Hugging Face's logo

tomaarsen
/
bert-base-uncased-gooaq

Sentence Similarity
sentence-transformers
Safetensors
English
bert
feature-extraction
Generated from Trainer
dataset_size:3012496
loss:MatryoshkaLoss
loss:CachedMultipleNegativesRankingLoss
Eval Results
text-embeddings-inference
Inference Endpoints
Model card Files Community
tomaarsen HF staff commited on
Commit
4911f66
•
1 Parent(s): a264527

Add new SentenceTransformer model

Browse files
Files changed (11) hide show
  1. 1_Pooling/config.json +10 -0
  2. README.md +1336 -0
  3. config.json +26 -0
  4. config_sentence_transformers.json +10 -0
  5. model.safetensors +3 -0
  6. modules.json +14 -0
  7. sentence_bert_config.json +4 -0
  8. special_tokens_map.json +7 -0
  9. tokenizer.json +0 -0
  10. tokenizer_config.json +55 -0
  11. vocab.txt +0 -0
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 768,
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 +1,1336 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: apache-2.0
5
+ tags:
6
+ - sentence-transformers
7
+ - sentence-similarity
8
+ - feature-extraction
9
+ - generated_from_trainer
10
+ - dataset_size:3012496
11
+ - loss:MatryoshkaLoss
12
+ - loss:CachedMultipleNegativesRankingLoss
13
+ base_model: google-bert/bert-base-uncased
14
+ widget:
15
+ - source_sentence: are the sequels better than the prequels?
16
+ sentences:
17
+ - '[''Automatically.'', ''When connected to car Bluetooth and,'', ''Manually.'']'
18
+ - The prequels are also not scared to take risks, making movies which are very different
19
+ from the original trilogy. The sequel saga, on the other hand, are technically
20
+ better made films, the acting is more consistent, the CGI is better and the writing
21
+ is stronger, however it falls down in many other places.
22
+ - While both public and private sectors use budgets as a key planning tool, public
23
+ bodies balance budgets, while private sector firms use budgets to predict operating
24
+ results. The public sector budget matches expenditures on mandated assets and
25
+ services with receipts of public money such as taxes and fees.
26
+ - source_sentence: are there bbqs at lake leschenaultia?
27
+ sentences:
28
+ - Vestavia Hills. The hummingbird, or, el zunzún as they are often called in the
29
+ Caribbean, have such a nickname because of their quick movements. The ruby-throated
30
+ hummingbird, the most commonly seen hummingbird in Alabama, is the inspiration
31
+ for this restaurant.
32
+ - Common causes of abdominal tenderness Abdominal tenderness is generally a sign
33
+ of inflammation or other acute processes in one or more organs. The organs are
34
+ located around the tender area. Acute processes mean sudden pressure caused by
35
+ something. For example, twisted or blocked organs can cause point tenderness.
36
+ - ​Located on 168 hectares of nature reserve, Lake Leschenaultia is the perfect
37
+ spot for a family day out in the Perth Hills. The Lake offers canoeing, swimming,
38
+ walk and cycle trails, as well as picnic, BBQ and camping facilities. ... There
39
+ are picnic tables set amongst lovely Wandoo trees.
40
+ - source_sentence: how much folic acid should you take prenatal?
41
+ sentences:
42
+ - Folic acid is a pregnancy superhero! Taking a prenatal vitamin with the recommended
43
+ 400 micrograms (mcg) of folic acid before and during pregnancy can help prevent
44
+ birth defects of your baby's brain and spinal cord. Take it every day and go ahead
45
+ and have a bowl of fortified cereal, too.
46
+ - '[''You must be unemployed through no fault of your own, as defined by Virginia
47
+ law.'', ''You must have earned at least a minimum amount in wages before you were
48
+ unemployed.'', ''You must be able and available to work, and you must be actively
49
+ seeking employment.'']'
50
+ - Wallpaper is printed in batches of rolls. It is important to have the same batch
51
+ number, to ensure colours match exactly. The batch number is usually located on
52
+ the wallpaper label close to the pattern number. Remember batch numbers also apply
53
+ to white wallpapers, as different batches can be different shades of white.
54
+ - source_sentence: what is the difference between minerals and electrolytes?
55
+ sentences:
56
+ - 'North: Just head north of Junk Junction like so. South: Head below Lucky Landing.
57
+ East: You''re basically landing between Lonely Lodge and the Racetrack. West:
58
+ The sign is west of Snobby Shores.'
59
+ - The fasting glucose tolerance test is the simplest and fastest way to measure
60
+ blood glucose and diagnose diabetes. Fasting means that you have had nothing to
61
+ eat or drink (except water) for 8 to 12 hours before the test.
62
+ - In other words, the term “electrolyte” typically implies ionized minerals dissolved
63
+ within water and beverages. Electrolytes are typically minerals, whereas minerals
64
+ may or may not be electrolytes.
65
+ - source_sentence: how can i download youtube videos with internet download manager?
66
+ sentences:
67
+ - '[''Go to settings and then click on extensions (top left side in chrome).'',
68
+ ''Minimise your browser and open the location (folder) where IDM is installed.
69
+ ... '', ''Find the file “IDMGCExt. ... '', ''Drag this file to your chrome browser
70
+ and drop to install the IDM extension.'']'
71
+ - Coca-Cola might rot your teeth and load your body with sugar and calories, but
72
+ it's actually an effective and safe first line of treatment for some stomach blockages,
73
+ researchers say.
74
+ - To fix a disabled iPhone or iPad without iTunes, you have to erase your device.
75
+ Click on the "Erase iPhone" option and confirm your selection. Wait for a while
76
+ as the "Find My iPhone" feature will remotely erase your iOS device. Needless
77
+ to say, it will also disable its lock.
78
+ datasets:
79
+ - sentence-transformers/gooaq
80
+ pipeline_tag: sentence-similarity
81
+ library_name: sentence-transformers
82
+ metrics:
83
+ - cosine_accuracy@1
84
+ - cosine_accuracy@3
85
+ - cosine_accuracy@5
86
+ - cosine_accuracy@10
87
+ - cosine_precision@1
88
+ - cosine_precision@3
89
+ - cosine_precision@5
90
+ - cosine_precision@10
91
+ - cosine_recall@1
92
+ - cosine_recall@3
93
+ - cosine_recall@5
94
+ - cosine_recall@10
95
+ - cosine_ndcg@10
96
+ - cosine_mrr@10
97
+ - cosine_map@100
98
+ co2_eq_emissions:
99
+ emissions: 242.52371141034885
100
+ energy_consumed: 0.623932244779674
101
+ source: codecarbon
102
+ training_type: fine-tuning
103
+ on_cloud: false
104
+ cpu_model: 13th Gen Intel(R) Core(TM) i7-13700K
105
+ ram_total_size: 31.777088165283203
106
+ hours_used: 1.619
107
+ hardware_used: 1 x NVIDIA GeForce RTX 3090
108
+ model-index:
109
+ - name: bert-base-uncased adapter finetuned on GooAQ pairs
110
+ results:
111
+ - task:
112
+ type: information-retrieval
113
+ name: Information Retrieval
114
+ dataset:
115
+ name: NanoClimateFEVER
116
+ type: NanoClimateFEVER
117
+ metrics:
118
+ - type: cosine_accuracy@1
119
+ value: 0.24
120
+ name: Cosine Accuracy@1
121
+ - type: cosine_accuracy@3
122
+ value: 0.42
123
+ name: Cosine Accuracy@3
124
+ - type: cosine_accuracy@5
125
+ value: 0.46
126
+ name: Cosine Accuracy@5
127
+ - type: cosine_accuracy@10
128
+ value: 0.56
129
+ name: Cosine Accuracy@10
130
+ - type: cosine_precision@1
131
+ value: 0.24
132
+ name: Cosine Precision@1
133
+ - type: cosine_precision@3
134
+ value: 0.15999999999999998
135
+ name: Cosine Precision@3
136
+ - type: cosine_precision@5
137
+ value: 0.10800000000000001
138
+ name: Cosine Precision@5
139
+ - type: cosine_precision@10
140
+ value: 0.07
141
+ name: Cosine Precision@10
142
+ - type: cosine_recall@1
143
+ value: 0.13166666666666665
144
+ name: Cosine Recall@1
145
+ - type: cosine_recall@3
146
+ value: 0.20833333333333337
147
+ name: Cosine Recall@3
148
+ - type: cosine_recall@5
149
+ value: 0.24166666666666664
150
+ name: Cosine Recall@5
151
+ - type: cosine_recall@10
152
+ value: 0.29666666666666663
153
+ name: Cosine Recall@10
154
+ - type: cosine_ndcg@10
155
+ value: 0.25516520961338873
156
+ name: Cosine Ndcg@10
157
+ - type: cosine_mrr@10
158
+ value: 0.3378809523809523
159
+ name: Cosine Mrr@10
160
+ - type: cosine_map@100
161
+ value: 0.20756281994556017
162
+ name: Cosine Map@100
163
+ - task:
164
+ type: information-retrieval
165
+ name: Information Retrieval
166
+ dataset:
167
+ name: NanoDBPedia
168
+ type: NanoDBPedia
169
+ metrics:
170
+ - type: cosine_accuracy@1
171
+ value: 0.54
172
+ name: Cosine Accuracy@1
173
+ - type: cosine_accuracy@3
174
+ value: 0.8
175
+ name: Cosine Accuracy@3
176
+ - type: cosine_accuracy@5
177
+ value: 0.84
178
+ name: Cosine Accuracy@5
179
+ - type: cosine_accuracy@10
180
+ value: 0.92
181
+ name: Cosine Accuracy@10
182
+ - type: cosine_precision@1
183
+ value: 0.54
184
+ name: Cosine Precision@1
185
+ - type: cosine_precision@3
186
+ value: 0.4866666666666667
187
+ name: Cosine Precision@3
188
+ - type: cosine_precision@5
189
+ value: 0.4440000000000001
190
+ name: Cosine Precision@5
191
+ - type: cosine_precision@10
192
+ value: 0.3899999999999999
193
+ name: Cosine Precision@10
194
+ - type: cosine_recall@1
195
+ value: 0.046781664425339056
196
+ name: Cosine Recall@1
197
+ - type: cosine_recall@3
198
+ value: 0.11117774881295754
199
+ name: Cosine Recall@3
200
+ - type: cosine_recall@5
201
+ value: 0.15829952609979633
202
+ name: Cosine Recall@5
203
+ - type: cosine_recall@10
204
+ value: 0.2554819210350403
205
+ name: Cosine Recall@10
206
+ - type: cosine_ndcg@10
207
+ value: 0.4644109757573673
208
+ name: Cosine Ndcg@10
209
+ - type: cosine_mrr@10
210
+ value: 0.6797460317460318
211
+ name: Cosine Mrr@10
212
+ - type: cosine_map@100
213
+ value: 0.3253011706807197
214
+ name: Cosine Map@100
215
+ - task:
216
+ type: information-retrieval
217
+ name: Information Retrieval
218
+ dataset:
219
+ name: NanoFEVER
220
+ type: NanoFEVER
221
+ metrics:
222
+ - type: cosine_accuracy@1
223
+ value: 0.54
224
+ name: Cosine Accuracy@1
225
+ - type: cosine_accuracy@3
226
+ value: 0.82
227
+ name: Cosine Accuracy@3
228
+ - type: cosine_accuracy@5
229
+ value: 0.9
230
+ name: Cosine Accuracy@5
231
+ - type: cosine_accuracy@10
232
+ value: 0.92
233
+ name: Cosine Accuracy@10
234
+ - type: cosine_precision@1
235
+ value: 0.54
236
+ name: Cosine Precision@1
237
+ - type: cosine_precision@3
238
+ value: 0.2733333333333333
239
+ name: Cosine Precision@3
240
+ - type: cosine_precision@5
241
+ value: 0.184
242
+ name: Cosine Precision@5
243
+ - type: cosine_precision@10
244
+ value: 0.09599999999999997
245
+ name: Cosine Precision@10
246
+ - type: cosine_recall@1
247
+ value: 0.53
248
+ name: Cosine Recall@1
249
+ - type: cosine_recall@3
250
+ value: 0.7766666666666666
251
+ name: Cosine Recall@3
252
+ - type: cosine_recall@5
253
+ value: 0.8566666666666666
254
+ name: Cosine Recall@5
255
+ - type: cosine_recall@10
256
+ value: 0.8866666666666667
257
+ name: Cosine Recall@10
258
+ - type: cosine_ndcg@10
259
+ value: 0.7348538316509182
260
+ name: Cosine Ndcg@10
261
+ - type: cosine_mrr@10
262
+ value: 0.6961904761904762
263
+ name: Cosine Mrr@10
264
+ - type: cosine_map@100
265
+ value: 0.6788071339639872
266
+ name: Cosine Map@100
267
+ - task:
268
+ type: information-retrieval
269
+ name: Information Retrieval
270
+ dataset:
271
+ name: NanoFiQA2018
272
+ type: NanoFiQA2018
273
+ metrics:
274
+ - type: cosine_accuracy@1
275
+ value: 0.24
276
+ name: Cosine Accuracy@1
277
+ - type: cosine_accuracy@3
278
+ value: 0.4
279
+ name: Cosine Accuracy@3
280
+ - type: cosine_accuracy@5
281
+ value: 0.5
282
+ name: Cosine Accuracy@5
283
+ - type: cosine_accuracy@10
284
+ value: 0.6
285
+ name: Cosine Accuracy@10
286
+ - type: cosine_precision@1
287
+ value: 0.24
288
+ name: Cosine Precision@1
289
+ - type: cosine_precision@3
290
+ value: 0.16
291
+ name: Cosine Precision@3
292
+ - type: cosine_precision@5
293
+ value: 0.14
294
+ name: Cosine Precision@5
295
+ - type: cosine_precision@10
296
+ value: 0.08800000000000001
297
+ name: Cosine Precision@10
298
+ - type: cosine_recall@1
299
+ value: 0.11474603174603175
300
+ name: Cosine Recall@1
301
+ - type: cosine_recall@3
302
+ value: 0.22874603174603172
303
+ name: Cosine Recall@3
304
+ - type: cosine_recall@5
305
+ value: 0.3166031746031746
306
+ name: Cosine Recall@5
307
+ - type: cosine_recall@10
308
+ value: 0.3986031746031745
309
+ name: Cosine Recall@10
310
+ - type: cosine_ndcg@10
311
+ value: 0.2925721974861802
312
+ name: Cosine Ndcg@10
313
+ - type: cosine_mrr@10
314
+ value: 0.3385
315
+ name: Cosine Mrr@10
316
+ - type: cosine_map@100
317
+ value: 0.2372091627126374
318
+ name: Cosine Map@100
319
+ - task:
320
+ type: information-retrieval
321
+ name: Information Retrieval
322
+ dataset:
323
+ name: NanoHotpotQA
324
+ type: NanoHotpotQA
325
+ metrics:
326
+ - type: cosine_accuracy@1
327
+ value: 0.6
328
+ name: Cosine Accuracy@1
329
+ - type: cosine_accuracy@3
330
+ value: 0.68
331
+ name: Cosine Accuracy@3
332
+ - type: cosine_accuracy@5
333
+ value: 0.74
334
+ name: Cosine Accuracy@5
335
+ - type: cosine_accuracy@10
336
+ value: 0.88
337
+ name: Cosine Accuracy@10
338
+ - type: cosine_precision@1
339
+ value: 0.6
340
+ name: Cosine Precision@1
341
+ - type: cosine_precision@3
342
+ value: 0.2866666666666667
343
+ name: Cosine Precision@3
344
+ - type: cosine_precision@5
345
+ value: 0.192
346
+ name: Cosine Precision@5
347
+ - type: cosine_precision@10
348
+ value: 0.118
349
+ name: Cosine Precision@10
350
+ - type: cosine_recall@1
351
+ value: 0.3
352
+ name: Cosine Recall@1
353
+ - type: cosine_recall@3
354
+ value: 0.43
355
+ name: Cosine Recall@3
356
+ - type: cosine_recall@5
357
+ value: 0.48
358
+ name: Cosine Recall@5
359
+ - type: cosine_recall@10
360
+ value: 0.59
361
+ name: Cosine Recall@10
362
+ - type: cosine_ndcg@10
363
+ value: 0.5291588954628265
364
+ name: Cosine Ndcg@10
365
+ - type: cosine_mrr@10
366
+ value: 0.6639365079365079
367
+ name: Cosine Mrr@10
368
+ - type: cosine_map@100
369
+ value: 0.45230644038161627
370
+ name: Cosine Map@100
371
+ - task:
372
+ type: information-retrieval
373
+ name: Information Retrieval
374
+ dataset:
375
+ name: NanoMSMARCO
376
+ type: NanoMSMARCO
377
+ metrics:
378
+ - type: cosine_accuracy@1
379
+ value: 0.28
380
+ name: Cosine Accuracy@1
381
+ - type: cosine_accuracy@3
382
+ value: 0.48
383
+ name: Cosine Accuracy@3
384
+ - type: cosine_accuracy@5
385
+ value: 0.54
386
+ name: Cosine Accuracy@5
387
+ - type: cosine_accuracy@10
388
+ value: 0.66
389
+ name: Cosine Accuracy@10
390
+ - type: cosine_precision@1
391
+ value: 0.28
392
+ name: Cosine Precision@1
393
+ - type: cosine_precision@3
394
+ value: 0.16
395
+ name: Cosine Precision@3
396
+ - type: cosine_precision@5
397
+ value: 0.10800000000000001
398
+ name: Cosine Precision@5
399
+ - type: cosine_precision@10
400
+ value: 0.066
401
+ name: Cosine Precision@10
402
+ - type: cosine_recall@1
403
+ value: 0.28
404
+ name: Cosine Recall@1
405
+ - type: cosine_recall@3
406
+ value: 0.48
407
+ name: Cosine Recall@3
408
+ - type: cosine_recall@5
409
+ value: 0.54
410
+ name: Cosine Recall@5
411
+ - type: cosine_recall@10
412
+ value: 0.66
413
+ name: Cosine Recall@10
414
+ - type: cosine_ndcg@10
415
+ value: 0.46795689507567784
416
+ name: Cosine Ndcg@10
417
+ - type: cosine_mrr@10
418
+ value: 0.4079126984126984
419
+ name: Cosine Mrr@10
420
+ - type: cosine_map@100
421
+ value: 0.42763462709531985
422
+ name: Cosine Map@100
423
+ - task:
424
+ type: information-retrieval
425
+ name: Information Retrieval
426
+ dataset:
427
+ name: NanoNFCorpus
428
+ type: NanoNFCorpus
429
+ metrics:
430
+ - type: cosine_accuracy@1
431
+ value: 0.32
432
+ name: Cosine Accuracy@1
433
+ - type: cosine_accuracy@3
434
+ value: 0.48
435
+ name: Cosine Accuracy@3
436
+ - type: cosine_accuracy@5
437
+ value: 0.5
438
+ name: Cosine Accuracy@5
439
+ - type: cosine_accuracy@10
440
+ value: 0.56
441
+ name: Cosine Accuracy@10
442
+ - type: cosine_precision@1
443
+ value: 0.32
444
+ name: Cosine Precision@1
445
+ - type: cosine_precision@3
446
+ value: 0.30666666666666664
447
+ name: Cosine Precision@3
448
+ - type: cosine_precision@5
449
+ value: 0.244
450
+ name: Cosine Precision@5
451
+ - type: cosine_precision@10
452
+ value: 0.184
453
+ name: Cosine Precision@10
454
+ - type: cosine_recall@1
455
+ value: 0.02092621665706462
456
+ name: Cosine Recall@1
457
+ - type: cosine_recall@3
458
+ value: 0.053426190783308986
459
+ name: Cosine Recall@3
460
+ - type: cosine_recall@5
461
+ value: 0.06393651269284006
462
+ name: Cosine Recall@5
463
+ - type: cosine_recall@10
464
+ value: 0.08045448545888809
465
+ name: Cosine Recall@10
466
+ - type: cosine_ndcg@10
467
+ value: 0.23067635403503162
468
+ name: Cosine Ndcg@10
469
+ - type: cosine_mrr@10
470
+ value: 0.39788888888888885
471
+ name: Cosine Mrr@10
472
+ - type: cosine_map@100
473
+ value: 0.09661097314535905
474
+ name: Cosine Map@100
475
+ - task:
476
+ type: information-retrieval
477
+ name: Information Retrieval
478
+ dataset:
479
+ name: NanoNQ
480
+ type: NanoNQ
481
+ metrics:
482
+ - type: cosine_accuracy@1
483
+ value: 0.38
484
+ name: Cosine Accuracy@1
485
+ - type: cosine_accuracy@3
486
+ value: 0.54
487
+ name: Cosine Accuracy@3
488
+ - type: cosine_accuracy@5
489
+ value: 0.62
490
+ name: Cosine Accuracy@5
491
+ - type: cosine_accuracy@10
492
+ value: 0.74
493
+ name: Cosine Accuracy@10
494
+ - type: cosine_precision@1
495
+ value: 0.38
496
+ name: Cosine Precision@1
497
+ - type: cosine_precision@3
498
+ value: 0.18
499
+ name: Cosine Precision@3
500
+ - type: cosine_precision@5
501
+ value: 0.128
502
+ name: Cosine Precision@5
503
+ - type: cosine_precision@10
504
+ value: 0.07600000000000001
505
+ name: Cosine Precision@10
506
+ - type: cosine_recall@1
507
+ value: 0.38
508
+ name: Cosine Recall@1
509
+ - type: cosine_recall@3
510
+ value: 0.51
511
+ name: Cosine Recall@3
512
+ - type: cosine_recall@5
513
+ value: 0.6
514
+ name: Cosine Recall@5
515
+ - type: cosine_recall@10
516
+ value: 0.71
517
+ name: Cosine Recall@10
518
+ - type: cosine_ndcg@10
519
+ value: 0.5386606354769653
520
+ name: Cosine Ndcg@10
521
+ - type: cosine_mrr@10
522
+ value: 0.490547619047619
523
+ name: Cosine Mrr@10
524
+ - type: cosine_map@100
525
+ value: 0.48961052316839493
526
+ name: Cosine Map@100
527
+ - task:
528
+ type: information-retrieval
529
+ name: Information Retrieval
530
+ dataset:
531
+ name: NanoQuoraRetrieval
532
+ type: NanoQuoraRetrieval
533
+ metrics:
534
+ - type: cosine_accuracy@1
535
+ value: 0.84
536
+ name: Cosine Accuracy@1
537
+ - type: cosine_accuracy@3
538
+ value: 0.94
539
+ name: Cosine Accuracy@3
540
+ - type: cosine_accuracy@5
541
+ value: 0.98
542
+ name: Cosine Accuracy@5
543
+ - type: cosine_accuracy@10
544
+ value: 1.0
545
+ name: Cosine Accuracy@10
546
+ - type: cosine_precision@1
547
+ value: 0.84
548
+ name: Cosine Precision@1
549
+ - type: cosine_precision@3
550
+ value: 0.38666666666666655
551
+ name: Cosine Precision@3
552
+ - type: cosine_precision@5
553
+ value: 0.24799999999999997
554
+ name: Cosine Precision@5
555
+ - type: cosine_precision@10
556
+ value: 0.12999999999999998
557
+ name: Cosine Precision@10
558
+ - type: cosine_recall@1
559
+ value: 0.7573333333333332
560
+ name: Cosine Recall@1
561
+ - type: cosine_recall@3
562
+ value: 0.912
563
+ name: Cosine Recall@3
564
+ - type: cosine_recall@5
565
+ value: 0.946
566
+ name: Cosine Recall@5
567
+ - type: cosine_recall@10
568
+ value: 0.9793333333333334
569
+ name: Cosine Recall@10
570
+ - type: cosine_ndcg@10
571
+ value: 0.9157663307482551
572
+ name: Cosine Ndcg@10
573
+ - type: cosine_mrr@10
574
+ value: 0.9009999999999999
575
+ name: Cosine Mrr@10
576
+ - type: cosine_map@100
577
+ value: 0.8893741502029173
578
+ name: Cosine Map@100
579
+ - task:
580
+ type: information-retrieval
581
+ name: Information Retrieval
582
+ dataset:
583
+ name: NanoSCIDOCS
584
+ type: NanoSCIDOCS
585
+ metrics:
586
+ - type: cosine_accuracy@1
587
+ value: 0.26
588
+ name: Cosine Accuracy@1
589
+ - type: cosine_accuracy@3
590
+ value: 0.46
591
+ name: Cosine Accuracy@3
592
+ - type: cosine_accuracy@5
593
+ value: 0.6
594
+ name: Cosine Accuracy@5
595
+ - type: cosine_accuracy@10
596
+ value: 0.68
597
+ name: Cosine Accuracy@10
598
+ - type: cosine_precision@1
599
+ value: 0.26
600
+ name: Cosine Precision@1
601
+ - type: cosine_precision@3
602
+ value: 0.20666666666666664
603
+ name: Cosine Precision@3
604
+ - type: cosine_precision@5
605
+ value: 0.184
606
+ name: Cosine Precision@5
607
+ - type: cosine_precision@10
608
+ value: 0.126
609
+ name: Cosine Precision@10
610
+ - type: cosine_recall@1
611
+ value: 0.054000000000000006
612
+ name: Cosine Recall@1
613
+ - type: cosine_recall@3
614
+ value: 0.12866666666666668
615
+ name: Cosine Recall@3
616
+ - type: cosine_recall@5
617
+ value: 0.18966666666666668
618
+ name: Cosine Recall@5
619
+ - type: cosine_recall@10
620
+ value: 0.25866666666666666
621
+ name: Cosine Recall@10
622
+ - type: cosine_ndcg@10
623
+ value: 0.24181947685643387
624
+ name: Cosine Ndcg@10
625
+ - type: cosine_mrr@10
626
+ value: 0.3803571428571429
627
+ name: Cosine Mrr@10
628
+ - type: cosine_map@100
629
+ value: 0.18652061021747493
630
+ name: Cosine Map@100
631
+ - task:
632
+ type: information-retrieval
633
+ name: Information Retrieval
634
+ dataset:
635
+ name: NanoArguAna
636
+ type: NanoArguAna
637
+ metrics:
638
+ - type: cosine_accuracy@1
639
+ value: 0.16
640
+ name: Cosine Accuracy@1
641
+ - type: cosine_accuracy@3
642
+ value: 0.58
643
+ name: Cosine Accuracy@3
644
+ - type: cosine_accuracy@5
645
+ value: 0.74
646
+ name: Cosine Accuracy@5
647
+ - type: cosine_accuracy@10
648
+ value: 0.84
649
+ name: Cosine Accuracy@10
650
+ - type: cosine_precision@1
651
+ value: 0.16
652
+ name: Cosine Precision@1
653
+ - type: cosine_precision@3
654
+ value: 0.19333333333333336
655
+ name: Cosine Precision@3
656
+ - type: cosine_precision@5
657
+ value: 0.14800000000000002
658
+ name: Cosine Precision@5
659
+ - type: cosine_precision@10
660
+ value: 0.08399999999999999
661
+ name: Cosine Precision@10
662
+ - type: cosine_recall@1
663
+ value: 0.16
664
+ name: Cosine Recall@1
665
+ - type: cosine_recall@3
666
+ value: 0.58
667
+ name: Cosine Recall@3
668
+ - type: cosine_recall@5
669
+ value: 0.74
670
+ name: Cosine Recall@5
671
+ - type: cosine_recall@10
672
+ value: 0.84
673
+ name: Cosine Recall@10
674
+ - type: cosine_ndcg@10
675
+ value: 0.5045313323048141
676
+ name: Cosine Ndcg@10
677
+ - type: cosine_mrr@10
678
+ value: 0.3963333333333333
679
+ name: Cosine Mrr@10
680
+ - type: cosine_map@100
681
+ value: 0.40074428294573644
682
+ name: Cosine Map@100
683
+ - task:
684
+ type: information-retrieval
685
+ name: Information Retrieval
686
+ dataset:
687
+ name: NanoSciFact
688
+ type: NanoSciFact
689
+ metrics:
690
+ - type: cosine_accuracy@1
691
+ value: 0.42
692
+ name: Cosine Accuracy@1
693
+ - type: cosine_accuracy@3
694
+ value: 0.58
695
+ name: Cosine Accuracy@3
696
+ - type: cosine_accuracy@5
697
+ value: 0.62
698
+ name: Cosine Accuracy@5
699
+ - type: cosine_accuracy@10
700
+ value: 0.64
701
+ name: Cosine Accuracy@10
702
+ - type: cosine_precision@1
703
+ value: 0.42
704
+ name: Cosine Precision@1
705
+ - type: cosine_precision@3
706
+ value: 0.20666666666666667
707
+ name: Cosine Precision@3
708
+ - type: cosine_precision@5
709
+ value: 0.14
710
+ name: Cosine Precision@5
711
+ - type: cosine_precision@10
712
+ value: 0.07600000000000001
713
+ name: Cosine Precision@10
714
+ - type: cosine_recall@1
715
+ value: 0.4
716
+ name: Cosine Recall@1
717
+ - type: cosine_recall@3
718
+ value: 0.56
719
+ name: Cosine Recall@3
720
+ - type: cosine_recall@5
721
+ value: 0.605
722
+ name: Cosine Recall@5
723
+ - type: cosine_recall@10
724
+ value: 0.64
725
+ name: Cosine Recall@10
726
+ - type: cosine_ndcg@10
727
+ value: 0.5380316349319392
728
+ name: Cosine Ndcg@10
729
+ - type: cosine_mrr@10
730
+ value: 0.5056666666666666
731
+ name: Cosine Mrr@10
732
+ - type: cosine_map@100
733
+ value: 0.5079821472790408
734
+ name: Cosine Map@100
735
+ - task:
736
+ type: information-retrieval
737
+ name: Information Retrieval
738
+ dataset:
739
+ name: NanoTouche2020
740
+ type: NanoTouche2020
741
+ metrics:
742
+ - type: cosine_accuracy@1
743
+ value: 0.4489795918367347
744
+ name: Cosine Accuracy@1
745
+ - type: cosine_accuracy@3
746
+ value: 0.8979591836734694
747
+ name: Cosine Accuracy@3
748
+ - type: cosine_accuracy@5
749
+ value: 0.9183673469387755
750
+ name: Cosine Accuracy@5
751
+ - type: cosine_accuracy@10
752
+ value: 0.9795918367346939
753
+ name: Cosine Accuracy@10
754
+ - type: cosine_precision@1
755
+ value: 0.4489795918367347
756
+ name: Cosine Precision@1
757
+ - type: cosine_precision@3
758
+ value: 0.4965986394557823
759
+ name: Cosine Precision@3
760
+ - type: cosine_precision@5
761
+ value: 0.45714285714285713
762
+ name: Cosine Precision@5
763
+ - type: cosine_precision@10
764
+ value: 0.38979591836734706
765
+ name: Cosine Precision@10
766
+ - type: cosine_recall@1
767
+ value: 0.03475887574057735
768
+ name: Cosine Recall@1
769
+ - type: cosine_recall@3
770
+ value: 0.11109807516506923
771
+ name: Cosine Recall@3
772
+ - type: cosine_recall@5
773
+ value: 0.1656210426064535
774
+ name: Cosine Recall@5
775
+ - type: cosine_recall@10
776
+ value: 0.2684807614936963
777
+ name: Cosine Recall@10
778
+ - type: cosine_ndcg@10
779
+ value: 0.43233093716838594
780
+ name: Cosine Ndcg@10
781
+ - type: cosine_mrr@10
782
+ value: 0.6532555879494653
783
+ name: Cosine Mrr@10
784
+ - type: cosine_map@100
785
+ value: 0.33493945959592186
786
+ name: Cosine Map@100
787
+ - task:
788
+ type: nano-beir
789
+ name: Nano BEIR
790
+ dataset:
791
+ name: NanoBEIR mean
792
+ type: NanoBEIR_mean
793
+ metrics:
794
+ - type: cosine_accuracy@1
795
+ value: 0.4053061224489796
796
+ name: Cosine Accuracy@1
797
+ - type: cosine_accuracy@3
798
+ value: 0.6213814756671899
799
+ name: Cosine Accuracy@3
800
+ - type: cosine_accuracy@5
801
+ value: 0.6891051805337519
802
+ name: Cosine Accuracy@5
803
+ - type: cosine_accuracy@10
804
+ value: 0.7676609105180533
805
+ name: Cosine Accuracy@10
806
+ - type: cosine_precision@1
807
+ value: 0.4053061224489796
808
+ name: Cosine Precision@1
809
+ - type: cosine_precision@3
810
+ value: 0.2694819466248038
811
+ name: Cosine Precision@3
812
+ - type: cosine_precision@5
813
+ value: 0.20962637362637365
814
+ name: Cosine Precision@5
815
+ - type: cosine_precision@10
816
+ value: 0.14567660910518054
817
+ name: Cosine Precision@10
818
+ - type: cosine_recall@1
819
+ value: 0.24693944527453943
820
+ name: Cosine Recall@1
821
+ - type: cosine_recall@3
822
+ value: 0.3915472856287718
823
+ name: Cosine Recall@3
824
+ - type: cosine_recall@5
825
+ value: 0.4541123273847895
826
+ name: Cosine Recall@5
827
+ - type: cosine_recall@10
828
+ value: 0.5280272058403178
829
+ name: Cosine Recall@10
830
+ - type: cosine_ndcg@10
831
+ value: 0.4727642081975526
832
+ name: Cosine Ndcg@10
833
+ - type: cosine_mrr@10
834
+ value: 0.5268627619545987
835
+ name: Cosine Mrr@10
836
+ - type: cosine_map@100
837
+ value: 0.40266180779497585
838
+ name: Cosine Map@100
839
+ ---
840
+
841
+ # bert-base-uncased adapter finetuned on GooAQ pairs
842
+
843
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [google-bert/bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased) on the [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
844
+
845
+ ## Model Details
846
+
847
+ ### Model Description
848
+ - **Model Type:** Sentence Transformer
849
+ - **Base model:** [google-bert/bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased) <!-- at revision 86b5e0934494bd15c9632b12f734a8a67f723594 -->
850
+ - **Maximum Sequence Length:** 512 tokens
851
+ - **Output Dimensionality:** 768 dimensions
852
+ - **Similarity Function:** Cosine Similarity
853
+ - **Training Dataset:**
854
+ - [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq)
855
+ - **Language:** en
856
+ - **License:** apache-2.0
857
+
858
+ ### Model Sources
859
+
860
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
861
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
862
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
863
+
864
+ ### Full Model Architecture
865
+
866
+ ```
867
+ SentenceTransformer(
868
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
869
+ (1): Pooling({'word_embedding_dimension': 768, '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})
870
+ )
871
+ ```
872
+
873
+ ## Usage
874
+
875
+ ### Direct Usage (Sentence Transformers)
876
+
877
+ First install the Sentence Transformers library:
878
+
879
+ ```bash
880
+ pip install -U sentence-transformers
881
+ ```
882
+
883
+ Then you can load this model and run inference.
884
+ ```python
885
+ from sentence_transformers import SentenceTransformer
886
+
887
+ # Download from the 🤗 Hub
888
+ model = SentenceTransformer("tomaarsen/bert-base-uncased-gooaq")
889
+ # Run inference
890
+ sentences = [
891
+ 'how can i download youtube videos with internet download manager?',
892
+ "['Go to settings and then click on extensions (top left side in chrome).', 'Minimise your browser and open the location (folder) where IDM is installed. ... ', 'Find the file “IDMGCExt. ... ', 'Drag this file to your chrome browser and drop to install the IDM extension.']",
893
+ "Coca-Cola might rot your teeth and load your body with sugar and calories, but it's actually an effective and safe first line of treatment for some stomach blockages, researchers say.",
894
+ ]
895
+ embeddings = model.encode(sentences)
896
+ print(embeddings.shape)
897
+ # [3, 768]
898
+
899
+ # Get the similarity scores for the embeddings
900
+ similarities = model.similarity(embeddings, embeddings)
901
+ print(similarities.shape)
902
+ # [3, 3]
903
+ ```
904
+
905
+ <!--
906
+ ### Direct Usage (Transformers)
907
+
908
+ <details><summary>Click to see the direct usage in Transformers</summary>
909
+
910
+ </details>
911
+ -->
912
+
913
+ <!--
914
+ ### Downstream Usage (Sentence Transformers)
915
+
916
+ You can finetune this model on your own dataset.
917
+
918
+ <details><summary>Click to expand</summary>
919
+
920
+ </details>
921
+ -->
922
+
923
+ <!--
924
+ ### Out-of-Scope Use
925
+
926
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
927
+ -->
928
+
929
+ ## Evaluation
930
+
931
+ ### Metrics
932
+
933
+ #### Information Retrieval
934
+
935
+ * Datasets: `NanoClimateFEVER`, `NanoDBPedia`, `NanoFEVER`, `NanoFiQA2018`, `NanoHotpotQA`, `NanoMSMARCO`, `NanoNFCorpus`, `NanoNQ`, `NanoQuoraRetrieval`, `NanoSCIDOCS`, `NanoArguAna`, `NanoSciFact` and `NanoTouche2020`
936
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
937
+
938
+ | Metric | NanoClimateFEVER | NanoDBPedia | NanoFEVER | NanoFiQA2018 | NanoHotpotQA | NanoMSMARCO | NanoNFCorpus | NanoNQ | NanoQuoraRetrieval | NanoSCIDOCS | NanoArguAna | NanoSciFact | NanoTouche2020 |
939
+ |:--------------------|:-----------------|:------------|:-----------|:-------------|:-------------|:------------|:-------------|:-----------|:-------------------|:------------|:------------|:------------|:---------------|
940
+ | cosine_accuracy@1 | 0.24 | 0.54 | 0.54 | 0.24 | 0.6 | 0.28 | 0.32 | 0.38 | 0.84 | 0.26 | 0.16 | 0.42 | 0.449 |
941
+ | cosine_accuracy@3 | 0.42 | 0.8 | 0.82 | 0.4 | 0.68 | 0.48 | 0.48 | 0.54 | 0.94 | 0.46 | 0.58 | 0.58 | 0.898 |
942
+ | cosine_accuracy@5 | 0.46 | 0.84 | 0.9 | 0.5 | 0.74 | 0.54 | 0.5 | 0.62 | 0.98 | 0.6 | 0.74 | 0.62 | 0.9184 |
943
+ | cosine_accuracy@10 | 0.56 | 0.92 | 0.92 | 0.6 | 0.88 | 0.66 | 0.56 | 0.74 | 1.0 | 0.68 | 0.84 | 0.64 | 0.9796 |
944
+ | cosine_precision@1 | 0.24 | 0.54 | 0.54 | 0.24 | 0.6 | 0.28 | 0.32 | 0.38 | 0.84 | 0.26 | 0.16 | 0.42 | 0.449 |
945
+ | cosine_precision@3 | 0.16 | 0.4867 | 0.2733 | 0.16 | 0.2867 | 0.16 | 0.3067 | 0.18 | 0.3867 | 0.2067 | 0.1933 | 0.2067 | 0.4966 |
946
+ | cosine_precision@5 | 0.108 | 0.444 | 0.184 | 0.14 | 0.192 | 0.108 | 0.244 | 0.128 | 0.248 | 0.184 | 0.148 | 0.14 | 0.4571 |
947
+ | cosine_precision@10 | 0.07 | 0.39 | 0.096 | 0.088 | 0.118 | 0.066 | 0.184 | 0.076 | 0.13 | 0.126 | 0.084 | 0.076 | 0.3898 |
948
+ | cosine_recall@1 | 0.1317 | 0.0468 | 0.53 | 0.1147 | 0.3 | 0.28 | 0.0209 | 0.38 | 0.7573 | 0.054 | 0.16 | 0.4 | 0.0348 |
949
+ | cosine_recall@3 | 0.2083 | 0.1112 | 0.7767 | 0.2287 | 0.43 | 0.48 | 0.0534 | 0.51 | 0.912 | 0.1287 | 0.58 | 0.56 | 0.1111 |
950
+ | cosine_recall@5 | 0.2417 | 0.1583 | 0.8567 | 0.3166 | 0.48 | 0.54 | 0.0639 | 0.6 | 0.946 | 0.1897 | 0.74 | 0.605 | 0.1656 |
951
+ | cosine_recall@10 | 0.2967 | 0.2555 | 0.8867 | 0.3986 | 0.59 | 0.66 | 0.0805 | 0.71 | 0.9793 | 0.2587 | 0.84 | 0.64 | 0.2685 |
952
+ | **cosine_ndcg@10** | **0.2552** | **0.4644** | **0.7349** | **0.2926** | **0.5292** | **0.468** | **0.2307** | **0.5387** | **0.9158** | **0.2418** | **0.5045** | **0.538** | **0.4323** |
953
+ | cosine_mrr@10 | 0.3379 | 0.6797 | 0.6962 | 0.3385 | 0.6639 | 0.4079 | 0.3979 | 0.4905 | 0.901 | 0.3804 | 0.3963 | 0.5057 | 0.6533 |
954
+ | cosine_map@100 | 0.2076 | 0.3253 | 0.6788 | 0.2372 | 0.4523 | 0.4276 | 0.0966 | 0.4896 | 0.8894 | 0.1865 | 0.4007 | 0.508 | 0.3349 |
955
+
956
+ #### Nano BEIR
957
+
958
+ * Dataset: `NanoBEIR_mean`
959
+ * Evaluated with [<code>NanoBEIREvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.NanoBEIREvaluator)
960
+
961
+ | Metric | Value |
962
+ |:--------------------|:-----------|
963
+ | cosine_accuracy@1 | 0.4053 |
964
+ | cosine_accuracy@3 | 0.6214 |
965
+ | cosine_accuracy@5 | 0.6891 |
966
+ | cosine_accuracy@10 | 0.7677 |
967
+ | cosine_precision@1 | 0.4053 |
968
+ | cosine_precision@3 | 0.2695 |
969
+ | cosine_precision@5 | 0.2096 |
970
+ | cosine_precision@10 | 0.1457 |
971
+ | cosine_recall@1 | 0.2469 |
972
+ | cosine_recall@3 | 0.3915 |
973
+ | cosine_recall@5 | 0.4541 |
974
+ | cosine_recall@10 | 0.528 |
975
+ | **cosine_ndcg@10** | **0.4728** |
976
+ | cosine_mrr@10 | 0.5269 |
977
+ | cosine_map@100 | 0.4027 |
978
+
979
+ <!--
980
+ ## Bias, Risks and Limitations
981
+
982
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
983
+ -->
984
+
985
+ <!--
986
+ ### Recommendations
987
+
988
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
989
+ -->
990
+
991
+ ## Training Details
992
+
993
+ ### Training Dataset
994
+
995
+ #### gooaq
996
+
997
+ * Dataset: [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) at [b089f72](https://huggingface.co/datasets/sentence-transformers/gooaq/tree/b089f728748a068b7bc5234e5bcf5b25e3c8279c)
998
+ * Size: 3,012,496 training samples
999
+ * Columns: <code>question</code> and <code>answer</code>
1000
+ * Approximate statistics based on the first 1000 samples:
1001
+ | | question | answer |
1002
+ |:--------|:----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
1003
+ | type | string | string |
1004
+ | details | <ul><li>min: 8 tokens</li><li>mean: 11.86 tokens</li><li>max: 21 tokens</li></ul> | <ul><li>min: 14 tokens</li><li>mean: 60.48 tokens</li><li>max: 138 tokens</li></ul> |
1005
+ * Samples:
1006
+ | question | answer |
1007
+ |:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
1008
+ | <code>what is the difference between broilers and layers?</code> | <code>An egg laying poultry is called egger or layer whereas broilers are reared for obtaining meat. So a layer should be able to produce more number of large sized eggs, without growing too much. On the other hand, a broiler should yield more meat and hence should be able to grow well.</code> |
1009
+ | <code>what is the difference between chronological order and spatial order?</code> | <code>As a writer, you should always remember that unlike chronological order and the other organizational methods for data, spatial order does not take into account the time. Spatial order is primarily focused on the location. All it does is take into account the location of objects and not the time.</code> |
1010
+ | <code>is kamagra same as viagra?</code> | <code>Kamagra is thought to contain the same active ingredient as Viagra, sildenafil citrate. In theory, it should work in much the same way as Viagra, taking about 45 minutes to take effect, and lasting for around 4-6 hours. However, this will vary from person to person.</code> |
1011
+ * Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
1012
+ ```json
1013
+ {
1014
+ "loss": "CachedMultipleNegativesRankingLoss",
1015
+ "matryoshka_dims": [
1016
+ 768,
1017
+ 512,
1018
+ 256,
1019
+ 128,
1020
+ 64,
1021
+ 32
1022
+ ],
1023
+ "matryoshka_weights": [
1024
+ 1,
1025
+ 1,
1026
+ 1,
1027
+ 1,
1028
+ 1,
1029
+ 1
1030
+ ],
1031
+ "n_dims_per_step": -1
1032
+ }
1033
+ ```
1034
+
1035
+ ### Evaluation Dataset
1036
+
1037
+ #### gooaq
1038
+
1039
+ * Dataset: [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) at [b089f72](https://huggingface.co/datasets/sentence-transformers/gooaq/tree/b089f728748a068b7bc5234e5bcf5b25e3c8279c)
1040
+ * Size: 3,012,496 evaluation samples
1041
+ * Columns: <code>question</code> and <code>answer</code>
1042
+ * Approximate statistics based on the first 1000 samples:
1043
+ | | question | answer |
1044
+ |:--------|:----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
1045
+ | type | string | string |
1046
+ | details | <ul><li>min: 8 tokens</li><li>mean: 11.88 tokens</li><li>max: 22 tokens</li></ul> | <ul><li>min: 14 tokens</li><li>mean: 61.03 tokens</li><li>max: 127 tokens</li></ul> |
1047
+ * Samples:
1048
+ | question | answer |
1049
+ |:-----------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
1050
+ | <code>how do i program my directv remote with my tv?</code> | <code>['Press MENU on your remote.', 'Select Settings & Help > Settings > Remote Control > Program Remote.', 'Choose the device (TV, audio, DVD) you wish to program. ... ', 'Follow the on-screen prompts to complete programming.']</code> |
1051
+ | <code>are rodrigues fruit bats nocturnal?</code> | <code>Before its numbers were threatened by habitat destruction, storms, and hunting, some of those groups could number 500 or more members. Sunrise, sunset. Rodrigues fruit bats are most active at dawn, at dusk, and at night.</code> |
1052
+ | <code>why does your heart rate increase during exercise bbc bitesize?</code> | <code>During exercise there is an increase in physical activity and muscle cells respire more than they do when the body is at rest. The heart rate increases during exercise. The rate and depth of breathing increases - this makes sure that more oxygen is absorbed into the blood, and more carbon dioxide is removed from it.</code> |
1053
+ * Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
1054
+ ```json
1055
+ {
1056
+ "loss": "CachedMultipleNegativesRankingLoss",
1057
+ "matryoshka_dims": [
1058
+ 768,
1059
+ 512,
1060
+ 256,
1061
+ 128,
1062
+ 64,
1063
+ 32
1064
+ ],
1065
+ "matryoshka_weights": [
1066
+ 1,
1067
+ 1,
1068
+ 1,
1069
+ 1,
1070
+ 1,
1071
+ 1
1072
+ ],
1073
+ "n_dims_per_step": -1
1074
+ }
1075
+ ```
1076
+
1077
+ ### Training Hyperparameters
1078
+ #### Non-Default Hyperparameters
1079
+
1080
+ - `eval_strategy`: steps
1081
+ - `per_device_train_batch_size`: 1024
1082
+ - `per_device_eval_batch_size`: 1024
1083
+ - `learning_rate`: 2e-05
1084
+ - `num_train_epochs`: 1
1085
+ - `warmup_ratio`: 0.1
1086
+ - `seed`: 12
1087
+ - `bf16`: True
1088
+ - `batch_sampler`: no_duplicates
1089
+
1090
+ #### All Hyperparameters
1091
+ <details><summary>Click to expand</summary>
1092
+
1093
+ - `overwrite_output_dir`: False
1094
+ - `do_predict`: False
1095
+ - `eval_strategy`: steps
1096
+ - `prediction_loss_only`: True
1097
+ - `per_device_train_batch_size`: 1024
1098
+ - `per_device_eval_batch_size`: 1024
1099
+ - `per_gpu_train_batch_size`: None
1100
+ - `per_gpu_eval_batch_size`: None
1101
+ - `gradient_accumulation_steps`: 1
1102
+ - `eval_accumulation_steps`: None
1103
+ - `torch_empty_cache_steps`: None
1104
+ - `learning_rate`: 2e-05
1105
+ - `weight_decay`: 0.0
1106
+ - `adam_beta1`: 0.9
1107
+ - `adam_beta2`: 0.999
1108
+ - `adam_epsilon`: 1e-08
1109
+ - `max_grad_norm`: 1.0
1110
+ - `num_train_epochs`: 1
1111
+ - `max_steps`: -1
1112
+ - `lr_scheduler_type`: linear
1113
+ - `lr_scheduler_kwargs`: {}
1114
+ - `warmup_ratio`: 0.1
1115
+ - `warmup_steps`: 0
1116
+ - `log_level`: passive
1117
+ - `log_level_replica`: warning
1118
+ - `log_on_each_node`: True
1119
+ - `logging_nan_inf_filter`: True
1120
+ - `save_safetensors`: True
1121
+ - `save_on_each_node`: False
1122
+ - `save_only_model`: False
1123
+ - `restore_callback_states_from_checkpoint`: False
1124
+ - `no_cuda`: False
1125
+ - `use_cpu`: False
1126
+ - `use_mps_device`: False
1127
+ - `seed`: 12
1128
+ - `data_seed`: None
1129
+ - `jit_mode_eval`: False
1130
+ - `use_ipex`: False
1131
+ - `bf16`: True
1132
+ - `fp16`: False
1133
+ - `fp16_opt_level`: O1
1134
+ - `half_precision_backend`: auto
1135
+ - `bf16_full_eval`: False
1136
+ - `fp16_full_eval`: False
1137
+ - `tf32`: None
1138
+ - `local_rank`: 0
1139
+ - `ddp_backend`: None
1140
+ - `tpu_num_cores`: None
1141
+ - `tpu_metrics_debug`: False
1142
+ - `debug`: []
1143
+ - `dataloader_drop_last`: False
1144
+ - `dataloader_num_workers`: 0
1145
+ - `dataloader_prefetch_factor`: None
1146
+ - `past_index`: -1
1147
+ - `disable_tqdm`: False
1148
+ - `remove_unused_columns`: True
1149
+ - `label_names`: None
1150
+ - `load_best_model_at_end`: False
1151
+ - `ignore_data_skip`: False
1152
+ - `fsdp`: []
1153
+ - `fsdp_min_num_params`: 0
1154
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
1155
+ - `fsdp_transformer_layer_cls_to_wrap`: None
1156
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
1157
+ - `deepspeed`: None
1158
+ - `label_smoothing_factor`: 0.0
1159
+ - `optim`: adamw_torch
1160
+ - `optim_args`: None
1161
+ - `adafactor`: False
1162
+ - `group_by_length`: False
1163
+ - `length_column_name`: length
1164
+ - `ddp_find_unused_parameters`: None
1165
+ - `ddp_bucket_cap_mb`: None
1166
+ - `ddp_broadcast_buffers`: False
1167
+ - `dataloader_pin_memory`: True
1168
+ - `dataloader_persistent_workers`: False
1169
+ - `skip_memory_metrics`: True
1170
+ - `use_legacy_prediction_loop`: False
1171
+ - `push_to_hub`: False
1172
+ - `resume_from_checkpoint`: None
1173
+ - `hub_model_id`: None
1174
+ - `hub_strategy`: every_save
1175
+ - `hub_private_repo`: False
1176
+ - `hub_always_push`: False
1177
+ - `gradient_checkpointing`: False
1178
+ - `gradient_checkpointing_kwargs`: None
1179
+ - `include_inputs_for_metrics`: False
1180
+ - `include_for_metrics`: []
1181
+ - `eval_do_concat_batches`: True
1182
+ - `fp16_backend`: auto
1183
+ - `push_to_hub_model_id`: None
1184
+ - `push_to_hub_organization`: None
1185
+ - `mp_parameters`:
1186
+ - `auto_find_batch_size`: False
1187
+ - `full_determinism`: False
1188
+ - `torchdynamo`: None
1189
+ - `ray_scope`: last
1190
+ - `ddp_timeout`: 1800
1191
+ - `torch_compile`: False
1192
+ - `torch_compile_backend`: None
1193
+ - `torch_compile_mode`: None
1194
+ - `dispatch_batches`: None
1195
+ - `split_batches`: None
1196
+ - `include_tokens_per_second`: False
1197
+ - `include_num_input_tokens_seen`: False
1198
+ - `neftune_noise_alpha`: None
1199
+ - `optim_target_modules`: None
1200
+ - `batch_eval_metrics`: False
1201
+ - `eval_on_start`: False
1202
+ - `use_liger_kernel`: False
1203
+ - `eval_use_gather_object`: False
1204
+ - `average_tokens_across_devices`: False
1205
+ - `prompts`: None
1206
+ - `batch_sampler`: no_duplicates
1207
+ - `multi_dataset_batch_sampler`: proportional
1208
+
1209
+ </details>
1210
+
1211
+ ### Training Logs
1212
+ | Epoch | Step | Training Loss | Validation Loss | NanoClimateFEVER_cosine_ndcg@10 | NanoDBPedia_cosine_ndcg@10 | NanoFEVER_cosine_ndcg@10 | NanoFiQA2018_cosine_ndcg@10 | NanoHotpotQA_cosine_ndcg@10 | NanoMSMARCO_cosine_ndcg@10 | NanoNFCorpus_cosine_ndcg@10 | NanoNQ_cosine_ndcg@10 | NanoQuoraRetrieval_cosine_ndcg@10 | NanoSCIDOCS_cosine_ndcg@10 | NanoArguAna_cosine_ndcg@10 | NanoSciFact_cosine_ndcg@10 | NanoTouche2020_cosine_ndcg@10 | NanoBEIR_mean_cosine_ndcg@10 |
1213
+ |:------:|:----:|:-------------:|:---------------:|:-------------------------------:|:--------------------------:|:------------------------:|:---------------------------:|:---------------------------:|:--------------------------:|:---------------------------:|:---------------------:|:---------------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:-----------------------------:|:----------------------------:|
1214
+ | 0 | 0 | - | - | 0.1046 | 0.2182 | 0.1573 | 0.0575 | 0.2597 | 0.1602 | 0.0521 | 0.0493 | 0.7310 | 0.1320 | 0.2309 | 0.1240 | 0.0970 | 0.1826 |
1215
+ | 0.0010 | 1 | 28.4268 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1216
+ | 0.0256 | 25 | 24.7252 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1217
+ | 0.0512 | 50 | 13.3628 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1218
+ | 0.0768 | 75 | 7.843 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1219
+ | 0.1024 | 100 | 5.7393 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1220
+ | 0.1279 | 125 | 4.6576 | 2.3368 | 0.2890 | 0.4610 | 0.7408 | 0.2882 | 0.5446 | 0.4091 | 0.2179 | 0.4664 | 0.9079 | 0.2394 | 0.5433 | 0.5003 | 0.4318 | 0.4646 |
1221
+ | 0.1535 | 150 | 4.0846 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1222
+ | 0.1791 | 175 | 3.7129 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1223
+ | 0.2047 | 200 | 3.4899 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1224
+ | 0.2303 | 225 | 3.3263 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1225
+ | 0.2559 | 250 | 3.2013 | 1.6545 | 0.2622 | 0.4744 | 0.7456 | 0.2934 | 0.5371 | 0.4326 | 0.2290 | 0.5157 | 0.9130 | 0.2577 | 0.5189 | 0.5155 | 0.4302 | 0.4712 |
1226
+ | 0.2815 | 275 | 2.9109 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1227
+ | 0.3071 | 300 | 2.9064 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1228
+ | 0.3327 | 325 | 2.8215 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1229
+ | 0.3582 | 350 | 2.7893 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1230
+ | 0.3838 | 375 | 2.6663 | 1.4146 | 0.2629 | 0.4657 | 0.7330 | 0.2853 | 0.5299 | 0.4346 | 0.2311 | 0.5216 | 0.9172 | 0.2513 | 0.5133 | 0.5429 | 0.4287 | 0.4706 |
1231
+ | 0.4094 | 400 | 2.6672 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1232
+ | 0.4350 | 425 | 2.5587 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1233
+ | 0.4606 | 450 | 2.5001 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1234
+ | 0.4862 | 475 | 2.4476 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1235
+ | 0.5118 | 500 | 2.4127 | 1.2843 | 0.2565 | 0.4668 | 0.7289 | 0.2838 | 0.5392 | 0.4599 | 0.2284 | 0.5238 | 0.9021 | 0.2416 | 0.4971 | 0.5349 | 0.4320 | 0.4688 |
1236
+ | 0.5374 | 525 | 2.414 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1237
+ | 0.5629 | 550 | 2.3723 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1238
+ | 0.5885 | 575 | 2.3418 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1239
+ | 0.6141 | 600 | 2.2862 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1240
+ | 0.6397 | 625 | 2.207 | 1.2078 | 0.2613 | 0.4542 | 0.7382 | 0.2817 | 0.5230 | 0.4664 | 0.2282 | 0.5266 | 0.9095 | 0.2453 | 0.5127 | 0.5414 | 0.4239 | 0.4702 |
1241
+ | 0.6653 | 650 | 2.2305 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1242
+ | 0.6909 | 675 | 2.2409 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1243
+ | 0.7165 | 700 | 2.2001 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1244
+ | 0.7421 | 725 | 2.1923 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1245
+ | 0.7677 | 750 | 2.195 | 1.1538 | 0.2549 | 0.4671 | 0.7333 | 0.2804 | 0.5265 | 0.4659 | 0.2321 | 0.5331 | 0.9086 | 0.2429 | 0.5070 | 0.5430 | 0.4369 | 0.4717 |
1246
+ | 0.7932 | 775 | 2.1826 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1247
+ | 0.8188 | 800 | 2.1754 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1248
+ | 0.8444 | 825 | 2.1141 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1249
+ | 0.8700 | 850 | 2.1572 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1250
+ | 0.8956 | 875 | 2.1126 | 1.1256 | 0.2505 | 0.4622 | 0.7293 | 0.2857 | 0.5286 | 0.4823 | 0.2308 | 0.5397 | 0.9158 | 0.2412 | 0.5050 | 0.5365 | 0.4387 | 0.4728 |
1251
+ | 0.9212 | 900 | 2.0755 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1252
+ | 0.9468 | 925 | 2.1032 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1253
+ | 0.9724 | 950 | 2.1211 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1254
+ | 0.9980 | 975 | 2.0826 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1255
+ | 1.0 | 977 | - | - | 0.2552 | 0.4644 | 0.7349 | 0.2926 | 0.5292 | 0.4680 | 0.2307 | 0.5387 | 0.9158 | 0.2418 | 0.5045 | 0.5380 | 0.4323 | 0.4728 |
1256
+
1257
+
1258
+ ### Environmental Impact
1259
+ Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).
1260
+ - **Energy Consumed**: 0.624 kWh
1261
+ - **Carbon Emitted**: 0.243 kg of CO2
1262
+ - **Hours Used**: 1.619 hours
1263
+
1264
+ ### Training Hardware
1265
+ - **On Cloud**: No
1266
+ - **GPU Model**: 1 x NVIDIA GeForce RTX 3090
1267
+ - **CPU Model**: 13th Gen Intel(R) Core(TM) i7-13700K
1268
+ - **RAM Size**: 31.78 GB
1269
+
1270
+ ### Framework Versions
1271
+ - Python: 3.11.6
1272
+ - Sentence Transformers: 3.4.0.dev0
1273
+ - Transformers: 4.46.2
1274
+ - PyTorch: 2.5.0+cu121
1275
+ - Accelerate: 0.35.0.dev0
1276
+ - Datasets: 2.20.0
1277
+ - Tokenizers: 0.20.3
1278
+
1279
+ ## Citation
1280
+
1281
+ ### BibTeX
1282
+
1283
+ #### Sentence Transformers
1284
+ ```bibtex
1285
+ @inproceedings{reimers-2019-sentence-bert,
1286
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
1287
+ author = "Reimers, Nils and Gurevych, Iryna",
1288
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
1289
+ month = "11",
1290
+ year = "2019",
1291
+ publisher = "Association for Computational Linguistics",
1292
+ url = "https://arxiv.org/abs/1908.10084",
1293
+ }
1294
+ ```
1295
+
1296
+ #### MatryoshkaLoss
1297
+ ```bibtex
1298
+ @misc{kusupati2024matryoshka,
1299
+ title={Matryoshka Representation Learning},
1300
+ 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},
1301
+ year={2024},
1302
+ eprint={2205.13147},
1303
+ archivePrefix={arXiv},
1304
+ primaryClass={cs.LG}
1305
+ }
1306
+ ```
1307
+
1308
+ #### CachedMultipleNegativesRankingLoss
1309
+ ```bibtex
1310
+ @misc{gao2021scaling,
1311
+ title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
1312
+ author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan},
1313
+ year={2021},
1314
+ eprint={2101.06983},
1315
+ archivePrefix={arXiv},
1316
+ primaryClass={cs.LG}
1317
+ }
1318
+ ```
1319
+
1320
+ <!--
1321
+ ## Glossary
1322
+
1323
+ *Clearly define terms in order to be accessible across audiences.*
1324
+ -->
1325
+
1326
+ <!--
1327
+ ## Model Card Authors
1328
+
1329
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
1330
+ -->
1331
+
1332
+ <!--
1333
+ ## Model Card Contact
1334
+
1335
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
1336
+ -->
config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google-bert/bert-base-uncased",
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "pad_token_id": 0,
20
+ "position_embedding_type": "absolute",
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.46.2",
23
+ "type_vocab_size": 2,
24
+ "use_cache": true,
25
+ "vocab_size": 30522
26
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "3.4.0.dev0",
4
+ "transformers": "4.46.2",
5
+ "pytorch": "2.5.0+cu121"
6
+ },
7
+ "prompts": {},
8
+ "default_prompt_name": null,
9
+ "similarity_fn_name": "cosine"
10
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48631b3f7961fcdada247f87245236d6eeeddff36196878e476f4ea3e2e28504
3
+ size 437951328
modules.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": false
4
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "mask_token": "[MASK]",
48
+ "model_max_length": 512,
49
+ "pad_token": "[PAD]",
50
+ "sep_token": "[SEP]",
51
+ "strip_accents": null,
52
+ "tokenize_chinese_chars": true,
53
+ "tokenizer_class": "BertTokenizer",
54
+ "unk_token": "[UNK]"
55
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff