mrm8488 commited on
Commit
a98256c
1 Parent(s): c15652e

Add new SentenceTransformer model.

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 1024,
3
+ "pooling_mode_cls_token": false,
4
+ "pooling_mode_mean_tokens": true,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
README.md ADDED
@@ -0,0 +1,985 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: []
3
+ library_name: sentence-transformers
4
+ tags:
5
+ - sentence-transformers
6
+ - sentence-similarity
7
+ - feature-extraction
8
+ - dataset_size:1K<n<10K
9
+ - loss:MatryoshkaLoss
10
+ - loss:CoSENTLoss
11
+ base_model: intfloat/multilingual-e5-large
12
+ metrics:
13
+ - pearson_cosine
14
+ - spearman_cosine
15
+ - pearson_manhattan
16
+ - spearman_manhattan
17
+ - pearson_euclidean
18
+ - spearman_euclidean
19
+ - pearson_dot
20
+ - spearman_dot
21
+ - pearson_max
22
+ - spearman_max
23
+ widget:
24
+ - source_sentence: El hombre captura una pelota
25
+ sentences:
26
+ - Un hombre lanza una pelota en el aire.
27
+ - Un hombre se encuentra tocando una flauta de madera.
28
+ - La mujer está maquillándose usando sombra de ojos.
29
+ - source_sentence: Un hombre está buscando algo.
30
+ sentences:
31
+ - En un mercado de granjeros, se encuentra un hombre.
32
+ - Se acerca a la pista un avión suizo de color blanco.
33
+ - dos chicas jóvenes se abrazan en la hierba.
34
+ - source_sentence: El avión está tocando tierra.
35
+ sentences:
36
+ - El avión animado se encuentra en proceso de aterrizaje.
37
+ - La capital de Siria fue golpeada por dos explosiones
38
+ - Violentos incidentes afectan a estudiantes chinos en Francia
39
+ - source_sentence: Un hombre saltando la cuerda.
40
+ sentences:
41
+ - Un hombre está saltando la cuerda.
42
+ - Una mujer entrena a su perro para saltar en el aire.
43
+ - Los gatitos están comiendo de los platos.
44
+ - source_sentence: tres perros gruñendo entre sí
45
+ sentences:
46
+ - Dos perros se aproximan uno al otro en el pasto.
47
+ - Una mujer sonriente brinda cariño a un pequeño bebé.
48
+ - Una mujer está montando a caballo en el campo.
49
+ pipeline_tag: sentence-similarity
50
+ model-index:
51
+ - name: SentenceTransformer based on intfloat/multilingual-e5-large
52
+ results:
53
+ - task:
54
+ type: semantic-similarity
55
+ name: Semantic Similarity
56
+ dataset:
57
+ name: sts dev 768
58
+ type: sts-dev-768
59
+ metrics:
60
+ - type: pearson_cosine
61
+ value: 0.8279951103268512
62
+ name: Pearson Cosine
63
+ - type: spearman_cosine
64
+ value: 0.8342643795984531
65
+ name: Spearman Cosine
66
+ - type: pearson_manhattan
67
+ value: 0.8228439538329566
68
+ name: Pearson Manhattan
69
+ - type: spearman_manhattan
70
+ value: 0.834870903153992
71
+ name: Spearman Manhattan
72
+ - type: pearson_euclidean
73
+ value: 0.8231076969394738
74
+ name: Pearson Euclidean
75
+ - type: spearman_euclidean
76
+ value: 0.8349270059177344
77
+ name: Spearman Euclidean
78
+ - type: pearson_dot
79
+ value: 0.8196281042113861
80
+ name: Pearson Dot
81
+ - type: spearman_dot
82
+ value: 0.8248683461954115
83
+ name: Spearman Dot
84
+ - type: pearson_max
85
+ value: 0.8279951103268512
86
+ name: Pearson Max
87
+ - type: spearman_max
88
+ value: 0.8349270059177344
89
+ name: Spearman Max
90
+ - task:
91
+ type: semantic-similarity
92
+ name: Semantic Similarity
93
+ dataset:
94
+ name: sts dev 512
95
+ type: sts-dev-512
96
+ metrics:
97
+ - type: pearson_cosine
98
+ value: 0.8236357426336446
99
+ name: Pearson Cosine
100
+ - type: spearman_cosine
101
+ value: 0.8332692872015282
102
+ name: Spearman Cosine
103
+ - type: pearson_manhattan
104
+ value: 0.8217552769156274
105
+ name: Pearson Manhattan
106
+ - type: spearman_manhattan
107
+ value: 0.8331746060276878
108
+ name: Spearman Manhattan
109
+ - type: pearson_euclidean
110
+ value: 0.8217859136681092
111
+ name: Pearson Euclidean
112
+ - type: spearman_euclidean
113
+ value: 0.8334069456110773
114
+ name: Spearman Euclidean
115
+ - type: pearson_dot
116
+ value: 0.8101789790612713
117
+ name: Pearson Dot
118
+ - type: spearman_dot
119
+ value: 0.8179205607773823
120
+ name: Spearman Dot
121
+ - type: pearson_max
122
+ value: 0.8236357426336446
123
+ name: Pearson Max
124
+ - type: spearman_max
125
+ value: 0.8334069456110773
126
+ name: Spearman Max
127
+ - task:
128
+ type: semantic-similarity
129
+ name: Semantic Similarity
130
+ dataset:
131
+ name: sts dev 256
132
+ type: sts-dev-256
133
+ metrics:
134
+ - type: pearson_cosine
135
+ value: 0.816222860848086
136
+ name: Pearson Cosine
137
+ - type: spearman_cosine
138
+ value: 0.8303708513421737
139
+ name: Spearman Cosine
140
+ - type: pearson_manhattan
141
+ value: 0.8178715987143794
142
+ name: Pearson Manhattan
143
+ - type: spearman_manhattan
144
+ value: 0.8301047046554985
145
+ name: Spearman Manhattan
146
+ - type: pearson_euclidean
147
+ value: 0.8183826652089494
148
+ name: Pearson Euclidean
149
+ - type: spearman_euclidean
150
+ value: 0.8301804247624904
151
+ name: Spearman Euclidean
152
+ - type: pearson_dot
153
+ value: 0.7878741921967743
154
+ name: Pearson Dot
155
+ - type: spearman_dot
156
+ value: 0.7904844114269662
157
+ name: Spearman Dot
158
+ - type: pearson_max
159
+ value: 0.8183826652089494
160
+ name: Pearson Max
161
+ - type: spearman_max
162
+ value: 0.8303708513421737
163
+ name: Spearman Max
164
+ - task:
165
+ type: semantic-similarity
166
+ name: Semantic Similarity
167
+ dataset:
168
+ name: sts dev 128
169
+ type: sts-dev-128
170
+ metrics:
171
+ - type: pearson_cosine
172
+ value: 0.794202606017138
173
+ name: Pearson Cosine
174
+ - type: spearman_cosine
175
+ value: 0.8198385906414491
176
+ name: Spearman Cosine
177
+ - type: pearson_manhattan
178
+ value: 0.8088714046889546
179
+ name: Pearson Manhattan
180
+ - type: spearman_manhattan
181
+ value: 0.8222921243120748
182
+ name: Spearman Manhattan
183
+ - type: pearson_euclidean
184
+ value: 0.8092312345267045
185
+ name: Pearson Euclidean
186
+ - type: spearman_euclidean
187
+ value: 0.8220266161646009
188
+ name: Spearman Euclidean
189
+ - type: pearson_dot
190
+ value: 0.7341586721030032
191
+ name: Pearson Dot
192
+ - type: spearman_dot
193
+ value: 0.7351749794310246
194
+ name: Spearman Dot
195
+ - type: pearson_max
196
+ value: 0.8092312345267045
197
+ name: Pearson Max
198
+ - type: spearman_max
199
+ value: 0.8222921243120748
200
+ name: Spearman Max
201
+ - task:
202
+ type: semantic-similarity
203
+ name: Semantic Similarity
204
+ dataset:
205
+ name: sts dev 64
206
+ type: sts-dev-64
207
+ metrics:
208
+ - type: pearson_cosine
209
+ value: 0.7727295051414095
210
+ name: Pearson Cosine
211
+ - type: spearman_cosine
212
+ value: 0.8076629783565549
213
+ name: Spearman Cosine
214
+ - type: pearson_manhattan
215
+ value: 0.7976419723073269
216
+ name: Pearson Manhattan
217
+ - type: spearman_manhattan
218
+ value: 0.8147883308842346
219
+ name: Spearman Manhattan
220
+ - type: pearson_euclidean
221
+ value: 0.7979124462870892
222
+ name: Pearson Euclidean
223
+ - type: spearman_euclidean
224
+ value: 0.8123832197697319
225
+ name: Spearman Euclidean
226
+ - type: pearson_dot
227
+ value: 0.6725844492342726
228
+ name: Pearson Dot
229
+ - type: spearman_dot
230
+ value: 0.6673162832940408
231
+ name: Spearman Dot
232
+ - type: pearson_max
233
+ value: 0.7979124462870892
234
+ name: Pearson Max
235
+ - type: spearman_max
236
+ value: 0.8147883308842346
237
+ name: Spearman Max
238
+ - task:
239
+ type: semantic-similarity
240
+ name: Semantic Similarity
241
+ dataset:
242
+ name: sts test 768
243
+ type: sts-test-768
244
+ metrics:
245
+ - type: pearson_cosine
246
+ value: 0.8630482725201897
247
+ name: Pearson Cosine
248
+ - type: spearman_cosine
249
+ value: 0.8813284718659181
250
+ name: Spearman Cosine
251
+ - type: pearson_manhattan
252
+ value: 0.8770818288812614
253
+ name: Pearson Manhattan
254
+ - type: spearman_manhattan
255
+ value: 0.8810971983428288
256
+ name: Spearman Manhattan
257
+ - type: pearson_euclidean
258
+ value: 0.8770132070253477
259
+ name: Pearson Euclidean
260
+ - type: spearman_euclidean
261
+ value: 0.8812162173545179
262
+ name: Spearman Euclidean
263
+ - type: pearson_dot
264
+ value: 0.8581811981775829
265
+ name: Pearson Dot
266
+ - type: spearman_dot
267
+ value: 0.8707402246720045
268
+ name: Spearman Dot
269
+ - type: pearson_max
270
+ value: 0.8770818288812614
271
+ name: Pearson Max
272
+ - type: spearman_max
273
+ value: 0.8813284718659181
274
+ name: Spearman Max
275
+ - task:
276
+ type: semantic-similarity
277
+ name: Semantic Similarity
278
+ dataset:
279
+ name: sts test 512
280
+ type: sts-test-512
281
+ metrics:
282
+ - type: pearson_cosine
283
+ value: 0.8589909139210625
284
+ name: Pearson Cosine
285
+ - type: spearman_cosine
286
+ value: 0.8799604919891442
287
+ name: Spearman Cosine
288
+ - type: pearson_manhattan
289
+ value: 0.8744468387217347
290
+ name: Pearson Manhattan
291
+ - type: spearman_manhattan
292
+ value: 0.8791142262015441
293
+ name: Spearman Manhattan
294
+ - type: pearson_euclidean
295
+ value: 0.8747974723064821
296
+ name: Pearson Euclidean
297
+ - type: spearman_euclidean
298
+ value: 0.8795698184784307
299
+ name: Spearman Euclidean
300
+ - type: pearson_dot
301
+ value: 0.8464185524060444
302
+ name: Pearson Dot
303
+ - type: spearman_dot
304
+ value: 0.8549652098582826
305
+ name: Spearman Dot
306
+ - type: pearson_max
307
+ value: 0.8747974723064821
308
+ name: Pearson Max
309
+ - type: spearman_max
310
+ value: 0.8799604919891442
311
+ name: Spearman Max
312
+ - task:
313
+ type: semantic-similarity
314
+ name: Semantic Similarity
315
+ dataset:
316
+ name: sts test 256
317
+ type: sts-test-256
318
+ metrics:
319
+ - type: pearson_cosine
320
+ value: 0.8528262537030415
321
+ name: Pearson Cosine
322
+ - type: spearman_cosine
323
+ value: 0.8762917275750132
324
+ name: Spearman Cosine
325
+ - type: pearson_manhattan
326
+ value: 0.8715060008387856
327
+ name: Pearson Manhattan
328
+ - type: spearman_manhattan
329
+ value: 0.8780718380107112
330
+ name: Spearman Manhattan
331
+ - type: pearson_euclidean
332
+ value: 0.87251419758469
333
+ name: Pearson Euclidean
334
+ - type: spearman_euclidean
335
+ value: 0.8788770265821976
336
+ name: Spearman Euclidean
337
+ - type: pearson_dot
338
+ value: 0.801980870958869
339
+ name: Pearson Dot
340
+ - type: spearman_dot
341
+ value: 0.8007112694661982
342
+ name: Spearman Dot
343
+ - type: pearson_max
344
+ value: 0.87251419758469
345
+ name: Pearson Max
346
+ - type: spearman_max
347
+ value: 0.8788770265821976
348
+ name: Spearman Max
349
+ - task:
350
+ type: semantic-similarity
351
+ name: Semantic Similarity
352
+ dataset:
353
+ name: sts test 128
354
+ type: sts-test-128
355
+ metrics:
356
+ - type: pearson_cosine
357
+ value: 0.8392066286150661
358
+ name: Pearson Cosine
359
+ - type: spearman_cosine
360
+ value: 0.8692426944903685
361
+ name: Spearman Cosine
362
+ - type: pearson_manhattan
363
+ value: 0.8631603748425567
364
+ name: Pearson Manhattan
365
+ - type: spearman_manhattan
366
+ value: 0.8715673768304316
367
+ name: Spearman Manhattan
368
+ - type: pearson_euclidean
369
+ value: 0.8643871758114816
370
+ name: Pearson Euclidean
371
+ - type: spearman_euclidean
372
+ value: 0.8724091426441261
373
+ name: Spearman Euclidean
374
+ - type: pearson_dot
375
+ value: 0.7461565194503229
376
+ name: Pearson Dot
377
+ - type: spearman_dot
378
+ value: 0.7403017354497338
379
+ name: Spearman Dot
380
+ - type: pearson_max
381
+ value: 0.8643871758114816
382
+ name: Pearson Max
383
+ - type: spearman_max
384
+ value: 0.8724091426441261
385
+ name: Spearman Max
386
+ - task:
387
+ type: semantic-similarity
388
+ name: Semantic Similarity
389
+ dataset:
390
+ name: sts test 64
391
+ type: sts-test-64
392
+ metrics:
393
+ - type: pearson_cosine
394
+ value: 0.8213671607347727
395
+ name: Pearson Cosine
396
+ - type: spearman_cosine
397
+ value: 0.8621003145087452
398
+ name: Spearman Cosine
399
+ - type: pearson_manhattan
400
+ value: 0.8530869243121955
401
+ name: Pearson Manhattan
402
+ - type: spearman_manhattan
403
+ value: 0.8631973638935834
404
+ name: Spearman Manhattan
405
+ - type: pearson_euclidean
406
+ value: 0.854140567169475
407
+ name: Pearson Euclidean
408
+ - type: spearman_euclidean
409
+ value: 0.8632627342101252
410
+ name: Spearman Euclidean
411
+ - type: pearson_dot
412
+ value: 0.6853599968011839
413
+ name: Pearson Dot
414
+ - type: spearman_dot
415
+ value: 0.6726454086764928
416
+ name: Spearman Dot
417
+ - type: pearson_max
418
+ value: 0.854140567169475
419
+ name: Pearson Max
420
+ - type: spearman_max
421
+ value: 0.8632627342101252
422
+ name: Spearman Max
423
+ ---
424
+
425
+ # SentenceTransformer based on intfloat/multilingual-e5-large
426
+
427
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [intfloat/multilingual-e5-large](https://huggingface.co/intfloat/multilingual-e5-large) on the clibrain/stsb_multi_es_aug_gpt3.5-turbo_2 dataset. It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
428
+
429
+ ## Model Details
430
+
431
+ ### Model Description
432
+ - **Model Type:** Sentence Transformer
433
+ - **Base model:** [intfloat/multilingual-e5-large](https://huggingface.co/intfloat/multilingual-e5-large) <!-- at revision ab10c1a7f42e74530fe7ae5be82e6d4f11a719eb -->
434
+ - **Maximum Sequence Length:** 512 tokens
435
+ - **Output Dimensionality:** 1024 tokens
436
+ - **Similarity Function:** Cosine Similarity
437
+ - **Training Dataset:**
438
+ - clibrain/stsb_multi_es_aug_gpt3.5-turbo_2
439
+ <!-- - **Language:** Unknown -->
440
+ <!-- - **License:** Unknown -->
441
+
442
+ ### Model Sources
443
+
444
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
445
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
446
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
447
+
448
+ ### Full Model Architecture
449
+
450
+ ```
451
+ SentenceTransformer(
452
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: XLMRobertaModel
453
+ (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
454
+ (2): Normalize()
455
+ )
456
+ ```
457
+
458
+ ## Usage
459
+
460
+ ### Direct Usage (Sentence Transformers)
461
+
462
+ First install the Sentence Transformers library:
463
+
464
+ ```bash
465
+ pip install -U sentence-transformers
466
+ ```
467
+
468
+ Then you can load this model and run inference.
469
+ ```python
470
+ from sentence_transformers import SentenceTransformer
471
+
472
+ # Download from the 🤗 Hub
473
+ model = SentenceTransformer("mrm8488/multilingual-e5-large-ft-sts-spanish-matryoshka-768-64-5e")
474
+ # Run inference
475
+ sentences = [
476
+ 'tres perros gruñendo entre sí',
477
+ 'Dos perros se aproximan uno al otro en el pasto.',
478
+ 'Una mujer sonriente brinda cariño a un pequeño bebé.',
479
+ ]
480
+ embeddings = model.encode(sentences)
481
+ print(embeddings.shape)
482
+ # [3, 1024]
483
+
484
+ # Get the similarity scores for the embeddings
485
+ similarities = model.similarity(embeddings, embeddings)
486
+ print(similarities.shape)
487
+ # [3, 3]
488
+ ```
489
+
490
+ <!--
491
+ ### Direct Usage (Transformers)
492
+
493
+ <details><summary>Click to see the direct usage in Transformers</summary>
494
+
495
+ </details>
496
+ -->
497
+
498
+ <!--
499
+ ### Downstream Usage (Sentence Transformers)
500
+
501
+ You can finetune this model on your own dataset.
502
+
503
+ <details><summary>Click to expand</summary>
504
+
505
+ </details>
506
+ -->
507
+
508
+ <!--
509
+ ### Out-of-Scope Use
510
+
511
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
512
+ -->
513
+
514
+ ## Evaluation
515
+
516
+ ### Metrics
517
+
518
+ #### Semantic Similarity
519
+ * Dataset: `sts-dev-768`
520
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
521
+
522
+ | Metric | Value |
523
+ |:--------------------|:-----------|
524
+ | pearson_cosine | 0.828 |
525
+ | **spearman_cosine** | **0.8343** |
526
+ | pearson_manhattan | 0.8228 |
527
+ | spearman_manhattan | 0.8349 |
528
+ | pearson_euclidean | 0.8231 |
529
+ | spearman_euclidean | 0.8349 |
530
+ | pearson_dot | 0.8196 |
531
+ | spearman_dot | 0.8249 |
532
+ | pearson_max | 0.828 |
533
+ | spearman_max | 0.8349 |
534
+
535
+ #### Semantic Similarity
536
+ * Dataset: `sts-dev-512`
537
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
538
+
539
+ | Metric | Value |
540
+ |:--------------------|:-----------|
541
+ | pearson_cosine | 0.8236 |
542
+ | **spearman_cosine** | **0.8333** |
543
+ | pearson_manhattan | 0.8218 |
544
+ | spearman_manhattan | 0.8332 |
545
+ | pearson_euclidean | 0.8218 |
546
+ | spearman_euclidean | 0.8334 |
547
+ | pearson_dot | 0.8102 |
548
+ | spearman_dot | 0.8179 |
549
+ | pearson_max | 0.8236 |
550
+ | spearman_max | 0.8334 |
551
+
552
+ #### Semantic Similarity
553
+ * Dataset: `sts-dev-256`
554
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
555
+
556
+ | Metric | Value |
557
+ |:--------------------|:-----------|
558
+ | pearson_cosine | 0.8162 |
559
+ | **spearman_cosine** | **0.8304** |
560
+ | pearson_manhattan | 0.8179 |
561
+ | spearman_manhattan | 0.8301 |
562
+ | pearson_euclidean | 0.8184 |
563
+ | spearman_euclidean | 0.8302 |
564
+ | pearson_dot | 0.7879 |
565
+ | spearman_dot | 0.7905 |
566
+ | pearson_max | 0.8184 |
567
+ | spearman_max | 0.8304 |
568
+
569
+ #### Semantic Similarity
570
+ * Dataset: `sts-dev-128`
571
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
572
+
573
+ | Metric | Value |
574
+ |:--------------------|:-----------|
575
+ | pearson_cosine | 0.7942 |
576
+ | **spearman_cosine** | **0.8198** |
577
+ | pearson_manhattan | 0.8089 |
578
+ | spearman_manhattan | 0.8223 |
579
+ | pearson_euclidean | 0.8092 |
580
+ | spearman_euclidean | 0.822 |
581
+ | pearson_dot | 0.7342 |
582
+ | spearman_dot | 0.7352 |
583
+ | pearson_max | 0.8092 |
584
+ | spearman_max | 0.8223 |
585
+
586
+ #### Semantic Similarity
587
+ * Dataset: `sts-dev-64`
588
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
589
+
590
+ | Metric | Value |
591
+ |:--------------------|:-----------|
592
+ | pearson_cosine | 0.7727 |
593
+ | **spearman_cosine** | **0.8077** |
594
+ | pearson_manhattan | 0.7976 |
595
+ | spearman_manhattan | 0.8148 |
596
+ | pearson_euclidean | 0.7979 |
597
+ | spearman_euclidean | 0.8124 |
598
+ | pearson_dot | 0.6726 |
599
+ | spearman_dot | 0.6673 |
600
+ | pearson_max | 0.7979 |
601
+ | spearman_max | 0.8148 |
602
+
603
+ #### Semantic Similarity
604
+ * Dataset: `sts-test-768`
605
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
606
+
607
+ | Metric | Value |
608
+ |:--------------------|:-----------|
609
+ | pearson_cosine | 0.863 |
610
+ | **spearman_cosine** | **0.8813** |
611
+ | pearson_manhattan | 0.8771 |
612
+ | spearman_manhattan | 0.8811 |
613
+ | pearson_euclidean | 0.877 |
614
+ | spearman_euclidean | 0.8812 |
615
+ | pearson_dot | 0.8582 |
616
+ | spearman_dot | 0.8707 |
617
+ | pearson_max | 0.8771 |
618
+ | spearman_max | 0.8813 |
619
+
620
+ #### Semantic Similarity
621
+ * Dataset: `sts-test-512`
622
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
623
+
624
+ | Metric | Value |
625
+ |:--------------------|:---------|
626
+ | pearson_cosine | 0.859 |
627
+ | **spearman_cosine** | **0.88** |
628
+ | pearson_manhattan | 0.8744 |
629
+ | spearman_manhattan | 0.8791 |
630
+ | pearson_euclidean | 0.8748 |
631
+ | spearman_euclidean | 0.8796 |
632
+ | pearson_dot | 0.8464 |
633
+ | spearman_dot | 0.855 |
634
+ | pearson_max | 0.8748 |
635
+ | spearman_max | 0.88 |
636
+
637
+ #### Semantic Similarity
638
+ * Dataset: `sts-test-256`
639
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
640
+
641
+ | Metric | Value |
642
+ |:--------------------|:-----------|
643
+ | pearson_cosine | 0.8528 |
644
+ | **spearman_cosine** | **0.8763** |
645
+ | pearson_manhattan | 0.8715 |
646
+ | spearman_manhattan | 0.8781 |
647
+ | pearson_euclidean | 0.8725 |
648
+ | spearman_euclidean | 0.8789 |
649
+ | pearson_dot | 0.802 |
650
+ | spearman_dot | 0.8007 |
651
+ | pearson_max | 0.8725 |
652
+ | spearman_max | 0.8789 |
653
+
654
+ #### Semantic Similarity
655
+ * Dataset: `sts-test-128`
656
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
657
+
658
+ | Metric | Value |
659
+ |:--------------------|:-----------|
660
+ | pearson_cosine | 0.8392 |
661
+ | **spearman_cosine** | **0.8692** |
662
+ | pearson_manhattan | 0.8632 |
663
+ | spearman_manhattan | 0.8716 |
664
+ | pearson_euclidean | 0.8644 |
665
+ | spearman_euclidean | 0.8724 |
666
+ | pearson_dot | 0.7462 |
667
+ | spearman_dot | 0.7403 |
668
+ | pearson_max | 0.8644 |
669
+ | spearman_max | 0.8724 |
670
+
671
+ #### Semantic Similarity
672
+ * Dataset: `sts-test-64`
673
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
674
+
675
+ | Metric | Value |
676
+ |:--------------------|:-----------|
677
+ | pearson_cosine | 0.8214 |
678
+ | **spearman_cosine** | **0.8621** |
679
+ | pearson_manhattan | 0.8531 |
680
+ | spearman_manhattan | 0.8632 |
681
+ | pearson_euclidean | 0.8541 |
682
+ | spearman_euclidean | 0.8633 |
683
+ | pearson_dot | 0.6854 |
684
+ | spearman_dot | 0.6726 |
685
+ | pearson_max | 0.8541 |
686
+ | spearman_max | 0.8633 |
687
+
688
+ <!--
689
+ ## Bias, Risks and Limitations
690
+
691
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
692
+ -->
693
+
694
+ <!--
695
+ ### Recommendations
696
+
697
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
698
+ -->
699
+
700
+ ## Training Details
701
+
702
+ ### Training Dataset
703
+
704
+ #### clibrain/stsb_multi_es_aug_gpt3.5-turbo_2
705
+
706
+ * Dataset: clibrain/stsb_multi_es_aug_gpt3.5-turbo_2
707
+ * Size: 2,697 training samples
708
+ * Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>score</code>
709
+ * Approximate statistics based on the first 1000 samples:
710
+ | | sentence1 | sentence2 | score |
711
+ |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:---------------------------------------------------------------|
712
+ | type | string | string | float |
713
+ | details | <ul><li>min: 8 tokens</li><li>mean: 22.25 tokens</li><li>max: 68 tokens</li></ul> | <ul><li>min: 8 tokens</li><li>mean: 22.01 tokens</li><li>max: 79 tokens</li></ul> | <ul><li>min: 0.0</li><li>mean: 2.67</li><li>max: 5.0</li></ul> |
714
+ * Samples:
715
+ | sentence1 | sentence2 | score |
716
+ |:------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------|:-------------------------------|
717
+ | <code>El pájaro de tamaño reducido se posó con delicadeza en una rama cubierta de escarcha.</code> | <code>Un ave de color amarillo descansaba tranquilamente en una rama.</code> | <code>3.200000047683716</code> |
718
+ | <code>Una chica está tocando la flauta en un parque.</code> | <code>Un grupo de músicos está tocando en un escenario al aire libre.</code> | <code>1.286</code> |
719
+ | <code>La aclamada escritora británica, Doris Lessing, galardonada con el premio Nobel, fallece</code> | <code>La destacada autora británica, Doris Lessing, reconocida con el prestigioso Premio Nobel, muere</code> | <code>4.199999809265137</code> |
720
+ * Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
721
+ ```json
722
+ {
723
+ "loss": "CoSENTLoss",
724
+ "matryoshka_dims": [
725
+ 768,
726
+ 512,
727
+ 256,
728
+ 128,
729
+ 64
730
+ ],
731
+ "matryoshka_weights": [
732
+ 1,
733
+ 1,
734
+ 1,
735
+ 1,
736
+ 1
737
+ ],
738
+ "n_dims_per_step": -1
739
+ }
740
+ ```
741
+
742
+ ### Evaluation Dataset
743
+
744
+ #### clibrain/stsb_multi_es_aug_gpt3.5-turbo_2
745
+
746
+ * Dataset: clibrain/stsb_multi_es_aug_gpt3.5-turbo_2
747
+ * Size: 697 evaluation samples
748
+ * Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>score</code>
749
+ * Approximate statistics based on the first 1000 samples:
750
+ | | sentence1 | sentence2 | score |
751
+ |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------|
752
+ | type | string | string | float |
753
+ | details | <ul><li>min: 8 tokens</li><li>mean: 22.76 tokens</li><li>max: 67 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 22.26 tokens</li><li>max: 63 tokens</li></ul> | <ul><li>min: 0.0</li><li>mean: 2.3</li><li>max: 5.0</li></ul> |
754
+ * Samples:
755
+ | sentence1 | sentence2 | score |
756
+ |:--------------------------------------------------------------------------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------|:-------------------------------|
757
+ | <code>Un incendio ocurrido en un hospital psiquiátrico ruso resultó en la trágica muerte de 38 personas.</code> | <code>Se teme que el incendio en un hospital psiquiátrico ruso cause la pérdida de la vida de 38 individuos.</code> | <code>4.199999809265137</code> |
758
+ | <code>"Street dijo que el otro individuo a veces se siente avergonzado de su fiesta, lo cual provoca risas en la multitud"</code> | <code>"A veces, el otro tipo se encuentra avergonzado de su fiesta y no se le puede culpar."</code> | <code>3.5</code> |
759
+ | <code>El veterano diplomático de Malasia tuvo un encuentro con Suu Kyi el miércoles en la casa del lago en Yangon donde permanece bajo arresto domiciliario.</code> | <code>Razali Ismail tuvo una reunión de 90 minutos con Suu Kyi, quien ganó el Premio Nobel de la Paz en 1991, en su casa del lago donde está recluida.</code> | <code>3.691999912261963</code> |
760
+ * Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
761
+ ```json
762
+ {
763
+ "loss": "CoSENTLoss",
764
+ "matryoshka_dims": [
765
+ 768,
766
+ 512,
767
+ 256,
768
+ 128,
769
+ 64
770
+ ],
771
+ "matryoshka_weights": [
772
+ 1,
773
+ 1,
774
+ 1,
775
+ 1,
776
+ 1
777
+ ],
778
+ "n_dims_per_step": -1
779
+ }
780
+ ```
781
+
782
+ ### Training Hyperparameters
783
+ #### Non-Default Hyperparameters
784
+
785
+ - `eval_strategy`: steps
786
+ - `per_device_train_batch_size`: 16
787
+ - `per_device_eval_batch_size`: 16
788
+ - `num_train_epochs`: 5
789
+ - `warmup_ratio`: 0.1
790
+ - `fp16`: True
791
+
792
+ #### All Hyperparameters
793
+ <details><summary>Click to expand</summary>
794
+
795
+ - `overwrite_output_dir`: False
796
+ - `do_predict`: False
797
+ - `eval_strategy`: steps
798
+ - `prediction_loss_only`: True
799
+ - `per_device_train_batch_size`: 16
800
+ - `per_device_eval_batch_size`: 16
801
+ - `per_gpu_train_batch_size`: None
802
+ - `per_gpu_eval_batch_size`: None
803
+ - `gradient_accumulation_steps`: 1
804
+ - `eval_accumulation_steps`: None
805
+ - `learning_rate`: 5e-05
806
+ - `weight_decay`: 0.0
807
+ - `adam_beta1`: 0.9
808
+ - `adam_beta2`: 0.999
809
+ - `adam_epsilon`: 1e-08
810
+ - `max_grad_norm`: 1.0
811
+ - `num_train_epochs`: 5
812
+ - `max_steps`: -1
813
+ - `lr_scheduler_type`: linear
814
+ - `lr_scheduler_kwargs`: {}
815
+ - `warmup_ratio`: 0.1
816
+ - `warmup_steps`: 0
817
+ - `log_level`: passive
818
+ - `log_level_replica`: warning
819
+ - `log_on_each_node`: True
820
+ - `logging_nan_inf_filter`: True
821
+ - `save_safetensors`: True
822
+ - `save_on_each_node`: False
823
+ - `save_only_model`: False
824
+ - `restore_callback_states_from_checkpoint`: False
825
+ - `no_cuda`: False
826
+ - `use_cpu`: False
827
+ - `use_mps_device`: False
828
+ - `seed`: 42
829
+ - `data_seed`: None
830
+ - `jit_mode_eval`: False
831
+ - `use_ipex`: False
832
+ - `bf16`: False
833
+ - `fp16`: True
834
+ - `fp16_opt_level`: O1
835
+ - `half_precision_backend`: auto
836
+ - `bf16_full_eval`: False
837
+ - `fp16_full_eval`: False
838
+ - `tf32`: None
839
+ - `local_rank`: 0
840
+ - `ddp_backend`: None
841
+ - `tpu_num_cores`: None
842
+ - `tpu_metrics_debug`: False
843
+ - `debug`: []
844
+ - `dataloader_drop_last`: False
845
+ - `dataloader_num_workers`: 0
846
+ - `dataloader_prefetch_factor`: None
847
+ - `past_index`: -1
848
+ - `disable_tqdm`: False
849
+ - `remove_unused_columns`: True
850
+ - `label_names`: None
851
+ - `load_best_model_at_end`: False
852
+ - `ignore_data_skip`: False
853
+ - `fsdp`: []
854
+ - `fsdp_min_num_params`: 0
855
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
856
+ - `fsdp_transformer_layer_cls_to_wrap`: None
857
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
858
+ - `deepspeed`: None
859
+ - `label_smoothing_factor`: 0.0
860
+ - `optim`: adamw_torch
861
+ - `optim_args`: None
862
+ - `adafactor`: False
863
+ - `group_by_length`: False
864
+ - `length_column_name`: length
865
+ - `ddp_find_unused_parameters`: None
866
+ - `ddp_bucket_cap_mb`: None
867
+ - `ddp_broadcast_buffers`: False
868
+ - `dataloader_pin_memory`: True
869
+ - `dataloader_persistent_workers`: False
870
+ - `skip_memory_metrics`: True
871
+ - `use_legacy_prediction_loop`: False
872
+ - `push_to_hub`: False
873
+ - `resume_from_checkpoint`: None
874
+ - `hub_model_id`: None
875
+ - `hub_strategy`: every_save
876
+ - `hub_private_repo`: False
877
+ - `hub_always_push`: False
878
+ - `gradient_checkpointing`: False
879
+ - `gradient_checkpointing_kwargs`: None
880
+ - `include_inputs_for_metrics`: False
881
+ - `eval_do_concat_batches`: True
882
+ - `fp16_backend`: auto
883
+ - `push_to_hub_model_id`: None
884
+ - `push_to_hub_organization`: None
885
+ - `mp_parameters`:
886
+ - `auto_find_batch_size`: False
887
+ - `full_determinism`: False
888
+ - `torchdynamo`: None
889
+ - `ray_scope`: last
890
+ - `ddp_timeout`: 1800
891
+ - `torch_compile`: False
892
+ - `torch_compile_backend`: None
893
+ - `torch_compile_mode`: None
894
+ - `dispatch_batches`: None
895
+ - `split_batches`: None
896
+ - `include_tokens_per_second`: False
897
+ - `include_num_input_tokens_seen`: False
898
+ - `neftune_noise_alpha`: None
899
+ - `optim_target_modules`: None
900
+ - `batch_eval_metrics`: False
901
+ - `batch_sampler`: batch_sampler
902
+ - `multi_dataset_batch_sampler`: proportional
903
+
904
+ </details>
905
+
906
+ ### Training Logs
907
+ | Epoch | Step | Training Loss | loss | sts-dev-128_spearman_cosine | sts-dev-256_spearman_cosine | sts-dev-512_spearman_cosine | sts-dev-64_spearman_cosine | sts-dev-768_spearman_cosine | sts-test-128_spearman_cosine | sts-test-256_spearman_cosine | sts-test-512_spearman_cosine | sts-test-64_spearman_cosine | sts-test-768_spearman_cosine |
908
+ |:------:|:----:|:-------------:|:-------:|:---------------------------:|:---------------------------:|:---------------------------:|:--------------------------:|:---------------------------:|:----------------------------:|:----------------------------:|:----------------------------:|:---------------------------:|:----------------------------:|
909
+ | 0.5917 | 100 | 21.7032 | 21.7030 | 0.8030 | 0.8124 | 0.8205 | 0.7839 | 0.8215 | - | - | - | - | - |
910
+ | 1.1834 | 200 | 21.4019 | 24.0898 | 0.7839 | 0.7972 | 0.8038 | 0.7680 | 0.8062 | - | - | - | - | - |
911
+ | 1.7751 | 300 | 21.2168 | 22.5421 | 0.7909 | 0.8027 | 0.8058 | 0.7786 | 0.8068 | - | - | - | - | - |
912
+ | 2.3669 | 400 | 20.7049 | 23.6522 | 0.7938 | 0.8049 | 0.8108 | 0.7873 | 0.8123 | - | - | - | - | - |
913
+ | 2.9586 | 500 | 20.5077 | 23.6100 | 0.8017 | 0.8116 | 0.8155 | 0.7893 | 0.8185 | - | - | - | - | - |
914
+ | 3.5503 | 600 | 19.2725 | 24.7539 | 0.8133 | 0.8254 | 0.8291 | 0.8032 | 0.8314 | - | - | - | - | - |
915
+ | 4.1420 | 700 | 19.0841 | 26.5286 | 0.8210 | 0.8298 | 0.8333 | 0.8102 | 0.8333 | - | - | - | - | - |
916
+ | 4.7337 | 800 | 18.6847 | 26.8158 | 0.8198 | 0.8304 | 0.8333 | 0.8077 | 0.8343 | - | - | - | - | - |
917
+ | 5.0 | 845 | - | - | - | - | - | - | - | 0.8692 | 0.8763 | 0.8800 | 0.8621 | 0.8813 |
918
+
919
+
920
+ ### Framework Versions
921
+ - Python: 3.10.12
922
+ - Sentence Transformers: 3.0.0
923
+ - Transformers: 4.41.1
924
+ - PyTorch: 2.3.0+cu121
925
+ - Accelerate: 0.30.1
926
+ - Datasets: 2.19.1
927
+ - Tokenizers: 0.19.1
928
+
929
+ ## Citation
930
+
931
+ ### BibTeX
932
+
933
+ #### Sentence Transformers
934
+ ```bibtex
935
+ @inproceedings{reimers-2019-sentence-bert,
936
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
937
+ author = "Reimers, Nils and Gurevych, Iryna",
938
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
939
+ month = "11",
940
+ year = "2019",
941
+ publisher = "Association for Computational Linguistics",
942
+ url = "https://arxiv.org/abs/1908.10084",
943
+ }
944
+ ```
945
+
946
+ #### MatryoshkaLoss
947
+ ```bibtex
948
+ @misc{kusupati2024matryoshka,
949
+ title={Matryoshka Representation Learning},
950
+ author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
951
+ year={2024},
952
+ eprint={2205.13147},
953
+ archivePrefix={arXiv},
954
+ primaryClass={cs.LG}
955
+ }
956
+ ```
957
+
958
+ #### CoSENTLoss
959
+ ```bibtex
960
+ @online{kexuefm-8847,
961
+ title={CoSENT: A more efficient sentence vector scheme than Sentence-BERT},
962
+ author={Su Jianlin},
963
+ year={2022},
964
+ month={Jan},
965
+ url={https://kexue.fm/archives/8847},
966
+ }
967
+ ```
968
+
969
+ <!--
970
+ ## Glossary
971
+
972
+ *Clearly define terms in order to be accessible across audiences.*
973
+ -->
974
+
975
+ <!--
976
+ ## Model Card Authors
977
+
978
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
979
+ -->
980
+
981
+ <!--
982
+ ## Model Card Contact
983
+
984
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
985
+ -->
config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "intfloat/multilingual-e5-large",
3
+ "architectures": [
4
+ "XLMRobertaModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 4096,
15
+ "layer_norm_eps": 1e-05,
16
+ "max_position_embeddings": 514,
17
+ "model_type": "xlm-roberta",
18
+ "num_attention_heads": 16,
19
+ "num_hidden_layers": 24,
20
+ "output_past": true,
21
+ "pad_token_id": 1,
22
+ "position_embedding_type": "absolute",
23
+ "torch_dtype": "float32",
24
+ "transformers_version": "4.41.1",
25
+ "type_vocab_size": 1,
26
+ "use_cache": true,
27
+ "vocab_size": 250002
28
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "3.0.0",
4
+ "transformers": "4.41.1",
5
+ "pytorch": "2.3.0+cu121"
6
+ },
7
+ "prompts": {},
8
+ "default_prompt_name": null,
9
+ "similarity_fn_name": null
10
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a936de3306ace5e039eda3a274983deea2ab160c3532c04d599ec8a3a9b9e53
3
+ size 2239607176
modules.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ },
14
+ {
15
+ "idx": 2,
16
+ "name": "2",
17
+ "path": "2_Normalize",
18
+ "type": "sentence_transformers.models.Normalize"
19
+ }
20
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": false
4
+ }
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": true,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:883b037111086fd4dfebbbc9b7cee11e1517b5e0c0514879478661440f137085
3
+ size 17082987
tokenizer_config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "250001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": true,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "mask_token": "<mask>",
49
+ "model_max_length": 512,
50
+ "pad_token": "<pad>",
51
+ "sep_token": "</s>",
52
+ "tokenizer_class": "XLMRobertaTokenizer",
53
+ "unk_token": "<unk>"
54
+ }