Hugging Face's logo

TownsWu
/
PEG

Sentence Similarity
Transformers
PyTorch
Chinese
bert
feature-extraction
PEG
mteb
Eval Results (legacy)
text-embeddings-inference
Model card Files
xet
 Community
7
TownsWu commited on
Commit
3e56f68
·
1 Parent(s): c86aea0

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +608 -0
README.md CHANGED
@@ -7,6 +7,614 @@ tags:
7
  - feature-extraction
8
  - sentence-similarity
9
  - transformers
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
10
  license: apache-2.0
11
  library_name: transformers
12
  ---
 
7
  - feature-extraction
8
  - sentence-similarity
9
  - transformers
10
+ - mteb
11
+ model-index:
12
+ - name: PEG
13
+ results:
14
+ - task:
15
+ type: Reranking
16
+ dataset:
17
+ type: C-MTEB/CMedQAv1-reranking
18
+ name: MTEB CMedQAv1
19
+ config: default
20
+ split: test
21
+ revision: None
22
+ metrics:
23
+ - type: map
24
+ value: 84.07718102381898
25
+ - type: mrr
26
+ value: 86.67924603174603
27
+ - task:
28
+ type: Reranking
29
+ dataset:
30
+ type: C-MTEB/CMedQAv2-reranking
31
+ name: MTEB CMedQAv2
32
+ config: default
33
+ split: test
34
+ revision: None
35
+ metrics:
36
+ - type: map
37
+ value: 85.13503552770737
38
+ - type: mrr
39
+ value: 87.48738095238096
40
+ - task:
41
+ type: Retrieval
42
+ dataset:
43
+ type: C-MTEB/CmedqaRetrieval
44
+ name: MTEB CmedqaRetrieval
45
+ config: default
46
+ split: dev
47
+ revision: None
48
+ metrics:
49
+ - type: map_at_1
50
+ value: 24.168
51
+ - type: map_at_10
52
+ value: 36.022999999999996
53
+ - type: map_at_100
54
+ value: 37.819
55
+ - type: map_at_1000
56
+ value: 37.95
57
+ - type: map_at_3
58
+ value: 31.947
59
+ - type: map_at_5
60
+ value: 34.172999999999995
61
+ - type: mrr_at_1
62
+ value: 36.934
63
+ - type: mrr_at_10
64
+ value: 44.899
65
+ - type: mrr_at_100
66
+ value: 45.838
67
+ - type: mrr_at_1000
68
+ value: 45.894
69
+ - type: mrr_at_3
70
+ value: 42.336
71
+ - type: mrr_at_5
72
+ value: 43.722
73
+ - type: ndcg_at_1
74
+ value: 36.934
75
+ - type: ndcg_at_10
76
+ value: 42.565
77
+ - type: ndcg_at_100
78
+ value: 49.677
79
+ - type: ndcg_at_1000
80
+ value: 51.989
81
+ - type: ndcg_at_3
82
+ value: 37.302
83
+ - type: ndcg_at_5
84
+ value: 39.393
85
+ - type: precision_at_1
86
+ value: 36.934
87
+ - type: precision_at_10
88
+ value: 9.547
89
+ - type: precision_at_100
90
+ value: 1.542
91
+ - type: precision_at_1000
92
+ value: 0.183
93
+ - type: precision_at_3
94
+ value: 21.154999999999998
95
+ - type: precision_at_5
96
+ value: 15.399
97
+ - type: recall_at_1
98
+ value: 24.168
99
+ - type: recall_at_10
100
+ value: 53.019000000000005
101
+ - type: recall_at_100
102
+ value: 82.699
103
+ - type: recall_at_1000
104
+ value: 98.221
105
+ - type: recall_at_3
106
+ value: 37.144
107
+ - type: recall_at_5
108
+ value: 43.61
109
+ - task:
110
+ type: Retrieval
111
+ dataset:
112
+ type: C-MTEB/CovidRetrieval
113
+ name: MTEB CovidRetrieval
114
+ config: default
115
+ split: dev
116
+ revision: None
117
+ metrics:
118
+ - type: map_at_1
119
+ value: 70.232
120
+ - type: map_at_10
121
+ value: 78.431
122
+ - type: map_at_100
123
+ value: 78.69
124
+ - type: map_at_1000
125
+ value: 78.692
126
+ - type: map_at_3
127
+ value: 76.765
128
+ - type: map_at_5
129
+ value: 77.798
130
+ - type: mrr_at_1
131
+ value: 70.39
132
+ - type: mrr_at_10
133
+ value: 78.41
134
+ - type: mrr_at_100
135
+ value: 78.669
136
+ - type: mrr_at_1000
137
+ value: 78.672
138
+ - type: mrr_at_3
139
+ value: 76.818
140
+ - type: mrr_at_5
141
+ value: 77.803
142
+ - type: ndcg_at_1
143
+ value: 70.49499999999999
144
+ - type: ndcg_at_10
145
+ value: 82.098
146
+ - type: ndcg_at_100
147
+ value: 83.296
148
+ - type: ndcg_at_1000
149
+ value: 83.402
150
+ - type: ndcg_at_3
151
+ value: 78.847
152
+ - type: ndcg_at_5
153
+ value: 80.634
154
+ - type: precision_at_1
155
+ value: 70.49499999999999
156
+ - type: precision_at_10
157
+ value: 9.442
158
+ - type: precision_at_100
159
+ value: 0.9990000000000001
160
+ - type: precision_at_1000
161
+ value: 0.101
162
+ - type: precision_at_3
163
+ value: 28.416000000000004
164
+ - type: precision_at_5
165
+ value: 17.956
166
+ - type: recall_at_1
167
+ value: 70.232
168
+ - type: recall_at_10
169
+ value: 93.361
170
+ - type: recall_at_100
171
+ value: 98.84100000000001
172
+ - type: recall_at_1000
173
+ value: 99.789
174
+ - type: recall_at_3
175
+ value: 84.721
176
+ - type: recall_at_5
177
+ value: 88.962
178
+ - task:
179
+ type: Retrieval
180
+ dataset:
181
+ type: C-MTEB/DuRetrieval
182
+ name: MTEB DuRetrieval
183
+ config: default
184
+ split: dev
185
+ revision: None
186
+ metrics:
187
+ - type: map_at_1
188
+ value: 25.957
189
+ - type: map_at_10
190
+ value: 79.243
191
+ - type: map_at_100
192
+ value: 82.137
193
+ - type: map_at_1000
194
+ value: 82.174
195
+ - type: map_at_3
196
+ value: 54.57300000000001
197
+ - type: map_at_5
198
+ value: 69.10300000000001
199
+ - type: mrr_at_1
200
+ value: 89.14999999999999
201
+ - type: mrr_at_10
202
+ value: 92.721
203
+ - type: mrr_at_100
204
+ value: 92.778
205
+ - type: mrr_at_1000
206
+ value: 92.781
207
+ - type: mrr_at_3
208
+ value: 92.44200000000001
209
+ - type: mrr_at_5
210
+ value: 92.622
211
+ - type: ndcg_at_1
212
+ value: 89.14999999999999
213
+ - type: ndcg_at_10
214
+ value: 86.835
215
+ - type: ndcg_at_100
216
+ value: 89.645
217
+ - type: ndcg_at_1000
218
+ value: 89.997
219
+ - type: ndcg_at_3
220
+ value: 85.187
221
+ - type: ndcg_at_5
222
+ value: 84.419
223
+ - type: precision_at_1
224
+ value: 89.14999999999999
225
+ - type: precision_at_10
226
+ value: 41.77
227
+ - type: precision_at_100
228
+ value: 4.81
229
+ - type: precision_at_1000
230
+ value: 0.48900000000000005
231
+ - type: precision_at_3
232
+ value: 76.3
233
+ - type: precision_at_5
234
+ value: 64.8
235
+ - type: recall_at_1
236
+ value: 25.957
237
+ - type: recall_at_10
238
+ value: 88.293
239
+ - type: recall_at_100
240
+ value: 97.476
241
+ - type: recall_at_1000
242
+ value: 99.367
243
+ - type: recall_at_3
244
+ value: 57.04
245
+ - type: recall_at_5
246
+ value: 74.17
247
+ - task:
248
+ type: Retrieval
249
+ dataset:
250
+ type: C-MTEB/EcomRetrieval
251
+ name: MTEB EcomRetrieval
252
+ config: default
253
+ split: dev
254
+ revision: None
255
+ metrics:
256
+ - type: map_at_1
257
+ value: 52.5
258
+ - type: map_at_10
259
+ value: 61.766
260
+ - type: map_at_100
261
+ value: 62.358999999999995
262
+ - type: map_at_1000
263
+ value: 62.370999999999995
264
+ - type: map_at_3
265
+ value: 59.45
266
+ - type: map_at_5
267
+ value: 60.834999999999994
268
+ - type: mrr_at_1
269
+ value: 52.5
270
+ - type: mrr_at_10
271
+ value: 61.766
272
+ - type: mrr_at_100
273
+ value: 62.358999999999995
274
+ - type: mrr_at_1000
275
+ value: 62.370999999999995
276
+ - type: mrr_at_3
277
+ value: 59.45
278
+ - type: mrr_at_5
279
+ value: 60.834999999999994
280
+ - type: ndcg_at_1
281
+ value: 52.5
282
+ - type: ndcg_at_10
283
+ value: 66.4
284
+ - type: ndcg_at_100
285
+ value: 69.234
286
+ - type: ndcg_at_1000
287
+ value: 69.572
288
+ - type: ndcg_at_3
289
+ value: 61.675000000000004
290
+ - type: ndcg_at_5
291
+ value: 64.158
292
+ - type: precision_at_1
293
+ value: 52.5
294
+ - type: precision_at_10
295
+ value: 8.1
296
+ - type: precision_at_100
297
+ value: 0.941
298
+ - type: precision_at_1000
299
+ value: 0.097
300
+ - type: precision_at_3
301
+ value: 22.7
302
+ - type: precision_at_5
303
+ value: 14.82
304
+ - type: recall_at_1
305
+ value: 52.5
306
+ - type: recall_at_10
307
+ value: 81.0
308
+ - type: recall_at_100
309
+ value: 94.1
310
+ - type: recall_at_1000
311
+ value: 96.8
312
+ - type: recall_at_3
313
+ value: 68.10000000000001
314
+ - type: recall_at_5
315
+ value: 74.1
316
+ - task:
317
+ type: Reranking
318
+ dataset:
319
+ type: C-MTEB/Mmarco-reranking
320
+ name: MTEB MMarcoReranking
321
+ config: default
322
+ split: dev
323
+ revision: None
324
+ metrics:
325
+ - type: map
326
+ value: 32.02701536591035
327
+ - type: mrr
328
+ value: 31.004761904761907
329
+ - task:
330
+ type: Retrieval
331
+ dataset:
332
+ type: C-MTEB/MMarcoRetrieval
333
+ name: MTEB MMarcoRetrieval
334
+ config: default
335
+ split: dev
336
+ revision: None
337
+ metrics:
338
+ - type: map_at_1
339
+ value: 68.548
340
+ - type: map_at_10
341
+ value: 77.547
342
+ - type: map_at_100
343
+ value: 77.80799999999999
344
+ - type: map_at_1000
345
+ value: 77.818
346
+ - type: map_at_3
347
+ value: 75.82600000000001
348
+ - type: map_at_5
349
+ value: 76.935
350
+ - type: mrr_at_1
351
+ value: 70.903
352
+ - type: mrr_at_10
353
+ value: 78.074
354
+ - type: mrr_at_100
355
+ value: 78.303
356
+ - type: mrr_at_1000
357
+ value: 78.312
358
+ - type: mrr_at_3
359
+ value: 76.60900000000001
360
+ - type: mrr_at_5
361
+ value: 77.537
362
+ - type: ndcg_at_1
363
+ value: 70.903
364
+ - type: ndcg_at_10
365
+ value: 81.04400000000001
366
+ - type: ndcg_at_100
367
+ value: 82.206
368
+ - type: ndcg_at_1000
369
+ value: 82.46900000000001
370
+ - type: ndcg_at_3
371
+ value: 77.823
372
+ - type: ndcg_at_5
373
+ value: 79.654
374
+ - type: precision_at_1
375
+ value: 70.903
376
+ - type: precision_at_10
377
+ value: 9.713
378
+ - type: precision_at_100
379
+ value: 1.0290000000000001
380
+ - type: precision_at_1000
381
+ value: 0.105
382
+ - type: precision_at_3
383
+ value: 29.226000000000003
384
+ - type: precision_at_5
385
+ value: 18.516
386
+ - type: recall_at_1
387
+ value: 68.548
388
+ - type: recall_at_10
389
+ value: 91.27499999999999
390
+ - type: recall_at_100
391
+ value: 96.523
392
+ - type: recall_at_1000
393
+ value: 98.601
394
+ - type: recall_at_3
395
+ value: 82.779
396
+ - type: recall_at_5
397
+ value: 87.1
398
+ - task:
399
+ type: Retrieval
400
+ dataset:
401
+ type: C-MTEB/MedicalRetrieval
402
+ name: MTEB MedicalRetrieval
403
+ config: default
404
+ split: dev
405
+ revision: None
406
+ metrics:
407
+ - type: map_at_1
408
+ value: 53.400000000000006
409
+ - type: map_at_10
410
+ value: 58.248
411
+ - type: map_at_100
412
+ value: 58.786
413
+ - type: map_at_1000
414
+ value: 58.836
415
+ - type: map_at_3
416
+ value: 57.199999999999996
417
+ - type: map_at_5
418
+ value: 57.809999999999995
419
+ - type: mrr_at_1
420
+ value: 53.400000000000006
421
+ - type: mrr_at_10
422
+ value: 58.248
423
+ - type: mrr_at_100
424
+ value: 58.786
425
+ - type: mrr_at_1000
426
+ value: 58.836
427
+ - type: mrr_at_3
428
+ value: 57.199999999999996
429
+ - type: mrr_at_5
430
+ value: 57.809999999999995
431
+ - type: ndcg_at_1
432
+ value: 53.400000000000006
433
+ - type: ndcg_at_10
434
+ value: 60.655
435
+ - type: ndcg_at_100
436
+ value: 63.683
437
+ - type: ndcg_at_1000
438
+ value: 65.159
439
+ - type: ndcg_at_3
440
+ value: 58.504999999999995
441
+ - type: ndcg_at_5
442
+ value: 59.589000000000006
443
+ - type: precision_at_1
444
+ value: 53.400000000000006
445
+ - type: precision_at_10
446
+ value: 6.819999999999999
447
+ - type: precision_at_100
448
+ value: 0.8330000000000001
449
+ - type: precision_at_1000
450
+ value: 0.095
451
+ - type: precision_at_3
452
+ value: 20.767
453
+ - type: precision_at_5
454
+ value: 12.98
455
+ - type: recall_at_1
456
+ value: 53.400000000000006
457
+ - type: recall_at_10
458
+ value: 68.2
459
+ - type: recall_at_100
460
+ value: 83.3
461
+ - type: recall_at_1000
462
+ value: 95.19999999999999
463
+ - type: recall_at_3
464
+ value: 62.3
465
+ - type: recall_at_5
466
+ value: 64.9
467
+ - task:
468
+ type: Reranking
469
+ dataset:
470
+ type: C-MTEB/T2Reranking
471
+ name: MTEB T2Reranking
472
+ config: default
473
+ split: dev
474
+ revision: None
475
+ metrics:
476
+ - type: map
477
+ value: 68.88440464348339
478
+ - type: mrr
479
+ value: 79.58819625796606
480
+ - task:
481
+ type: Retrieval
482
+ dataset:
483
+ type: C-MTEB/T2Retrieval
484
+ name: MTEB T2Retrieval
485
+ config: default
486
+ split: dev
487
+ revision: None
488
+ metrics:
489
+ - type: map_at_1
490
+ value: 27.32
491
+ - type: map_at_10
492
+ value: 77.47699999999999
493
+ - type: map_at_100
494
+ value: 81.10000000000001
495
+ - type: map_at_1000
496
+ value: 81.159
497
+ - type: map_at_3
498
+ value: 54.203
499
+ - type: map_at_5
500
+ value: 66.76299999999999
501
+ - type: mrr_at_1
502
+ value: 89.983
503
+ - type: mrr_at_10
504
+ value: 92.487
505
+ - type: mrr_at_100
506
+ value: 92.575
507
+ - type: mrr_at_1000
508
+ value: 92.578
509
+ - type: mrr_at_3
510
+ value: 92.06299999999999
511
+ - type: mrr_at_5
512
+ value: 92.34599999999999
513
+ - type: ndcg_at_1
514
+ value: 89.983
515
+ - type: ndcg_at_10
516
+ value: 84.936
517
+ - type: ndcg_at_100
518
+ value: 88.442
519
+ - type: ndcg_at_1000
520
+ value: 89.023
521
+ - type: ndcg_at_3
522
+ value: 86.157
523
+ - type: ndcg_at_5
524
+ value: 84.83000000000001
525
+ - type: precision_at_1
526
+ value: 89.983
527
+ - type: precision_at_10
528
+ value: 42.405
529
+ - type: precision_at_100
530
+ value: 5.032
531
+ - type: precision_at_1000
532
+ value: 0.516
533
+ - type: precision_at_3
534
+ value: 75.51
535
+ - type: precision_at_5
536
+ value: 63.413
537
+ - type: recall_at_1
538
+ value: 27.32
539
+ - type: recall_at_10
540
+ value: 83.954
541
+ - type: recall_at_100
542
+ value: 95.48100000000001
543
+ - type: recall_at_1000
544
+ value: 98.502
545
+ - type: recall_at_3
546
+ value: 55.925000000000004
547
+ - type: recall_at_5
548
+ value: 70.19800000000001
549
+ - task:
550
+ type: Retrieval
551
+ dataset:
552
+ type: C-MTEB/VideoRetrieval
553
+ name: MTEB VideoRetrieval
554
+ config: default
555
+ split: dev
556
+ revision: None
557
+ metrics:
558
+ - type: map_at_1
559
+ value: 64.0
560
+ - type: map_at_10
561
+ value: 72.69
562
+ - type: map_at_100
563
+ value: 73.07000000000001
564
+ - type: map_at_1000
565
+ value: 73.08
566
+ - type: map_at_3
567
+ value: 71.217
568
+ - type: map_at_5
569
+ value: 72.072
570
+ - type: mrr_at_1
571
+ value: 64.0
572
+ - type: mrr_at_10
573
+ value: 72.69
574
+ - type: mrr_at_100
575
+ value: 73.07000000000001
576
+ - type: mrr_at_1000
577
+ value: 73.08
578
+ - type: mrr_at_3
579
+ value: 71.217
580
+ - type: mrr_at_5
581
+ value: 72.072
582
+ - type: ndcg_at_1
583
+ value: 64.0
584
+ - type: ndcg_at_10
585
+ value: 76.532
586
+ - type: ndcg_at_100
587
+ value: 78.405
588
+ - type: ndcg_at_1000
589
+ value: 78.632
590
+ - type: ndcg_at_3
591
+ value: 73.527
592
+ - type: ndcg_at_5
593
+ value: 75.059
594
+ - type: precision_at_1
595
+ value: 64.0
596
+ - type: precision_at_10
597
+ value: 8.84
598
+ - type: precision_at_100
599
+ value: 0.972
600
+ - type: precision_at_1000
601
+ value: 0.099
602
+ - type: precision_at_3
603
+ value: 26.733
604
+ - type: precision_at_5
605
+ value: 16.78
606
+ - type: recall_at_1
607
+ value: 64.0
608
+ - type: recall_at_10
609
+ value: 88.4
610
+ - type: recall_at_100
611
+ value: 97.2
612
+ - type: recall_at_1000
613
+ value: 98.9
614
+ - type: recall_at_3
615
+ value: 80.2
616
+ - type: recall_at_5
617
+ value: 83.89999999999999
618
  license: apache-2.0
619
  library_name: transformers
620
  ---