Hugging Face's logo

zeroshot
/
gte-large-quant

Feature Extraction
Transformers
ONNX
English
bert
sparse sparsity quantized onnx embeddings int8
mteb
Eval Results
text-embeddings-inference
Inference Endpoints
Model card Files Community
zeroshot commited on
Commit
a8754aa
·
1 Parent(s): f52564f

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +380 -0
README.md CHANGED
@@ -1,6 +1,386 @@
1
  ---
2
  tags:
3
  - sparse sparsity quantized onnx embeddings int8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4
  license: mit
5
  language:
6
  - en
 
1
  ---
2
  tags:
3
  - sparse sparsity quantized onnx embeddings int8
4
+ - mteb
5
+ - mteb
6
+ model-index:
7
+ - name: gte-large-quant
8
+ results:
9
+ - task:
10
+ type: STS
11
+ dataset:
12
+ type: mteb/biosses-sts
13
+ name: MTEB BIOSSES
14
+ config: default
15
+ split: test
16
+ revision: d3fb88f8f02e40887cd149695127462bbcf29b4a
17
+ metrics:
18
+ - type: cos_sim_pearson
19
+ value: 90.27260027646717
20
+ - type: cos_sim_spearman
21
+ value: 87.97790825077952
22
+ - type: euclidean_pearson
23
+ value: 88.42832241523092
24
+ - type: euclidean_spearman
25
+ value: 87.97248644049293
26
+ - type: manhattan_pearson
27
+ value: 88.13802465778512
28
+ - type: manhattan_spearman
29
+ value: 87.43391995202266
30
+ - task:
31
+ type: STS
32
+ dataset:
33
+ type: mteb/sickr-sts
34
+ name: MTEB SICK-R
35
+ config: default
36
+ split: test
37
+ revision: a6ea5a8cab320b040a23452cc28066d9beae2cee
38
+ metrics:
39
+ - type: cos_sim_pearson
40
+ value: 85.1416039713116
41
+ - type: cos_sim_spearman
42
+ value: 79.13359419669726
43
+ - type: euclidean_pearson
44
+ value: 83.08042050989465
45
+ - type: euclidean_spearman
46
+ value: 79.31565112619433
47
+ - type: manhattan_pearson
48
+ value: 83.10376638254372
49
+ - type: manhattan_spearman
50
+ value: 79.30772376012946
51
+ - task:
52
+ type: STS
53
+ dataset:
54
+ type: mteb/sts12-sts
55
+ name: MTEB STS12
56
+ config: default
57
+ split: test
58
+ revision: a0d554a64d88156834ff5ae9920b964011b16384
59
+ metrics:
60
+ - type: cos_sim_pearson
61
+ value: 84.93030439955828
62
+ - type: cos_sim_spearman
63
+ value: 75.98104622572393
64
+ - type: euclidean_pearson
65
+ value: 81.20791722502764
66
+ - type: euclidean_spearman
67
+ value: 75.74595761987686
68
+ - type: manhattan_pearson
69
+ value: 81.23169425598003
70
+ - type: manhattan_spearman
71
+ value: 75.73065403644094
72
+ - task:
73
+ type: STS
74
+ dataset:
75
+ type: mteb/sts13-sts
76
+ name: MTEB STS13
77
+ config: default
78
+ split: test
79
+ revision: 7e90230a92c190f1bf69ae9002b8cea547a64cca
80
+ metrics:
81
+ - type: cos_sim_pearson
82
+ value: 85.6693892097855
83
+ - type: cos_sim_spearman
84
+ value: 87.54973524492165
85
+ - type: euclidean_pearson
86
+ value: 86.55642466103943
87
+ - type: euclidean_spearman
88
+ value: 87.47921340148683
89
+ - type: manhattan_pearson
90
+ value: 86.52043275063926
91
+ - type: manhattan_spearman
92
+ value: 87.43869426658489
93
+ - task:
94
+ type: STS
95
+ dataset:
96
+ type: mteb/sts14-sts
97
+ name: MTEB STS14
98
+ config: default
99
+ split: test
100
+ revision: 6031580fec1f6af667f0bd2da0a551cf4f0b2375
101
+ metrics:
102
+ - type: cos_sim_pearson
103
+ value: 84.37393784507647
104
+ - type: cos_sim_spearman
105
+ value: 81.98702164762233
106
+ - type: euclidean_pearson
107
+ value: 84.22038158338351
108
+ - type: euclidean_spearman
109
+ value: 81.9872746771322
110
+ - type: manhattan_pearson
111
+ value: 84.21915949674062
112
+ - type: manhattan_spearman
113
+ value: 81.97923386273747
114
+ - task:
115
+ type: STS
116
+ dataset:
117
+ type: mteb/sts15-sts
118
+ name: MTEB STS15
119
+ config: default
120
+ split: test
121
+ revision: ae752c7c21bf194d8b67fd573edf7ae58183cbe3
122
+ metrics:
123
+ - type: cos_sim_pearson
124
+ value: 87.34477744314285
125
+ - type: cos_sim_spearman
126
+ value: 88.92669309789463
127
+ - type: euclidean_pearson
128
+ value: 88.20128441166663
129
+ - type: euclidean_spearman
130
+ value: 88.91524205114627
131
+ - type: manhattan_pearson
132
+ value: 88.24425729639415
133
+ - type: manhattan_spearman
134
+ value: 88.97457451709523
135
+ - task:
136
+ type: STS
137
+ dataset:
138
+ type: mteb/sts16-sts
139
+ name: MTEB STS16
140
+ config: default
141
+ split: test
142
+ revision: 4d8694f8f0e0100860b497b999b3dbed754a0513
143
+ metrics:
144
+ - type: cos_sim_pearson
145
+ value: 82.11827015492467
146
+ - type: cos_sim_spearman
147
+ value: 83.59397157586835
148
+ - type: euclidean_pearson
149
+ value: 82.97284591328044
150
+ - type: euclidean_spearman
151
+ value: 83.74509747941255
152
+ - type: manhattan_pearson
153
+ value: 82.974440264842
154
+ - type: manhattan_spearman
155
+ value: 83.72260506292083
156
+ - task:
157
+ type: STS
158
+ dataset:
159
+ type: mteb/sts17-crosslingual-sts
160
+ name: MTEB STS17 (en-en)
161
+ config: en-en
162
+ split: test
163
+ revision: af5e6fb845001ecf41f4c1e033ce921939a2a68d
164
+ metrics:
165
+ - type: cos_sim_pearson
166
+ value: 88.29744487677577
167
+ - type: cos_sim_spearman
168
+ value: 88.50799779856109
169
+ - type: euclidean_pearson
170
+ value: 89.0149154609955
171
+ - type: euclidean_spearman
172
+ value: 88.72798794474068
173
+ - type: manhattan_pearson
174
+ value: 89.14318227078863
175
+ - type: manhattan_spearman
176
+ value: 88.98372697017017
177
+ - task:
178
+ type: STS
179
+ dataset:
180
+ type: mteb/sts22-crosslingual-sts
181
+ name: MTEB STS22 (en)
182
+ config: en
183
+ split: test
184
+ revision: 6d1ba47164174a496b7fa5d3569dae26a6813b80
185
+ metrics:
186
+ - type: cos_sim_pearson
187
+ value: 70.114540107077
188
+ - type: cos_sim_spearman
189
+ value: 69.72244488054433
190
+ - type: euclidean_pearson
191
+ value: 70.03658853094686
192
+ - type: euclidean_spearman
193
+ value: 68.96035610557085
194
+ - type: manhattan_pearson
195
+ value: 69.83707789686764
196
+ - type: manhattan_spearman
197
+ value: 68.71831797289812
198
+ - task:
199
+ type: STS
200
+ dataset:
201
+ type: mteb/stsbenchmark-sts
202
+ name: MTEB STSBenchmark
203
+ config: default
204
+ split: test
205
+ revision: b0fddb56ed78048fa8b90373c8a3cfc37b684831
206
+ metrics:
207
+ - type: cos_sim_pearson
208
+ value: 84.86664469775837
209
+ - type: cos_sim_spearman
210
+ value: 85.39649452953681
211
+ - type: euclidean_pearson
212
+ value: 85.68509956626748
213
+ - type: euclidean_spearman
214
+ value: 85.50984027606854
215
+ - type: manhattan_pearson
216
+ value: 85.6688745008871
217
+ - type: manhattan_spearman
218
+ value: 85.465201888803
219
+ - task:
220
+ type: PairClassification
221
+ dataset:
222
+ type: mteb/sprintduplicatequestions-pairclassification
223
+ name: MTEB SprintDuplicateQuestions
224
+ config: default
225
+ split: test
226
+ revision: d66bd1f72af766a5cc4b0ca5e00c162f89e8cc46
227
+ metrics:
228
+ - type: cos_sim_accuracy
229
+ value: 99.8079207920792
230
+ - type: cos_sim_ap
231
+ value: 95.62897445718106
232
+ - type: cos_sim_f1
233
+ value: 90.03083247687564
234
+ - type: cos_sim_precision
235
+ value: 92.60042283298098
236
+ - type: cos_sim_recall
237
+ value: 87.6
238
+ - type: dot_accuracy
239
+ value: 99.67029702970297
240
+ - type: dot_ap
241
+ value: 90.20258347721159
242
+ - type: dot_f1
243
+ value: 83.06172839506172
244
+ - type: dot_precision
245
+ value: 82.04878048780488
246
+ - type: dot_recall
247
+ value: 84.1
248
+ - type: euclidean_accuracy
249
+ value: 99.80594059405941
250
+ - type: euclidean_ap
251
+ value: 95.53963697283662
252
+ - type: euclidean_f1
253
+ value: 89.92405063291139
254
+ - type: euclidean_precision
255
+ value: 91.07692307692308
256
+ - type: euclidean_recall
257
+ value: 88.8
258
+ - type: manhattan_accuracy
259
+ value: 99.80594059405941
260
+ - type: manhattan_ap
261
+ value: 95.55714505339634
262
+ - type: manhattan_f1
263
+ value: 90.06085192697769
264
+ - type: manhattan_precision
265
+ value: 91.35802469135803
266
+ - type: manhattan_recall
267
+ value: 88.8
268
+ - type: max_accuracy
269
+ value: 99.8079207920792
270
+ - type: max_ap
271
+ value: 95.62897445718106
272
+ - type: max_f1
273
+ value: 90.06085192697769
274
+ - task:
275
+ type: PairClassification
276
+ dataset:
277
+ type: mteb/twittersemeval2015-pairclassification
278
+ name: MTEB TwitterSemEval2015
279
+ config: default
280
+ split: test
281
+ revision: 70970daeab8776df92f5ea462b6173c0b46fd2d1
282
+ metrics:
283
+ - type: cos_sim_accuracy
284
+ value: 85.87351731537224
285
+ - type: cos_sim_ap
286
+ value: 72.87360532701162
287
+ - type: cos_sim_f1
288
+ value: 67.8826895565093
289
+ - type: cos_sim_precision
290
+ value: 61.918225315354505
291
+ - type: cos_sim_recall
292
+ value: 75.11873350923483
293
+ - type: dot_accuracy
294
+ value: 80.15139774691542
295
+ - type: dot_ap
296
+ value: 53.5201503222712
297
+ - type: dot_f1
298
+ value: 53.42203179614388
299
+ - type: dot_precision
300
+ value: 46.64303996849773
301
+ - type: dot_recall
302
+ value: 62.50659630606861
303
+ - type: euclidean_accuracy
304
+ value: 85.87351731537224
305
+ - type: euclidean_ap
306
+ value: 73.10465263888227
307
+ - type: euclidean_f1
308
+ value: 68.38209376101516
309
+ - type: euclidean_precision
310
+ value: 61.63948316034739
311
+ - type: euclidean_recall
312
+ value: 76.78100263852242
313
+ - type: manhattan_accuracy
314
+ value: 85.83775406806939
315
+ - type: manhattan_ap
316
+ value: 73.08358693248583
317
+ - type: manhattan_f1
318
+ value: 68.34053485927829
319
+ - type: manhattan_precision
320
+ value: 61.303163628745025
321
+ - type: manhattan_recall
322
+ value: 77.20316622691293
323
+ - type: max_accuracy
324
+ value: 85.87351731537224
325
+ - type: max_ap
326
+ value: 73.10465263888227
327
+ - type: max_f1
328
+ value: 68.38209376101516
329
+ - task:
330
+ type: PairClassification
331
+ dataset:
332
+ type: mteb/twitterurlcorpus-pairclassification
333
+ name: MTEB TwitterURLCorpus
334
+ config: default
335
+ split: test
336
+ revision: 8b6510b0b1fa4e4c4f879467980e9be563ec1cdf
337
+ metrics:
338
+ - type: cos_sim_accuracy
339
+ value: 88.85202002561415
340
+ - type: cos_sim_ap
341
+ value: 85.58170945333845
342
+ - type: cos_sim_f1
343
+ value: 77.87783280804442
344
+ - type: cos_sim_precision
345
+ value: 75.95140515222482
346
+ - type: cos_sim_recall
347
+ value: 79.90452725592854
348
+ - type: dot_accuracy
349
+ value: 85.29902588582296
350
+ - type: dot_ap
351
+ value: 76.95795800483633
352
+ - type: dot_f1
353
+ value: 71.30231900452489
354
+ - type: dot_precision
355
+ value: 65.91503267973856
356
+ - type: dot_recall
357
+ value: 77.6485987064983
358
+ - type: euclidean_accuracy
359
+ value: 88.80738929638684
360
+ - type: euclidean_ap
361
+ value: 85.5344499509856
362
+ - type: euclidean_f1
363
+ value: 77.9805854353285
364
+ - type: euclidean_precision
365
+ value: 75.97312495435624
366
+ - type: euclidean_recall
367
+ value: 80.09701262704034
368
+ - type: manhattan_accuracy
369
+ value: 88.7782822990647
370
+ - type: manhattan_ap
371
+ value: 85.52577812395661
372
+ - type: manhattan_f1
373
+ value: 77.97958958110746
374
+ - type: manhattan_precision
375
+ value: 74.76510067114094
376
+ - type: manhattan_recall
377
+ value: 81.48290729904527
378
+ - type: max_accuracy
379
+ value: 88.85202002561415
380
+ - type: max_ap
381
+ value: 85.58170945333845
382
+ - type: max_f1
383
+ value: 77.9805854353285
384
  license: mit
385
  language:
386
  - en