Hugging Face's logo

RedHatAI
/
bge-large-en-v1.5-sparse

Feature Extraction
Transformers
ONNX
English
bert
sparse sparsity quantized onnx embeddings int8
mteb
Eval Results
text-embeddings-inference
Model card Files Community
zeroshot commited on
Commit
9a3ed52
·
1 Parent(s): 31bd196

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +379 -0
README.md CHANGED
@@ -4,6 +4,385 @@ language:
4
  - en
5
  tags:
6
  - sparse sparsity quantized onnx embeddings int8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7
  ---
8
 
9
  # bge-large-en-v1.5-sparse
 
4
  - en
5
  tags:
6
  - sparse sparsity quantized onnx embeddings int8
7
+ - mteb
8
+ model-index:
9
+ - name: bge-large-en-v1.5-sparse
10
+ results:
11
+ - task:
12
+ type: STS
13
+ dataset:
14
+ type: mteb/biosses-sts
15
+ name: MTEB BIOSSES
16
+ config: default
17
+ split: test
18
+ revision: d3fb88f8f02e40887cd149695127462bbcf29b4a
19
+ metrics:
20
+ - type: cos_sim_pearson
21
+ value: 87.73305831153709
22
+ - type: cos_sim_spearman
23
+ value: 85.64351771070989
24
+ - type: euclidean_pearson
25
+ value: 86.06880877736519
26
+ - type: euclidean_spearman
27
+ value: 85.60676988543395
28
+ - type: manhattan_pearson
29
+ value: 85.69108036145253
30
+ - type: manhattan_spearman
31
+ value: 85.05314281283421
32
+ - task:
33
+ type: STS
34
+ dataset:
35
+ type: mteb/sickr-sts
36
+ name: MTEB SICK-R
37
+ config: default
38
+ split: test
39
+ revision: a6ea5a8cab320b040a23452cc28066d9beae2cee
40
+ metrics:
41
+ - type: cos_sim_pearson
42
+ value: 85.61833776000717
43
+ - type: cos_sim_spearman
44
+ value: 80.73718686921521
45
+ - type: euclidean_pearson
46
+ value: 83.9368704709159
47
+ - type: euclidean_spearman
48
+ value: 80.64477415487963
49
+ - type: manhattan_pearson
50
+ value: 83.92383757341743
51
+ - type: manhattan_spearman
52
+ value: 80.59625506933862
53
+ - task:
54
+ type: STS
55
+ dataset:
56
+ type: mteb/sts12-sts
57
+ name: MTEB STS12
58
+ config: default
59
+ split: test
60
+ revision: a0d554a64d88156834ff5ae9920b964011b16384
61
+ metrics:
62
+ - type: cos_sim_pearson
63
+ value: 83.81272888013494
64
+ - type: cos_sim_spearman
65
+ value: 76.07038564455931
66
+ - type: euclidean_pearson
67
+ value: 80.33676600912023
68
+ - type: euclidean_spearman
69
+ value: 75.86575335744111
70
+ - type: manhattan_pearson
71
+ value: 80.36973770593211
72
+ - type: manhattan_spearman
73
+ value: 75.88787860200954
74
+ - task:
75
+ type: STS
76
+ dataset:
77
+ type: mteb/sts13-sts
78
+ name: MTEB STS13
79
+ config: default
80
+ split: test
81
+ revision: 7e90230a92c190f1bf69ae9002b8cea547a64cca
82
+ metrics:
83
+ - type: cos_sim_pearson
84
+ value: 85.58781524090651
85
+ - type: cos_sim_spearman
86
+ value: 86.80508359626748
87
+ - type: euclidean_pearson
88
+ value: 85.22891409219575
89
+ - type: euclidean_spearman
90
+ value: 85.78295876926319
91
+ - type: manhattan_pearson
92
+ value: 85.2193177032458
93
+ - type: manhattan_spearman
94
+ value: 85.74049940198427
95
+ - task:
96
+ type: STS
97
+ dataset:
98
+ type: mteb/sts14-sts
99
+ name: MTEB STS14
100
+ config: default
101
+ split: test
102
+ revision: 6031580fec1f6af667f0bd2da0a551cf4f0b2375
103
+ metrics:
104
+ - type: cos_sim_pearson
105
+ value: 84.0862821699066
106
+ - type: cos_sim_spearman
107
+ value: 81.67856196476185
108
+ - type: euclidean_pearson
109
+ value: 83.38475353138897
110
+ - type: euclidean_spearman
111
+ value: 81.45279784228292
112
+ - type: manhattan_pearson
113
+ value: 83.29235221714131
114
+ - type: manhattan_spearman
115
+ value: 81.3971683104493
116
+ - task:
117
+ type: STS
118
+ dataset:
119
+ type: mteb/sts15-sts
120
+ name: MTEB STS15
121
+ config: default
122
+ split: test
123
+ revision: ae752c7c21bf194d8b67fd573edf7ae58183cbe3
124
+ metrics:
125
+ - type: cos_sim_pearson
126
+ value: 87.44459051393112
127
+ - type: cos_sim_spearman
128
+ value: 88.74673154561383
129
+ - type: euclidean_pearson
130
+ value: 88.13112382236628
131
+ - type: euclidean_spearman
132
+ value: 88.56241954487271
133
+ - type: manhattan_pearson
134
+ value: 88.11098632041256
135
+ - type: manhattan_spearman
136
+ value: 88.55607051247829
137
+ - task:
138
+ type: STS
139
+ dataset:
140
+ type: mteb/sts16-sts
141
+ name: MTEB STS16
142
+ config: default
143
+ split: test
144
+ revision: 4d8694f8f0e0100860b497b999b3dbed754a0513
145
+ metrics:
146
+ - type: cos_sim_pearson
147
+ value: 82.8825746257794
148
+ - type: cos_sim_spearman
149
+ value: 84.6066555379785
150
+ - type: euclidean_pearson
151
+ value: 84.12438131112606
152
+ - type: euclidean_spearman
153
+ value: 84.75862802179907
154
+ - type: manhattan_pearson
155
+ value: 84.12791217960807
156
+ - type: manhattan_spearman
157
+ value: 84.7739597139034
158
+ - task:
159
+ type: STS
160
+ dataset:
161
+ type: mteb/sts17-crosslingual-sts
162
+ name: MTEB STS17 (en-en)
163
+ config: en-en
164
+ split: test
165
+ revision: af5e6fb845001ecf41f4c1e033ce921939a2a68d
166
+ metrics:
167
+ - type: cos_sim_pearson
168
+ value: 89.19971502207773
169
+ - type: cos_sim_spearman
170
+ value: 89.75109780507901
171
+ - type: euclidean_pearson
172
+ value: 89.5913898113725
173
+ - type: euclidean_spearman
174
+ value: 89.20244860773123
175
+ - type: manhattan_pearson
176
+ value: 89.68755363801112
177
+ - type: manhattan_spearman
178
+ value: 89.3105024782381
179
+ - task:
180
+ type: STS
181
+ dataset:
182
+ type: mteb/sts22-crosslingual-sts
183
+ name: MTEB STS22 (en)
184
+ config: en
185
+ split: test
186
+ revision: 6d1ba47164174a496b7fa5d3569dae26a6813b80
187
+ metrics:
188
+ - type: cos_sim_pearson
189
+ value: 61.73885819503523
190
+ - type: cos_sim_spearman
191
+ value: 64.09521607825829
192
+ - type: euclidean_pearson
193
+ value: 64.22116001518724
194
+ - type: euclidean_spearman
195
+ value: 63.84189650719827
196
+ - type: manhattan_pearson
197
+ value: 64.23930191730729
198
+ - type: manhattan_spearman
199
+ value: 63.7536172795383
200
+ - task:
201
+ type: STS
202
+ dataset:
203
+ type: mteb/stsbenchmark-sts
204
+ name: MTEB STSBenchmark
205
+ config: default
206
+ split: test
207
+ revision: b0fddb56ed78048fa8b90373c8a3cfc37b684831
208
+ metrics:
209
+ - type: cos_sim_pearson
210
+ value: 85.68505574064375
211
+ - type: cos_sim_spearman
212
+ value: 86.87614324154406
213
+ - type: euclidean_pearson
214
+ value: 86.96751967489614
215
+ - type: euclidean_spearman
216
+ value: 86.78979082790067
217
+ - type: manhattan_pearson
218
+ value: 86.92578795715433
219
+ - type: manhattan_spearman
220
+ value: 86.74076104131726
221
+ - task:
222
+ type: PairClassification
223
+ dataset:
224
+ type: mteb/sprintduplicatequestions-pairclassification
225
+ name: MTEB SprintDuplicateQuestions
226
+ config: default
227
+ split: test
228
+ revision: d66bd1f72af766a5cc4b0ca5e00c162f89e8cc46
229
+ metrics:
230
+ - type: cos_sim_accuracy
231
+ value: 99.80990099009901
232
+ - type: cos_sim_ap
233
+ value: 95.00187845875503
234
+ - type: cos_sim_f1
235
+ value: 90.37698412698413
236
+ - type: cos_sim_precision
237
+ value: 89.66535433070865
238
+ - type: cos_sim_recall
239
+ value: 91.10000000000001
240
+ - type: dot_accuracy
241
+ value: 99.63366336633663
242
+ - type: dot_ap
243
+ value: 87.6642728041652
244
+ - type: dot_f1
245
+ value: 81.40803173029252
246
+ - type: dot_precision
247
+ value: 80.7276302851524
248
+ - type: dot_recall
249
+ value: 82.1
250
+ - type: euclidean_accuracy
251
+ value: 99.8079207920792
252
+ - type: euclidean_ap
253
+ value: 94.88531851782375
254
+ - type: euclidean_f1
255
+ value: 90.49019607843137
256
+ - type: euclidean_precision
257
+ value: 88.75
258
+ - type: euclidean_recall
259
+ value: 92.30000000000001
260
+ - type: manhattan_accuracy
261
+ value: 99.81188118811882
262
+ - type: manhattan_ap
263
+ value: 94.87944331919043
264
+ - type: manhattan_f1
265
+ value: 90.5
266
+ - type: manhattan_precision
267
+ value: 90.5
268
+ - type: manhattan_recall
269
+ value: 90.5
270
+ - type: max_accuracy
271
+ value: 99.81188118811882
272
+ - type: max_ap
273
+ value: 95.00187845875503
274
+ - type: max_f1
275
+ value: 90.5
276
+ - task:
277
+ type: PairClassification
278
+ dataset:
279
+ type: mteb/twittersemeval2015-pairclassification
280
+ name: MTEB TwitterSemEval2015
281
+ config: default
282
+ split: test
283
+ revision: 70970daeab8776df92f5ea462b6173c0b46fd2d1
284
+ metrics:
285
+ - type: cos_sim_accuracy
286
+ value: 86.3861238600465
287
+ - type: cos_sim_ap
288
+ value: 74.50058066578084
289
+ - type: cos_sim_f1
290
+ value: 69.25949774629748
291
+ - type: cos_sim_precision
292
+ value: 67.64779874213836
293
+ - type: cos_sim_recall
294
+ value: 70.94986807387863
295
+ - type: dot_accuracy
296
+ value: 81.57000655659535
297
+ - type: dot_ap
298
+ value: 59.10193583653485
299
+ - type: dot_f1
300
+ value: 58.39352155832786
301
+ - type: dot_precision
302
+ value: 49.88780852655198
303
+ - type: dot_recall
304
+ value: 70.3957783641161
305
+ - type: euclidean_accuracy
306
+ value: 86.37420277761221
307
+ - type: euclidean_ap
308
+ value: 74.41671247141966
309
+ - type: euclidean_f1
310
+ value: 69.43907156673114
311
+ - type: euclidean_precision
312
+ value: 64.07853636769299
313
+ - type: euclidean_recall
314
+ value: 75.77836411609499
315
+ - type: manhattan_accuracy
316
+ value: 86.30267628300649
317
+ - type: manhattan_ap
318
+ value: 74.34438603336339
319
+ - type: manhattan_f1
320
+ value: 69.41888619854721
321
+ - type: manhattan_precision
322
+ value: 64.13870246085011
323
+ - type: manhattan_recall
324
+ value: 75.64643799472296
325
+ - type: max_accuracy
326
+ value: 86.3861238600465
327
+ - type: max_ap
328
+ value: 74.50058066578084
329
+ - type: max_f1
330
+ value: 69.43907156673114
331
+ - task:
332
+ type: PairClassification
333
+ dataset:
334
+ type: mteb/twitterurlcorpus-pairclassification
335
+ name: MTEB TwitterURLCorpus
336
+ config: default
337
+ split: test
338
+ revision: 8b6510b0b1fa4e4c4f879467980e9be563ec1cdf
339
+ metrics:
340
+ - type: cos_sim_accuracy
341
+ value: 88.87530562347187
342
+ - type: cos_sim_ap
343
+ value: 85.69496469410068
344
+ - type: cos_sim_f1
345
+ value: 77.96973052787007
346
+ - type: cos_sim_precision
347
+ value: 74.8900865125514
348
+ - type: cos_sim_recall
349
+ value: 81.3135201724669
350
+ - type: dot_accuracy
351
+ value: 86.70780455621532
352
+ - type: dot_ap
353
+ value: 80.03489678512908
354
+ - type: dot_f1
355
+ value: 73.26376129933124
356
+ - type: dot_precision
357
+ value: 70.07591733445804
358
+ - type: dot_recall
359
+ value: 76.75546658453958
360
+ - type: euclidean_accuracy
361
+ value: 88.85978189156674
362
+ - type: euclidean_ap
363
+ value: 85.67894953317325
364
+ - type: euclidean_f1
365
+ value: 78.04295942720763
366
+ - type: euclidean_precision
367
+ value: 75.67254845241538
368
+ - type: euclidean_recall
369
+ value: 80.56667693255312
370
+ - type: manhattan_accuracy
371
+ value: 88.88306748942446
372
+ - type: manhattan_ap
373
+ value: 85.66556510677526
374
+ - type: manhattan_f1
375
+ value: 78.06278290950576
376
+ - type: manhattan_precision
377
+ value: 74.76912231230173
378
+ - type: manhattan_recall
379
+ value: 81.65999384046813
380
+ - type: max_accuracy
381
+ value: 88.88306748942446
382
+ - type: max_ap
383
+ value: 85.69496469410068
384
+ - type: max_f1
385
+ value: 78.06278290950576
386
  ---
387
 
388
  # bge-large-en-v1.5-sparse