Hugging Face's logo

sartifyllc
/
bge-base-swahili-matryoshka

Sentence Similarity
sentence-transformers
Safetensors
English
bert
feature-extraction
Generated from Trainer
dataset_size:282883
loss:MatryoshkaLoss
loss:MultipleNegativesRankingLoss
Eval Results
text-embeddings-inference
Inference Endpoints
Model card Files Community
Mollel commited on
Commit
67d3e52
•
1 Parent(s): e512422

Add new SentenceTransformer model.

Browse files
Files changed (11) hide show
  1. 1_Pooling/config.json +10 -0
  2. README.md +1015 -0
  3. config.json +32 -0
  4. config_sentence_transformers.json +10 -0
  5. model.safetensors +3 -0
  6. modules.json +20 -0
  7. sentence_bert_config.json +4 -0
  8. special_tokens_map.json +37 -0
  9. tokenizer.json +0 -0
  10. tokenizer_config.json +57 -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": true,
4
+ "pooling_mode_mean_tokens": false,
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 +1,1015 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: apache-2.0
5
+ library_name: sentence-transformers
6
+ tags:
7
+ - sentence-transformers
8
+ - sentence-similarity
9
+ - feature-extraction
10
+ - generated_from_trainer
11
+ - dataset_size:282883
12
+ - loss:MatryoshkaLoss
13
+ - loss:MultipleNegativesRankingLoss
14
+ base_model: BAAI/bge-base-en-v1.5
15
+ datasets: []
16
+ metrics:
17
+ - cosine_accuracy@1
18
+ - cosine_accuracy@3
19
+ - cosine_accuracy@5
20
+ - cosine_accuracy@10
21
+ - cosine_precision@1
22
+ - cosine_precision@3
23
+ - cosine_precision@5
24
+ - cosine_precision@10
25
+ - cosine_recall@1
26
+ - cosine_recall@3
27
+ - cosine_recall@5
28
+ - cosine_recall@10
29
+ - cosine_ndcg@10
30
+ - cosine_mrr@10
31
+ - cosine_map@100
32
+ widget:
33
+ - source_sentence: Mwanamke anashona.
34
+ sentences:
35
+ - Mwanamke akishona blanketi pamoja.
36
+ - Lakini Sir James alimkatiza.
37
+ - Kwa kuongezea, wabunifu mashuhuri sasa wana maduka ya rejareja katika hoteli kadhaa
38
+ za ununuzi.
39
+ - source_sentence: Mtandao huwawezesha watu kununua vitu.
40
+ sentences:
41
+ - Mwanamke fulani anakata kipande cha jibini.
42
+ - Hakuna hata mmoja wa wafadhili hawa anayeweza kuruhusu kuacha sasa, haswa na uchumi
43
+ unaoteseka ndani na kitaifa.
44
+ - Kwa vyovyote vile, kama ningekuwa na nia ya kununua kitabu hicho, ningekuwa na
45
+ nafasi nzuri zaidi ya kujadili bei nzuri.
46
+ - source_sentence: Je, kweli wewe ni hivyo gullible, Dave Hanson?
47
+ sentences:
48
+ - Mwanamume amesimama katika mashua paddling kuelekea pwani lined na fanicha na
49
+ vitu vingine kubwa.
50
+ - Dave Hanson, je, unaamini kila kitu wanachosema?
51
+ - Wasichana watatu wakifanya mchezo wa kuigiza jukwaani.
52
+ - source_sentence: Wanandoa wakitembea pamoja.
53
+ sentences:
54
+ - Mwanamume aliyevalia koti la manjano anatembea kando ya gari la kubebea watu.
55
+ - Wenzi wa ndoa wazee wanatembea barabarani wakishikamana mikono.
56
+ - Msichana mdogo anapanda kwenye kifaa cha kamba.
57
+ - source_sentence: Kuna masuala ya sera.
58
+ sentences:
59
+ - Mwanamke mwenye makunyanzi sana akishikilia miwani yake na kutembea kwenye barabara
60
+ ya jiji.
61
+ - Mwanamume anayeigiza kwa ajili ya umati wa watu.
62
+ - Masuala ya sera ya mbinu nyingi na maombi.
63
+ pipeline_tag: sentence-similarity
64
+ model-index:
65
+ - name: BGE base Financial Matryoshka
66
+ results:
67
+ - task:
68
+ type: information-retrieval
69
+ name: Information Retrieval
70
+ dataset:
71
+ name: dim 768
72
+ type: dim_768
73
+ metrics:
74
+ - type: cosine_accuracy@1
75
+ value: 0.26803894120641386
76
+ name: Cosine Accuracy@1
77
+ - type: cosine_accuracy@3
78
+ value: 0.3499618223466531
79
+ name: Cosine Accuracy@3
80
+ - type: cosine_accuracy@5
81
+ value: 0.3858806312038687
82
+ name: Cosine Accuracy@5
83
+ - type: cosine_accuracy@10
84
+ value: 0.43318910664291166
85
+ name: Cosine Accuracy@10
86
+ - type: cosine_precision@1
87
+ value: 0.26803894120641386
88
+ name: Cosine Precision@1
89
+ - type: cosine_precision@3
90
+ value: 0.11665394078221768
91
+ name: Cosine Precision@3
92
+ - type: cosine_precision@5
93
+ value: 0.07717612624077373
94
+ name: Cosine Precision@5
95
+ - type: cosine_precision@10
96
+ value: 0.043318910664291166
97
+ name: Cosine Precision@10
98
+ - type: cosine_recall@1
99
+ value: 0.26803894120641386
100
+ name: Cosine Recall@1
101
+ - type: cosine_recall@3
102
+ value: 0.3499618223466531
103
+ name: Cosine Recall@3
104
+ - type: cosine_recall@5
105
+ value: 0.3858806312038687
106
+ name: Cosine Recall@5
107
+ - type: cosine_recall@10
108
+ value: 0.43318910664291166
109
+ name: Cosine Recall@10
110
+ - type: cosine_ndcg@10
111
+ value: 0.34611891078942064
112
+ name: Cosine Ndcg@10
113
+ - type: cosine_mrr@10
114
+ value: 0.3188061049905684
115
+ name: Cosine Mrr@10
116
+ - type: cosine_map@100
117
+ value: 0.3251959746415499
118
+ name: Cosine Map@100
119
+ - task:
120
+ type: information-retrieval
121
+ name: Information Retrieval
122
+ dataset:
123
+ name: dim 512
124
+ type: dim_512
125
+ metrics:
126
+ - type: cosine_accuracy@1
127
+ value: 0.26552557902774243
128
+ name: Cosine Accuracy@1
129
+ - type: cosine_accuracy@3
130
+ value: 0.34601679816747266
131
+ name: Cosine Accuracy@3
132
+ - type: cosine_accuracy@5
133
+ value: 0.3810766098243828
134
+ name: Cosine Accuracy@5
135
+ - type: cosine_accuracy@10
136
+ value: 0.4290850089081191
137
+ name: Cosine Accuracy@10
138
+ - type: cosine_precision@1
139
+ value: 0.26552557902774243
140
+ name: Cosine Precision@1
141
+ - type: cosine_precision@3
142
+ value: 0.11533893272249085
143
+ name: Cosine Precision@3
144
+ - type: cosine_precision@5
145
+ value: 0.07621532196487656
146
+ name: Cosine Precision@5
147
+ - type: cosine_precision@10
148
+ value: 0.04290850089081191
149
+ name: Cosine Precision@10
150
+ - type: cosine_recall@1
151
+ value: 0.26552557902774243
152
+ name: Cosine Recall@1
153
+ - type: cosine_recall@3
154
+ value: 0.34601679816747266
155
+ name: Cosine Recall@3
156
+ - type: cosine_recall@5
157
+ value: 0.3810766098243828
158
+ name: Cosine Recall@5
159
+ - type: cosine_recall@10
160
+ value: 0.4290850089081191
161
+ name: Cosine Recall@10
162
+ - type: cosine_ndcg@10
163
+ value: 0.3425120728009226
164
+ name: Cosine Ndcg@10
165
+ - type: cosine_mrr@10
166
+ value: 0.31538232445349546
167
+ name: Cosine Mrr@10
168
+ - type: cosine_map@100
169
+ value: 0.32174207802147353
170
+ name: Cosine Map@100
171
+ - task:
172
+ type: information-retrieval
173
+ name: Information Retrieval
174
+ dataset:
175
+ name: dim 256
176
+ type: dim_256
177
+ metrics:
178
+ - type: cosine_accuracy@1
179
+ value: 0.2576355306693815
180
+ name: Cosine Accuracy@1
181
+ - type: cosine_accuracy@3
182
+ value: 0.33790404683125475
183
+ name: Cosine Accuracy@3
184
+ - type: cosine_accuracy@5
185
+ value: 0.37165945533214556
186
+ name: Cosine Accuracy@5
187
+ - type: cosine_accuracy@10
188
+ value: 0.41950878086026977
189
+ name: Cosine Accuracy@10
190
+ - type: cosine_precision@1
191
+ value: 0.2576355306693815
192
+ name: Cosine Precision@1
193
+ - type: cosine_precision@3
194
+ value: 0.1126346822770849
195
+ name: Cosine Precision@3
196
+ - type: cosine_precision@5
197
+ value: 0.07433189106642912
198
+ name: Cosine Precision@5
199
+ - type: cosine_precision@10
200
+ value: 0.041950878086026974
201
+ name: Cosine Precision@10
202
+ - type: cosine_recall@1
203
+ value: 0.2576355306693815
204
+ name: Cosine Recall@1
205
+ - type: cosine_recall@3
206
+ value: 0.33790404683125475
207
+ name: Cosine Recall@3
208
+ - type: cosine_recall@5
209
+ value: 0.37165945533214556
210
+ name: Cosine Recall@5
211
+ - type: cosine_recall@10
212
+ value: 0.41950878086026977
213
+ name: Cosine Recall@10
214
+ - type: cosine_ndcg@10
215
+ value: 0.3338740008089949
216
+ name: Cosine Ndcg@10
217
+ - type: cosine_mrr@10
218
+ value: 0.30705069547968683
219
+ name: Cosine Mrr@10
220
+ - type: cosine_map@100
221
+ value: 0.3134101334652913
222
+ name: Cosine Map@100
223
+ - task:
224
+ type: information-retrieval
225
+ name: Information Retrieval
226
+ dataset:
227
+ name: dim 128
228
+ type: dim_128
229
+ metrics:
230
+ - type: cosine_accuracy@1
231
+ value: 0.24494146093153474
232
+ name: Cosine Accuracy@1
233
+ - type: cosine_accuracy@3
234
+ value: 0.3218694324255536
235
+ name: Cosine Accuracy@3
236
+ - type: cosine_accuracy@5
237
+ value: 0.3557202850598117
238
+ name: Cosine Accuracy@5
239
+ - type: cosine_accuracy@10
240
+ value: 0.402901501654365
241
+ name: Cosine Accuracy@10
242
+ - type: cosine_precision@1
243
+ value: 0.24494146093153474
244
+ name: Cosine Precision@1
245
+ - type: cosine_precision@3
246
+ value: 0.10728981080851785
247
+ name: Cosine Precision@3
248
+ - type: cosine_precision@5
249
+ value: 0.07114405701196233
250
+ name: Cosine Precision@5
251
+ - type: cosine_precision@10
252
+ value: 0.0402901501654365
253
+ name: Cosine Precision@10
254
+ - type: cosine_recall@1
255
+ value: 0.24494146093153474
256
+ name: Cosine Recall@1
257
+ - type: cosine_recall@3
258
+ value: 0.3218694324255536
259
+ name: Cosine Recall@3
260
+ - type: cosine_recall@5
261
+ value: 0.3557202850598117
262
+ name: Cosine Recall@5
263
+ - type: cosine_recall@10
264
+ value: 0.402901501654365
265
+ name: Cosine Recall@10
266
+ - type: cosine_ndcg@10
267
+ value: 0.3191027723891013
268
+ name: Cosine Ndcg@10
269
+ - type: cosine_mrr@10
270
+ value: 0.2928823673781056
271
+ name: Cosine Mrr@10
272
+ - type: cosine_map@100
273
+ value: 0.299205934269314
274
+ name: Cosine Map@100
275
+ - task:
276
+ type: information-retrieval
277
+ name: Information Retrieval
278
+ dataset:
279
+ name: dim 64
280
+ type: dim_64
281
+ metrics:
282
+ - type: cosine_accuracy@1
283
+ value: 0.21936243318910664
284
+ name: Cosine Accuracy@1
285
+ - type: cosine_accuracy@3
286
+ value: 0.2918045304148638
287
+ name: Cosine Accuracy@3
288
+ - type: cosine_accuracy@5
289
+ value: 0.3234601679816747
290
+ name: Cosine Accuracy@5
291
+ - type: cosine_accuracy@10
292
+ value: 0.3698778315092899
293
+ name: Cosine Accuracy@10
294
+ - type: cosine_precision@1
295
+ value: 0.21936243318910664
296
+ name: Cosine Precision@1
297
+ - type: cosine_precision@3
298
+ value: 0.0972681768049546
299
+ name: Cosine Precision@3
300
+ - type: cosine_precision@5
301
+ value: 0.06469203359633495
302
+ name: Cosine Precision@5
303
+ - type: cosine_precision@10
304
+ value: 0.03698778315092899
305
+ name: Cosine Precision@10
306
+ - type: cosine_recall@1
307
+ value: 0.21936243318910664
308
+ name: Cosine Recall@1
309
+ - type: cosine_recall@3
310
+ value: 0.2918045304148638
311
+ name: Cosine Recall@3
312
+ - type: cosine_recall@5
313
+ value: 0.3234601679816747
314
+ name: Cosine Recall@5
315
+ - type: cosine_recall@10
316
+ value: 0.3698778315092899
317
+ name: Cosine Recall@10
318
+ - type: cosine_ndcg@10
319
+ value: 0.2897472677253453
320
+ name: Cosine Ndcg@10
321
+ - type: cosine_mrr@10
322
+ value: 0.26472963050495346
323
+ name: Cosine Mrr@10
324
+ - type: cosine_map@100
325
+ value: 0.2710377326397304
326
+ name: Cosine Map@100
327
+ ---
328
+
329
+ # BGE base Financial Matryoshka
330
+
331
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5). 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.
332
+
333
+ ## Model Details
334
+
335
+ ### Model Description
336
+ - **Model Type:** Sentence Transformer
337
+ - **Base model:** [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5) <!-- at revision a5beb1e3e68b9ab74eb54cfd186867f64f240e1a -->
338
+ - **Maximum Sequence Length:** 512 tokens
339
+ - **Output Dimensionality:** 768 tokens
340
+ - **Similarity Function:** Cosine Similarity
341
+ <!-- - **Training Dataset:** Unknown -->
342
+ - **Language:** en
343
+ - **License:** apache-2.0
344
+
345
+ ### Model Sources
346
+
347
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
348
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
349
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
350
+
351
+ ### Full Model Architecture
352
+
353
+ ```
354
+ SentenceTransformer(
355
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': True}) with Transformer model: BertModel
356
+ (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
357
+ (2): Normalize()
358
+ )
359
+ ```
360
+
361
+ ## Usage
362
+
363
+ ### Direct Usage (Sentence Transformers)
364
+
365
+ First install the Sentence Transformers library:
366
+
367
+ ```bash
368
+ pip install -U sentence-transformers
369
+ ```
370
+
371
+ Then you can load this model and run inference.
372
+ ```python
373
+ from sentence_transformers import SentenceTransformer
374
+
375
+ # Download from the 🤗 Hub
376
+ model = SentenceTransformer("sartifyllc/bge-base-swahili-matryoshka")
377
+ # Run inference
378
+ sentences = [
379
+ 'Kuna masuala ya sera.',
380
+ 'Masuala ya sera ya mbinu nyingi na maombi.',
381
+ 'Mwanamke mwenye makunyanzi sana akishikilia miwani yake na kutembea kwenye barabara ya jiji.',
382
+ ]
383
+ embeddings = model.encode(sentences)
384
+ print(embeddings.shape)
385
+ # [3, 768]
386
+
387
+ # Get the similarity scores for the embeddings
388
+ similarities = model.similarity(embeddings, embeddings)
389
+ print(similarities.shape)
390
+ # [3, 3]
391
+ ```
392
+
393
+ <!--
394
+ ### Direct Usage (Transformers)
395
+
396
+ <details><summary>Click to see the direct usage in Transformers</summary>
397
+
398
+ </details>
399
+ -->
400
+
401
+ <!--
402
+ ### Downstream Usage (Sentence Transformers)
403
+
404
+ You can finetune this model on your own dataset.
405
+
406
+ <details><summary>Click to expand</summary>
407
+
408
+ </details>
409
+ -->
410
+
411
+ <!--
412
+ ### Out-of-Scope Use
413
+
414
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
415
+ -->
416
+
417
+ ## Evaluation
418
+
419
+ ### Metrics
420
+
421
+ #### Information Retrieval
422
+ * Dataset: `dim_768`
423
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
424
+
425
+ | Metric | Value |
426
+ |:--------------------|:-----------|
427
+ | cosine_accuracy@1 | 0.268 |
428
+ | cosine_accuracy@3 | 0.35 |
429
+ | cosine_accuracy@5 | 0.3859 |
430
+ | cosine_accuracy@10 | 0.4332 |
431
+ | cosine_precision@1 | 0.268 |
432
+ | cosine_precision@3 | 0.1167 |
433
+ | cosine_precision@5 | 0.0772 |
434
+ | cosine_precision@10 | 0.0433 |
435
+ | cosine_recall@1 | 0.268 |
436
+ | cosine_recall@3 | 0.35 |
437
+ | cosine_recall@5 | 0.3859 |
438
+ | cosine_recall@10 | 0.4332 |
439
+ | cosine_ndcg@10 | 0.3461 |
440
+ | cosine_mrr@10 | 0.3188 |
441
+ | **cosine_map@100** | **0.3252** |
442
+
443
+ #### Information Retrieval
444
+ * Dataset: `dim_512`
445
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
446
+
447
+ | Metric | Value |
448
+ |:--------------------|:-----------|
449
+ | cosine_accuracy@1 | 0.2655 |
450
+ | cosine_accuracy@3 | 0.346 |
451
+ | cosine_accuracy@5 | 0.3811 |
452
+ | cosine_accuracy@10 | 0.4291 |
453
+ | cosine_precision@1 | 0.2655 |
454
+ | cosine_precision@3 | 0.1153 |
455
+ | cosine_precision@5 | 0.0762 |
456
+ | cosine_precision@10 | 0.0429 |
457
+ | cosine_recall@1 | 0.2655 |
458
+ | cosine_recall@3 | 0.346 |
459
+ | cosine_recall@5 | 0.3811 |
460
+ | cosine_recall@10 | 0.4291 |
461
+ | cosine_ndcg@10 | 0.3425 |
462
+ | cosine_mrr@10 | 0.3154 |
463
+ | **cosine_map@100** | **0.3217** |
464
+
465
+ #### Information Retrieval
466
+ * Dataset: `dim_256`
467
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
468
+
469
+ | Metric | Value |
470
+ |:--------------------|:-----------|
471
+ | cosine_accuracy@1 | 0.2576 |
472
+ | cosine_accuracy@3 | 0.3379 |
473
+ | cosine_accuracy@5 | 0.3717 |
474
+ | cosine_accuracy@10 | 0.4195 |
475
+ | cosine_precision@1 | 0.2576 |
476
+ | cosine_precision@3 | 0.1126 |
477
+ | cosine_precision@5 | 0.0743 |
478
+ | cosine_precision@10 | 0.042 |
479
+ | cosine_recall@1 | 0.2576 |
480
+ | cosine_recall@3 | 0.3379 |
481
+ | cosine_recall@5 | 0.3717 |
482
+ | cosine_recall@10 | 0.4195 |
483
+ | cosine_ndcg@10 | 0.3339 |
484
+ | cosine_mrr@10 | 0.3071 |
485
+ | **cosine_map@100** | **0.3134** |
486
+
487
+ #### Information Retrieval
488
+ * Dataset: `dim_128`
489
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
490
+
491
+ | Metric | Value |
492
+ |:--------------------|:-----------|
493
+ | cosine_accuracy@1 | 0.2449 |
494
+ | cosine_accuracy@3 | 0.3219 |
495
+ | cosine_accuracy@5 | 0.3557 |
496
+ | cosine_accuracy@10 | 0.4029 |
497
+ | cosine_precision@1 | 0.2449 |
498
+ | cosine_precision@3 | 0.1073 |
499
+ | cosine_precision@5 | 0.0711 |
500
+ | cosine_precision@10 | 0.0403 |
501
+ | cosine_recall@1 | 0.2449 |
502
+ | cosine_recall@3 | 0.3219 |
503
+ | cosine_recall@5 | 0.3557 |
504
+ | cosine_recall@10 | 0.4029 |
505
+ | cosine_ndcg@10 | 0.3191 |
506
+ | cosine_mrr@10 | 0.2929 |
507
+ | **cosine_map@100** | **0.2992** |
508
+
509
+ #### Information Retrieval
510
+ * Dataset: `dim_64`
511
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
512
+
513
+ | Metric | Value |
514
+ |:--------------------|:----------|
515
+ | cosine_accuracy@1 | 0.2194 |
516
+ | cosine_accuracy@3 | 0.2918 |
517
+ | cosine_accuracy@5 | 0.3235 |
518
+ | cosine_accuracy@10 | 0.3699 |
519
+ | cosine_precision@1 | 0.2194 |
520
+ | cosine_precision@3 | 0.0973 |
521
+ | cosine_precision@5 | 0.0647 |
522
+ | cosine_precision@10 | 0.037 |
523
+ | cosine_recall@1 | 0.2194 |
524
+ | cosine_recall@3 | 0.2918 |
525
+ | cosine_recall@5 | 0.3235 |
526
+ | cosine_recall@10 | 0.3699 |
527
+ | cosine_ndcg@10 | 0.2897 |
528
+ | cosine_mrr@10 | 0.2647 |
529
+ | **cosine_map@100** | **0.271** |
530
+
531
+ <!--
532
+ ## Bias, Risks and Limitations
533
+
534
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
535
+ -->
536
+
537
+ <!--
538
+ ### Recommendations
539
+
540
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
541
+ -->
542
+
543
+ ## Training Details
544
+
545
+ ### Training Dataset
546
+
547
+ #### Unnamed Dataset
548
+
549
+
550
+ * Size: 282,883 training samples
551
+ * Columns: <code>positive</code> and <code>anchor</code>
552
+ * Approximate statistics based on the first 1000 samples:
553
+ | | positive | anchor |
554
+ |:--------|:---------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
555
+ | type | string | string |
556
+ | details | <ul><li>min: 5 tokens</li><li>mean: 20.1 tokens</li><li>max: 78 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 38.64 tokens</li><li>max: 184 tokens</li></ul> |
557
+ * Samples:
558
+ | positive | anchor |
559
+ |:----------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
560
+ | <code>Alingoja mtu huyo mwingine arudi.</code> | <code>Ca'daan alingoja hadi alasiri yote mtu huyo atoke tena.</code> |
561
+ | <code>Sheria hiyo huanzisha mfululizo wa ukaguzi wa majaribio.</code> | <code>Sheria hiyo pia inatoa sheria ya kudhibiti kwa ajili ya mashirika fulani ambayo yanahitaji kuandaa taarifa za kifedha za mashirika yote na kuzisimamisha kwa wakaguzi wa jumla.</code> |
562
+ | <code>Mbwa anakimbia na kuruka nje.</code> | <code>Mbwa mwenye rangi ya kahawia anaruka na kukimbia shambani.</code> |
563
+ * Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
564
+ ```json
565
+ {
566
+ "loss": "MultipleNegativesRankingLoss",
567
+ "matryoshka_dims": [
568
+ 768,
569
+ 512,
570
+ 256,
571
+ 128,
572
+ 64
573
+ ],
574
+ "matryoshka_weights": [
575
+ 1,
576
+ 1,
577
+ 1,
578
+ 1,
579
+ 1
580
+ ],
581
+ "n_dims_per_step": -1
582
+ }
583
+ ```
584
+
585
+ ### Training Hyperparameters
586
+ #### Non-Default Hyperparameters
587
+
588
+ - `eval_strategy`: epoch
589
+ - `per_device_train_batch_size`: 32
590
+ - `per_device_eval_batch_size`: 16
591
+ - `gradient_accumulation_steps`: 16
592
+ - `learning_rate`: 2e-05
593
+ - `num_train_epochs`: 4
594
+ - `lr_scheduler_type`: cosine
595
+ - `warmup_ratio`: 0.1
596
+ - `bf16`: True
597
+ - `tf32`: True
598
+ - `load_best_model_at_end`: True
599
+ - `optim`: adamw_torch_fused
600
+ - `batch_sampler`: no_duplicates
601
+
602
+ #### All Hyperparameters
603
+ <details><summary>Click to expand</summary>
604
+
605
+ - `overwrite_output_dir`: False
606
+ - `do_predict`: False
607
+ - `eval_strategy`: epoch
608
+ - `prediction_loss_only`: True
609
+ - `per_device_train_batch_size`: 32
610
+ - `per_device_eval_batch_size`: 16
611
+ - `per_gpu_train_batch_size`: None
612
+ - `per_gpu_eval_batch_size`: None
613
+ - `gradient_accumulation_steps`: 16
614
+ - `eval_accumulation_steps`: None
615
+ - `learning_rate`: 2e-05
616
+ - `weight_decay`: 0.0
617
+ - `adam_beta1`: 0.9
618
+ - `adam_beta2`: 0.999
619
+ - `adam_epsilon`: 1e-08
620
+ - `max_grad_norm`: 1.0
621
+ - `num_train_epochs`: 4
622
+ - `max_steps`: -1
623
+ - `lr_scheduler_type`: cosine
624
+ - `lr_scheduler_kwargs`: {}
625
+ - `warmup_ratio`: 0.1
626
+ - `warmup_steps`: 0
627
+ - `log_level`: passive
628
+ - `log_level_replica`: warning
629
+ - `log_on_each_node`: True
630
+ - `logging_nan_inf_filter`: True
631
+ - `save_safetensors`: True
632
+ - `save_on_each_node`: False
633
+ - `save_only_model`: False
634
+ - `restore_callback_states_from_checkpoint`: False
635
+ - `no_cuda`: False
636
+ - `use_cpu`: False
637
+ - `use_mps_device`: False
638
+ - `seed`: 42
639
+ - `data_seed`: None
640
+ - `jit_mode_eval`: False
641
+ - `use_ipex`: False
642
+ - `bf16`: True
643
+ - `fp16`: False
644
+ - `fp16_opt_level`: O1
645
+ - `half_precision_backend`: auto
646
+ - `bf16_full_eval`: False
647
+ - `fp16_full_eval`: False
648
+ - `tf32`: True
649
+ - `local_rank`: 0
650
+ - `ddp_backend`: None
651
+ - `tpu_num_cores`: None
652
+ - `tpu_metrics_debug`: False
653
+ - `debug`: []
654
+ - `dataloader_drop_last`: False
655
+ - `dataloader_num_workers`: 0
656
+ - `dataloader_prefetch_factor`: None
657
+ - `past_index`: -1
658
+ - `disable_tqdm`: False
659
+ - `remove_unused_columns`: True
660
+ - `label_names`: None
661
+ - `load_best_model_at_end`: True
662
+ - `ignore_data_skip`: False
663
+ - `fsdp`: []
664
+ - `fsdp_min_num_params`: 0
665
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
666
+ - `fsdp_transformer_layer_cls_to_wrap`: None
667
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
668
+ - `deepspeed`: None
669
+ - `label_smoothing_factor`: 0.0
670
+ - `optim`: adamw_torch_fused
671
+ - `optim_args`: None
672
+ - `adafactor`: False
673
+ - `group_by_length`: False
674
+ - `length_column_name`: length
675
+ - `ddp_find_unused_parameters`: None
676
+ - `ddp_bucket_cap_mb`: None
677
+ - `ddp_broadcast_buffers`: False
678
+ - `dataloader_pin_memory`: True
679
+ - `dataloader_persistent_workers`: False
680
+ - `skip_memory_metrics`: True
681
+ - `use_legacy_prediction_loop`: False
682
+ - `push_to_hub`: False
683
+ - `resume_from_checkpoint`: None
684
+ - `hub_model_id`: None
685
+ - `hub_strategy`: every_save
686
+ - `hub_private_repo`: False
687
+ - `hub_always_push`: False
688
+ - `gradient_checkpointing`: False
689
+ - `gradient_checkpointing_kwargs`: None
690
+ - `include_inputs_for_metrics`: False
691
+ - `eval_do_concat_batches`: True
692
+ - `fp16_backend`: auto
693
+ - `push_to_hub_model_id`: None
694
+ - `push_to_hub_organization`: None
695
+ - `mp_parameters`:
696
+ - `auto_find_batch_size`: False
697
+ - `full_determinism`: False
698
+ - `torchdynamo`: None
699
+ - `ray_scope`: last
700
+ - `ddp_timeout`: 1800
701
+ - `torch_compile`: False
702
+ - `torch_compile_backend`: None
703
+ - `torch_compile_mode`: None
704
+ - `dispatch_batches`: None
705
+ - `split_batches`: None
706
+ - `include_tokens_per_second`: False
707
+ - `include_num_input_tokens_seen`: False
708
+ - `neftune_noise_alpha`: None
709
+ - `optim_target_modules`: None
710
+ - `batch_eval_metrics`: False
711
+ - `batch_sampler`: no_duplicates
712
+ - `multi_dataset_batch_sampler`: proportional
713
+
714
+ </details>
715
+
716
+ ### Training Logs
717
+ <details><summary>Click to expand</summary>
718
+
719
+ | Epoch | Step | Training Loss | dim_128_cosine_map@100 | dim_256_cosine_map@100 | dim_512_cosine_map@100 | dim_64_cosine_map@100 | dim_768_cosine_map@100 |
720
+ |:----------:|:--------:|:-------------:|:----------------------:|:----------------------:|:----------------------:|:---------------------:|:----------------------:|
721
+ | 0.0181 | 10 | 9.7089 | - | - | - | - | - |
722
+ | 0.0362 | 20 | 9.2806 | - | - | - | - | - |
723
+ | 0.0543 | 30 | 8.8905 | - | - | - | - | - |
724
+ | 0.0724 | 40 | 7.9651 | - | - | - | - | - |
725
+ | 0.0905 | 50 | 7.4201 | - | - | - | - | - |
726
+ | 0.1086 | 60 | 6.8346 | - | - | - | - | - |
727
+ | 0.1267 | 70 | 6.515 | - | - | - | - | - |
728
+ | 0.1448 | 80 | 6.2009 | - | - | - | - | - |
729
+ | 0.1629 | 90 | 5.8256 | - | - | - | - | - |
730
+ | 0.1810 | 100 | 5.549 | - | - | - | - | - |
731
+ | 0.1991 | 110 | 5.1667 | - | - | - | - | - |
732
+ | 0.2172 | 120 | 5.2684 | - | - | - | - | - |
733
+ | 0.2353 | 130 | 5.0678 | - | - | - | - | - |
734
+ | 0.2534 | 140 | 4.9183 | - | - | - | - | - |
735
+ | 0.2715 | 150 | 4.844 | - | - | - | - | - |
736
+ | 0.2896 | 160 | 4.5427 | - | - | - | - | - |
737
+ | 0.3077 | 170 | 4.3324 | - | - | - | - | - |
738
+ | 0.3258 | 180 | 4.4963 | - | - | - | - | - |
739
+ | 0.3439 | 190 | 4.1704 | - | - | - | - | - |
740
+ | 0.3620 | 200 | 4.1285 | - | - | - | - | - |
741
+ | 0.3800 | 210 | 4.0235 | - | - | - | - | - |
742
+ | 0.3981 | 220 | 4.0738 | - | - | - | - | - |
743
+ | 0.4162 | 230 | 3.9132 | - | - | - | - | - |
744
+ | 0.4343 | 240 | 3.9682 | - | - | - | - | - |
745
+ | 0.4524 | 250 | 3.7542 | - | - | - | - | - |
746
+ | 0.4705 | 260 | 3.6508 | - | - | - | - | - |
747
+ | 0.4886 | 270 | 3.7596 | - | - | - | - | - |
748
+ | 0.5067 | 280 | 3.5596 | - | - | - | - | - |
749
+ | 0.5248 | 290 | 3.5077 | - | - | - | - | - |
750
+ | 0.5429 | 300 | 3.3831 | - | - | - | - | - |
751
+ | 0.5610 | 310 | 3.4 | - | - | - | - | - |
752
+ | 0.5791 | 320 | 3.296 | - | - | - | - | - |
753
+ | 0.5972 | 330 | 3.3646 | - | - | - | - | - |
754
+ | 0.6153 | 340 | 3.3533 | - | - | - | - | - |
755
+ | 0.6334 | 350 | 3.2171 | - | - | - | - | - |
756
+ | 0.6515 | 360 | 3.2324 | - | - | - | - | - |
757
+ | 0.6696 | 370 | 3.1544 | - | - | - | - | - |
758
+ | 0.6877 | 380 | 3.3393 | - | - | - | - | - |
759
+ | 0.7058 | 390 | 3.0864 | - | - | - | - | - |
760
+ | 0.7239 | 400 | 3.1069 | - | - | - | - | - |
761
+ | 0.7420 | 410 | 3.0722 | - | - | - | - | - |
762
+ | 0.7601 | 420 | 3.1446 | - | - | - | - | - |
763
+ | 0.7782 | 430 | 3.0847 | - | - | - | - | - |
764
+ | 0.7963 | 440 | 3.0331 | - | - | - | - | - |
765
+ | 0.8144 | 450 | 3.0197 | - | - | - | - | - |
766
+ | 0.8325 | 460 | 2.9667 | - | - | - | - | - |
767
+ | 0.8506 | 470 | 2.8331 | - | - | - | - | - |
768
+ | 0.8687 | 480 | 2.9333 | - | - | - | - | - |
769
+ | 0.8868 | 490 | 2.8714 | - | - | - | - | - |
770
+ | 0.9049 | 500 | 2.8578 | - | - | - | - | - |
771
+ | 0.9230 | 510 | 2.9689 | - | - | - | - | - |
772
+ | 0.9411 | 520 | 2.7977 | - | - | - | - | - |
773
+ | 0.9592 | 530 | 2.9832 | - | - | - | - | - |
774
+ | 0.9773 | 540 | 2.9761 | - | - | - | - | - |
775
+ | 0.9954 | 550 | 2.7711 | - | - | - | - | - |
776
+ | 0.9990 | 552 | - | 0.2772 | 0.2954 | 0.3052 | 0.2445 | 0.3080 |
777
+ | 1.0135 | 560 | 2.7194 | - | - | - | - | - |
778
+ | 1.0316 | 570 | 2.8489 | - | - | - | - | - |
779
+ | 1.0497 | 580 | 2.6559 | - | - | - | - | - |
780
+ | 1.0678 | 590 | 2.6239 | - | - | - | - | - |
781
+ | 1.0859 | 600 | 2.7081 | - | - | - | - | - |
782
+ | 1.1039 | 610 | 2.6581 | - | - | - | - | - |
783
+ | 1.1220 | 620 | 2.7709 | - | - | - | - | - |
784
+ | 1.1401 | 630 | 2.6191 | - | - | - | - | - |
785
+ | 1.1582 | 640 | 2.6712 | - | - | - | - | - |
786
+ | 1.1763 | 650 | 2.5445 | - | - | - | - | - |
787
+ | 1.1944 | 660 | 2.5264 | - | - | - | - | - |
788
+ | 1.2125 | 670 | 2.5782 | - | - | - | - | - |
789
+ | 1.2306 | 680 | 2.5652 | - | - | - | - | - |
790
+ | 1.2487 | 690 | 2.6229 | - | - | - | - | - |
791
+ | 1.2668 | 700 | 2.5557 | - | - | - | - | - |
792
+ | 1.2849 | 710 | 2.5251 | - | - | - | - | - |
793
+ | 1.3030 | 720 | 2.4555 | - | - | - | - | - |
794
+ | 1.3211 | 730 | 2.5335 | - | - | - | - | - |
795
+ | 1.3392 | 740 | 2.5027 | - | - | - | - | - |
796
+ | 1.3573 | 750 | 2.3569 | - | - | - | - | - |
797
+ | 1.3754 | 760 | 2.4255 | - | - | - | - | - |
798
+ | 1.3935 | 770 | 2.4626 | - | - | - | - | - |
799
+ | 1.4116 | 780 | 2.363 | - | - | - | - | - |
800
+ | 1.4297 | 790 | 2.4 | - | - | - | - | - |
801
+ | 1.4478 | 800 | 2.3317 | - | - | - | - | - |
802
+ | 1.4659 | 810 | 2.2922 | - | - | - | - | - |
803
+ | 1.4840 | 820 | 2.4086 | - | - | - | - | - |
804
+ | 1.5021 | 830 | 2.3166 | - | - | - | - | - |
805
+ | 1.5202 | 840 | 2.3401 | - | - | - | - | - |
806
+ | 1.5383 | 850 | 2.1951 | - | - | - | - | - |
807
+ | 1.5564 | 860 | 2.214 | - | - | - | - | - |
808
+ | 1.5745 | 870 | 2.1859 | - | - | - | - | - |
809
+ | 1.5926 | 880 | 2.3605 | - | - | - | - | - |
810
+ | 1.6107 | 890 | 2.2528 | - | - | - | - | - |
811
+ | 1.6288 | 900 | 2.2759 | - | - | - | - | - |
812
+ | 1.6469 | 910 | 2.1458 | - | - | - | - | - |
813
+ | 1.6650 | 920 | 2.187 | - | - | - | - | - |
814
+ | 1.6831 | 930 | 2.3406 | - | - | - | - | - |
815
+ | 1.7012 | 940 | 2.2151 | - | - | - | - | - |
816
+ | 1.7193 | 950 | 2.2971 | - | - | - | - | - |
817
+ | 1.7374 | 960 | 2.2736 | - | - | - | - | - |
818
+ | 1.7555 | 970 | 2.2329 | - | - | - | - | - |
819
+ | 1.7736 | 980 | 2.2602 | - | - | - | - | - |
820
+ | 1.7917 | 990 | 2.2402 | - | - | - | - | - |
821
+ | 1.8098 | 1000 | 2.1971 | - | - | - | - | - |
822
+ | 1.8278 | 1010 | 2.1642 | - | - | - | - | - |
823
+ | 1.8459 | 1020 | 2.1274 | - | - | - | - | - |
824
+ | 1.8640 | 1030 | 2.1833 | - | - | - | - | - |
825
+ | 1.8821 | 1040 | 2.156 | - | - | - | - | - |
826
+ | 1.9002 | 1050 | 2.1252 | - | - | - | - | - |
827
+ | 1.9183 | 1060 | 2.161 | - | - | - | - | - |
828
+ | 1.9364 | 1070 | 2.1267 | - | - | - | - | - |
829
+ | 1.9545 | 1080 | 2.2017 | - | - | - | - | - |
830
+ | 1.9726 | 1090 | 2.3044 | - | - | - | - | - |
831
+ | 1.9907 | 1100 | 2.161 | - | - | - | - | - |
832
+ | 1.9998 | 1105 | - | 0.2928 | 0.3085 | 0.3165 | 0.2632 | 0.3204 |
833
+ | 2.0088 | 1110 | 2.0594 | - | - | - | - | - |
834
+ | 2.0269 | 1120 | 2.2277 | - | - | - | - | - |
835
+ | 2.0450 | 1130 | 2.1591 | - | - | - | - | - |
836
+ | 2.0631 | 1140 | 2.0396 | - | - | - | - | - |
837
+ | 2.0812 | 1150 | 2.1007 | - | - | - | - | - |
838
+ | 2.0993 | 1160 | 2.0705 | - | - | - | - | - |
839
+ | 2.1174 | 1170 | 2.0894 | - | - | - | - | - |
840
+ | 2.1355 | 1180 | 2.0677 | - | - | - | - | - |
841
+ | 2.1536 | 1190 | 2.0893 | - | - | - | - | - |
842
+ | 2.1717 | 1200 | 1.984 | - | - | - | - | - |
843
+ | 2.1898 | 1210 | 1.9206 | - | - | - | - | - |
844
+ | 2.2079 | 1220 | 2.132 | - | - | - | - | - |
845
+ | 2.2260 | 1230 | 2.0457 | - | - | - | - | - |
846
+ | 2.2441 | 1240 | 2.1428 | - | - | - | - | - |
847
+ | 2.2622 | 1250 | 2.1116 | - | - | - | - | - |
848
+ | 2.2803 | 1260 | 1.993 | - | - | - | - | - |
849
+ | 2.2984 | 1270 | 2.0181 | - | - | - | - | - |
850
+ | 2.3165 | 1280 | 1.9742 | - | - | - | - | - |
851
+ | 2.3346 | 1290 | 2.081 | - | - | - | - | - |
852
+ | 2.3527 | 1300 | 1.9107 | - | - | - | - | - |
853
+ | 2.3708 | 1310 | 1.9507 | - | - | - | - | - |
854
+ | 2.3889 | 1320 | 1.9844 | - | - | - | - | - |
855
+ | 2.4070 | 1330 | 2.0035 | - | - | - | - | - |
856
+ | 2.4251 | 1340 | 1.9121 | - | - | - | - | - |
857
+ | 2.4432 | 1350 | 2.0057 | - | - | - | - | - |
858
+ | 2.4613 | 1360 | 1.9323 | - | - | - | - | - |
859
+ | 2.4794 | 1370 | 1.9216 | - | - | - | - | - |
860
+ | 2.4975 | 1380 | 1.995 | - | - | - | - | - |
861
+ | 2.5156 | 1390 | 1.9285 | - | - | - | - | - |
862
+ | 2.5337 | 1400 | 1.8886 | - | - | - | - | - |
863
+ | 2.5517 | 1410 | 1.8298 | - | - | - | - | - |
864
+ | 2.5698 | 1420 | 1.8452 | - | - | - | - | - |
865
+ | 2.5879 | 1430 | 1.9488 | - | - | - | - | - |
866
+ | 2.6060 | 1440 | 1.8928 | - | - | - | - | - |
867
+ | 2.6241 | 1450 | 2.0101 | - | - | - | - | - |
868
+ | 2.6422 | 1460 | 1.7591 | - | - | - | - | - |
869
+ | 2.6603 | 1470 | 1.9177 | - | - | - | - | - |
870
+ | 2.6784 | 1480 | 1.9329 | - | - | - | - | - |
871
+ | 2.6965 | 1490 | 1.8978 | - | - | - | - | - |
872
+ | 2.7146 | 1500 | 1.9589 | - | - | - | - | - |
873
+ | 2.7327 | 1510 | 1.9744 | - | - | - | - | - |
874
+ | 2.7508 | 1520 | 1.9272 | - | - | - | - | - |
875
+ | 2.7689 | 1530 | 1.9234 | - | - | - | - | - |
876
+ | 2.7870 | 1540 | 1.9667 | - | - | - | - | - |
877
+ | 2.8051 | 1550 | 1.853 | - | - | - | - | - |
878
+ | 2.8232 | 1560 | 1.9191 | - | - | - | - | - |
879
+ | 2.8413 | 1570 | 1.8083 | - | - | - | - | - |
880
+ | 2.8594 | 1580 | 1.8543 | - | - | - | - | - |
881
+ | 2.8775 | 1590 | 1.9091 | - | - | - | - | - |
882
+ | 2.8956 | 1600 | 1.8079 | - | - | - | - | - |
883
+ | 2.9137 | 1610 | 1.8992 | - | - | - | - | - |
884
+ | 2.9318 | 1620 | 1.8742 | - | - | - | - | - |
885
+ | 2.9499 | 1630 | 1.9313 | - | - | - | - | - |
886
+ | 2.9680 | 1640 | 1.9832 | - | - | - | - | - |
887
+ | 2.9861 | 1650 | 1.9037 | - | - | - | - | - |
888
+ | 2.9988 | 1657 | - | 0.2982 | 0.3130 | 0.3211 | 0.2697 | 0.3247 |
889
+ | 3.0042 | 1660 | 1.7924 | - | - | - | - | - |
890
+ | 3.0223 | 1670 | 1.9677 | - | - | - | - | - |
891
+ | 3.0404 | 1680 | 1.9123 | - | - | - | - | - |
892
+ | 3.0585 | 1690 | 1.7691 | - | - | - | - | - |
893
+ | 3.0766 | 1700 | 1.8822 | - | - | - | - | - |
894
+ | 3.0947 | 1710 | 1.8543 | - | - | - | - | - |
895
+ | 3.1128 | 1720 | 1.8127 | - | - | - | - | - |
896
+ | 3.1309 | 1730 | 1.8844 | - | - | - | - | - |
897
+ | 3.1490 | 1740 | 1.911 | - | - | - | - | - |
898
+ | 3.1671 | 1750 | 1.7695 | - | - | - | - | - |
899
+ | 3.1852 | 1760 | 1.8134 | - | - | - | - | - |
900
+ | 3.2033 | 1770 | 1.7794 | - | - | - | - | - |
901
+ | 3.2214 | 1780 | 1.8851 | - | - | - | - | - |
902
+ | 3.2395 | 1790 | 1.8381 | - | - | - | - | - |
903
+ | 3.2576 | 1800 | 1.9184 | - | - | - | - | - |
904
+ | 3.2756 | 1810 | 1.8074 | - | - | - | - | - |
905
+ | 3.2937 | 1820 | 1.8236 | - | - | - | - | - |
906
+ | 3.3118 | 1830 | 1.8203 | - | - | - | - | - |
907
+ | 3.3299 | 1840 | 1.8874 | - | - | - | - | - |
908
+ | 3.3480 | 1850 | 1.7457 | - | - | - | - | - |
909
+ | 3.3661 | 1860 | 1.7933 | - | - | - | - | - |
910
+ | 3.3842 | 1870 | 1.759 | - | - | - | - | - |
911
+ | 3.4023 | 1880 | 1.8514 | - | - | - | - | - |
912
+ | 3.4204 | 1890 | 1.8163 | - | - | - | - | - |
913
+ | 3.4385 | 1900 | 1.8299 | - | - | - | - | - |
914
+ | 3.4566 | 1910 | 1.8112 | - | - | - | - | - |
915
+ | 3.4747 | 1920 | 1.7446 | - | - | - | - | - |
916
+ | 3.4928 | 1930 | 1.8314 | - | - | - | - | - |
917
+ | 3.5109 | 1940 | 1.742 | - | - | - | - | - |
918
+ | 3.5290 | 1950 | 1.7519 | - | - | - | - | - |
919
+ | 3.5471 | 1960 | 1.722 | - | - | - | - | - |
920
+ | 3.5652 | 1970 | 1.7454 | - | - | - | - | - |
921
+ | 3.5833 | 1980 | 1.7875 | - | - | - | - | - |
922
+ | 3.6014 | 1990 | 1.7596 | - | - | - | - | - |
923
+ | 3.6195 | 2000 | 1.8348 | - | - | - | - | - |
924
+ | 3.6376 | 2010 | 1.6954 | - | - | - | - | - |
925
+ | 3.6557 | 2020 | 1.7334 | - | - | - | - | - |
926
+ | 3.6738 | 2030 | 1.8318 | - | - | - | - | - |
927
+ | 3.6919 | 2040 | 1.7982 | - | - | - | - | - |
928
+ | 3.7100 | 2050 | 1.7987 | - | - | - | - | - |
929
+ | 3.7281 | 2060 | 1.8402 | - | - | - | - | - |
930
+ | 3.7462 | 2070 | 1.8569 | - | - | - | - | - |
931
+ | 3.7643 | 2080 | 1.8285 | - | - | - | - | - |
932
+ | 3.7824 | 2090 | 1.8652 | - | - | - | - | - |
933
+ | 3.8005 | 2100 | 1.7731 | - | - | - | - | - |
934
+ | 3.8186 | 2110 | 1.8697 | - | - | - | - | - |
935
+ | 3.8367 | 2120 | 1.6953 | - | - | - | - | - |
936
+ | 3.8548 | 2130 | 1.7493 | - | - | - | - | - |
937
+ | 3.8729 | 2140 | 1.8031 | - | - | - | - | - |
938
+ | 3.8910 | 2150 | 1.7053 | - | - | - | - | - |
939
+ | 3.9091 | 2160 | 1.8436 | - | - | - | - | - |
940
+ | 3.9272 | 2170 | 1.7572 | - | - | - | - | - |
941
+ | 3.9453 | 2180 | 1.7797 | - | - | - | - | - |
942
+ | 3.9634 | 2190 | 1.8827 | - | - | - | - | - |
943
+ | 3.9815 | 2200 | 1.8678 | - | - | - | - | - |
944
+ | **3.9959** | **2208** | **-** | **0.2992** | **0.3134** | **0.3217** | **0.271** | **0.3252** |
945
+
946
+ * The bold row denotes the saved checkpoint.
947
+ </details>
948
+
949
+ ### Framework Versions
950
+ - Python: 3.10.12
951
+ - Sentence Transformers: 3.0.1
952
+ - Transformers: 4.41.2
953
+ - PyTorch: 2.1.2+cu121
954
+ - Accelerate: 0.31.0
955
+ - Datasets: 2.19.1
956
+ - Tokenizers: 0.19.1
957
+
958
+ ## Citation
959
+
960
+ ### BibTeX
961
+
962
+ #### Sentence Transformers
963
+ ```bibtex
964
+ @inproceedings{reimers-2019-sentence-bert,
965
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
966
+ author = "Reimers, Nils and Gurevych, Iryna",
967
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
968
+ month = "11",
969
+ year = "2019",
970
+ publisher = "Association for Computational Linguistics",
971
+ url = "https://arxiv.org/abs/1908.10084",
972
+ }
973
+ ```
974
+
975
+ #### MatryoshkaLoss
976
+ ```bibtex
977
+ @misc{kusupati2024matryoshka,
978
+ title={Matryoshka Representation Learning},
979
+ 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},
980
+ year={2024},
981
+ eprint={2205.13147},
982
+ archivePrefix={arXiv},
983
+ primaryClass={cs.LG}
984
+ }
985
+ ```
986
+
987
+ #### MultipleNegativesRankingLoss
988
+ ```bibtex
989
+ @misc{henderson2017efficient,
990
+ title={Efficient Natural Language Response Suggestion for Smart Reply},
991
+ author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
992
+ year={2017},
993
+ eprint={1705.00652},
994
+ archivePrefix={arXiv},
995
+ primaryClass={cs.CL}
996
+ }
997
+ ```
998
+
999
+ <!--
1000
+ ## Glossary
1001
+
1002
+ *Clearly define terms in order to be accessible across audiences.*
1003
+ -->
1004
+
1005
+ <!--
1006
+ ## Model Card Authors
1007
+
1008
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
1009
+ -->
1010
+
1011
+ <!--
1012
+ ## Model Card Contact
1013
+
1014
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
1015
+ -->
config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "BAAI/bge-base-en-v1.5",
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
+ "id2label": {
13
+ "0": "LABEL_0"
14
+ },
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 3072,
17
+ "label2id": {
18
+ "LABEL_0": 0
19
+ },
20
+ "layer_norm_eps": 1e-12,
21
+ "max_position_embeddings": 512,
22
+ "model_type": "bert",
23
+ "num_attention_heads": 12,
24
+ "num_hidden_layers": 12,
25
+ "pad_token_id": 0,
26
+ "position_embedding_type": "absolute",
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.41.2",
29
+ "type_vocab_size": 2,
30
+ "use_cache": true,
31
+ "vocab_size": 30522
32
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "3.0.1",
4
+ "transformers": "4.41.2",
5
+ "pytorch": "2.1.2+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 sha256:3800ab347872a796d31dc76526b2613b32e623fa612613d196e69ae08009d927
3
+ size 437951328
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": true
4
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "never_split": null,
51
+ "pad_token": "[PAD]",
52
+ "sep_token": "[SEP]",
53
+ "strip_accents": null,
54
+ "tokenize_chinese_chars": true,
55
+ "tokenizer_class": "BertTokenizer",
56
+ "unk_token": "[UNK]"
57
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff