
    - type: recall_at_5
      value: 0.4903
---
# Granite-Embedding-107m-multilingual

**Model Summary:**
Granite-Embedding-107M-Multilingual is a 107M parameter dense biencoder embedding model from the Granite Embeddings suite that can be used to generate high quality text embeddings. This model produces embedding vectors of size 384 and is trained using a combination of open source relevance-pair datasets with permissive, enterprise-friendly license, and IBM collected and generated datasets. This model is developed using contrastive finetuning, knowledge distillation and model merging for improved performance.

- **Developers:** Granite Embedding Team, IBM
- **GitHub Repository:** [ibm-granite/granite-embedding-models](https://github.com/ibm-granite/granite-embedding-models)
- **Website**: [Granite Docs](https://www.ibm.com/granite/docs/)
- **Paper:** Coming Soon
- **Release Date**: December 18th, 2024
- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0)

**Supported Languages:** 
English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese. Users may finetune Granite-Embedding-107M-Multilingual for languages beyond these 12 languages.

**Intended use:** 
The model is designed to produce fixed length vector representations for a given text, which can be used for text similarity, retrieval, and search applications.

**Usage with Sentence Transformers:** 
The model is compatible with SentenceTransformer library and is very easy to use:

First, install the sentence transformers library
```shell
pip install sentence_transformers
```

The model can then be used to encode pairs of text and find the similarity between their representations

```python
from sentence_transformers import SentenceTransformer, util

model_path = "ibm-granite/granite-embedding-107m-multilingual"
# Load the Sentence Transformer model
model = SentenceTransformer(model_path)

input_queries = [
    ' Who made the song My achy breaky heart? ',
    'summit define'
    ]

input_passages = [
    "Achy Breaky Heart is a country song written by Don Von Tress. Originally titled Don't Tell My Heart and performed by The Marcy Brothers in 1991. ",
    "Definition of summit for English Language Learners. : 1 the highest point of a mountain : the top of a mountain. : 2 the highest level. : 3 a meeting or series of meetings between the leaders of two or more governments."
    ]

# encode queries and passages
query_embeddings = model.encode(input_queries)
passage_embeddings = model.encode(input_passages)

# calculate cosine similarity
print(util.cos_sim(query_embeddings, passage_embeddings))
```

**Usage with Huggingface Transformers:** 
This is a simple example of how to use the Granite-Embedding-107m-Multilingual model with the Transformers library and PyTorch.

First, install the required libraries
```shell
pip install transformers torch
```

The model can then be used to encode pairs of text

```python
import torch
from transformers import AutoModel, AutoTokenizer

model_path = "ibm-granite/granite-embedding-107m-multilingual"

# Load the model and tokenizer
model = AutoModel.from_pretrained(model_path)
tokenizer = AutoTokenizer.from_pretrained(model_path)
model.eval()

input_queries = [
    ' Who made the song My achy breaky heart? ',
    'summit define'
    ]

# tokenize inputs
tokenized_queries = tokenizer(input_queries, padding=True, truncation=True, return_tensors='pt')

# encode queries
with torch.no_grad():
    # Queries
    model_output = model(**tokenized_queries)
    # Perform pooling. granite-embedding-30m-english uses CLS Pooling
    query_embeddings = model_output[0][:, 0]

# normalize the embeddings
query_embeddings = torch.nn.functional.normalize(query_embeddings, dim=1)

```

**Evaluation:**
The average performance of the Granite-Embedding-107M-Multilingual on Multilingual Miracl (across 18 langauges), Mintaka Retrieval (across 8 languages) and MTEB Retrieval for English (across 15 tasks), German (across 4 tasks), Spanish (across 2 tasks), Frenc (across 5 tasks), Japanese (across 2 tasks), Arabic (1 task), Korean (1 task) and Chinese (across 8 tasks) is reported below. Granite-Embedding-107M-Multilingual is twice as fast as other models with similar embedding dimensions.

| Model                              | Paramters (M)| Embedding Dimension | Miracl (18)   |  Mintaka Retrieval (8) | MTEB English (15) | MTEB German (4) |MTEB Spanish (2) | MTEB French (5) | MTEB Japanese (2) |  MTEB Arabic (1) | MTEB Korean (1) | MTEB Chinese (8) | 
|------------------------------------|:------------:|:-------------------:|:-------------:| :---------------------:|:-----------------:|:---------------:|:---------------:|:---------------:|:----------------:|:----------------:|----------------:|-----------------:|
|granite-embedding-107m-multilingual | 107 | 384 | 55.9 | 22.6 | 45.3 | 70.3 | 48.7 | 51.1 | 59.0 | 63.2 | 70.5 | 40.8 | 

**Model Architecture:**
Granite-Embedding-107m-Multilingual is based on an encoder-only XLM-RoBERTa like transformer architecture, trained internally at IBM Research.

| Model                     | granite-embedding-30m-english | granite-embedding-125m-english    | granite-embedding-107m-multilingual | granite-embedding-278m-multilingual |
| :---------                | :-------:| :--------:   | :---------:| :-----:|
| Embedding size            | 384      | 768          | **384**    | 768    |
| Number of layers          | 6        | 12           | **6**      | 12     |
| Number of attention heads | 12       | 12           | **12**     | 12     |
| Intermediate size         | 1536     | 3072         | **1536**   | 3072   |
| Activation Function       | GeLU     | GeLU         | **GeLU**   | GeLU   |
| Vocabulary Size           | 50265    | 50265        | **250002** | 250002 |
| Max. Sequence Length      | 512      | 512          | **512**    | 512    |
| # Parameters              | 30M      | 125M         | **107M**   | 278M   |


**Training Data:**
Overall, the training data consists of four key sources: (1) unsupervised title-body paired data scraped from the web, (2) publicly available paired with permissive, enterprise-friendly license, (3) IBM-internal paired data targetting specific technical domains, and (4) IBM-generated synthetic data. The data is listed below:

| **Dataset**                                                               | **Num. Pairs** | 
|:--------------------------------------------------------------------------|:--------------:|
| Multilingual MC4                                                          | 52,823,484     |
| Multilingual Webhose                                                      | 12,369,322     | 
| English Wikipedia                                                         | 20,745,403     | 
| Multilingual Wikimedia                                                    | 2,911,090      |
| Miracl Corpus (Title-Body)                                                | 10,120,398     |
| Stack Exchange Duplicate questions (titles)                               | 304,525        | 
| Stack Exchange Duplicate questions (titles)                               | 304,525        | 
| Stack Exchange Duplicate questions (bodies)                               | 250,519        | 
| Machine Translations of Stack Exchange Duplicate questions (titles)       | 187,195        | 
| Stack Exchange (Title, Answer) pairs                                      | 4,067,139      | 
| Stack Exchange (Title, Body) pairs                                        | 23,978,013     | 
| Stack Exchange (Title, Body) pairs                                        | 23,978,013     | 
| Machine Translations of Stack Exchange (Title+Body, Answer) pairs         | 1,827,15       | 
| SearchQA                                                                  | 582,261        | 
| S2ORC (Title, Abstract)                                                   | 41,769,185     | 
| WikiAnswers Duplicate question pairs                                      | 77,427,422     | 
| CCNews                                                                    | 614,664        |
| XSum                                                                      | 226,711        |
| SimpleWiki                                                                | 102,225        |
| Machine Translated Cross Lingual Parallel Corpora                         | 28,376,115     |
| SPECTER citation triplets                                                 | 684,100        | 
| Machine Translations of SPECTER citation triplets                         | 4,104,600      | 
| Natural Questions (NQ)                                                    | 100,231        | 
| SQuAD2.0                                                                  | 87,599         | 
| HotpotQA                                                                  | 85,000         | 
| Fever                                                                     | 109,810        | 
| PubMed                                                                    | 20,000,000     | 
| Multilingual Miracl Triples                                               | 81,409         | 
| Multilingual MrTydi Triples                                               | 48,715         | 
| Sadeeem Question Asnwering                                                | 4,037          | 
| DBPedia Title-Body Pairs                                                  | 4,635,922      | 
| Synthetic: English Query-Wikipedia Passage                                | 1,879,093      | 
| Synthetic: English Fact Verification                                      | 9,888          | 
| Synthetic: Multilingual Query-Wikipedia Passage                           | 300,266        |
| Synthetic: Multilingual News Summaries                                    | 37,489         |
| IBM Internal Triples                                                      | 40,290         | 
| IBM Internal Title-Body Pairs                                             | 1,524,586      |

Notably, we do not use the popular MS-MARCO retrieval dataset in our training corpus due to its non-commercial license, while other open-source models train on this dataset due to its high quality.

**Infrastructure:**
We train Granite Embedding Models using IBM's computing cluster, Cognitive Compute Cluster, which is outfitted with NVIDIA A100 80gb GPUs. This cluster provides a scalable and efficient infrastructure for training our models over multiple GPUs.

**Ethical Considerations and Limitations:** 
The data used to train the base language model was filtered to remove text containing hate, abuse, and profanity. Granite-Embedding-278m-Multilingual is trained only for English texts, and has a context length of 512 tokens (longer texts will be truncated to this size).


<!-- ## Citation
```
@misc{granite-embedding-models,
  author = {author 1, author2, ...},
  title = {},
  journal = {},
  volume = {},
  year = {2024},
  url = {https://arxiv.org/abs/0000.00000},
}
``` -->