akot commited on
Commit
ac45bff
1 Parent(s): 6eefb55

Add new SentenceTransformer model.

Browse files
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 1024,
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 +1,837 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: aari1995/German_Semantic_V3
3
+ datasets: []
4
+ language:
5
+ - en
6
+ library_name: sentence-transformers
7
+ license: apache-2.0
8
+ metrics:
9
+ - cosine_accuracy@1
10
+ - cosine_accuracy@3
11
+ - cosine_accuracy@5
12
+ - cosine_accuracy@10
13
+ - cosine_precision@1
14
+ - cosine_precision@3
15
+ - cosine_precision@5
16
+ - cosine_precision@10
17
+ - cosine_recall@1
18
+ - cosine_recall@3
19
+ - cosine_recall@5
20
+ - cosine_recall@10
21
+ - cosine_ndcg@10
22
+ - cosine_mrr@10
23
+ - cosine_map@100
24
+ pipeline_tag: sentence-similarity
25
+ tags:
26
+ - sentence-transformers
27
+ - sentence-similarity
28
+ - feature-extraction
29
+ - generated_from_trainer
30
+ - dataset_size:4957
31
+ - loss:MatryoshkaLoss
32
+ - loss:MultipleNegativesRankingLoss
33
+ widget:
34
+ - source_sentence: 312 Aus steuerlicher Sicht ist es möglich, mehrere Versorgungszusagen
35
+ nebeneinander, also neben einer Altzusage auch eine Neuzusage zu erteilen (z.
36
+ B. „alte“ Direktversicherung und „neuer“ Pensionsfonds).
37
+ sentences:
38
+ - Wann liegt bei der betrieblichen Altersversorgung eine schädliche Verwendung vor?
39
+ - Welche steuerliche Behandlung erfahren Auszahlungen aus Altersvorsorgeverträgen
40
+ nach § 22 Nr. 5 EStG?
41
+ - Können verschiedene Versorgungszusagen wie Direktversicherung und Pensionsfonds
42
+ gleichzeitig bestehen?
43
+ - source_sentence: 5 Pflichtversicherte nach dem Gesetz über die Alterssicherung der
44
+ Landwirte gehören, soweit sie nicht als Pflichtversicherte der gesetzlichen Rentenversicherung
45
+ ohnehin bereits anspruchsberechtigt sind, in dieser Eigenschaft ebenfalls zum
46
+ begünstigten Personenkreis. Darunter fallen insbesondere die in Anlage 1 Abschnitt
47
+ B aufgeführten Personen.
48
+ sentences:
49
+ - Wann wird das Anrecht der ausgleichsberechtigten Person bei intern geteilter Altersvorsorge
50
+ als abgeschlossen betrachtet?
51
+ - Welche Personen sind in der Anlage 1 Abschnitt B bezüglich der Alterssicherung
52
+ der Landwirte aufgeführt?
53
+ - In welchen Fällen führt die Möglichkeit einer Beitragserstattung nicht zur Versagung
54
+ der Anerkennung als betriebliche Altersversorgung?
55
+ - source_sentence: 233 Voraussetzung für die Förderung durch Sonderausgabenabzug nach
56
+ § 10a EStG und Zulage nach Abschnitt XI EStG ist in den Fällen der Rz. 231 f.,
57
+ dass der Steuerpflichtige zum begünstigten Personenkreis gehört. Die zeitliche
58
+ Zuordnung dieser Altersvorsorgebeiträge richtet sich grundsätzlich nach § 11 Abs.
59
+ 2 EStG.
60
+ sentences:
61
+ - Wer gehört zum begünstigten Personenkreis für die Altersvorsorgeförderung?
62
+ - Wie werden erstattete Kosten eines Altersvorsorgevertrags besteuert, wenn sie
63
+ dem Steuerpflichtigen ausgezahlt werden?
64
+ - Ist der Übertragungswert einer betrieblichen Altersversorgung bei einem Arbeitgeberwechsel
65
+ steuerfrei?
66
+ - source_sentence: 127 Die Entnahme des Teilkapitalbetrags von bis zu 30 % des zur
67
+ Verfügung stehenden Kapitals aus dem Vertrag hat zu Beginn der Auszahlungsphase
68
+ zu erfolgen. Eine Verteilung über mehrere Auszahlungszeitpunkte ist nicht möglich.
69
+ sentences:
70
+ - Kann ich den Teilkapitalbetrag aus meiner Altersvorsorge zu verschiedenen Zeitpunkten
71
+ entnehmen?
72
+ - Welche Einkunftsarten können Leistungen aus einer Versorgungszusage des Arbeitgebers
73
+ sein?
74
+ - Was ist im Todesfall des Zulageberechtigten bezüglich der Förderbeiträge zu tun?
75
+ - source_sentence: '67 Abwandlung des Beispiels 1 in Rn. 66: A erhält zudem zwei Kinderzulagen
76
+ für seine in den Jahren 2004 und 2005 geborenen Kinder. Beitragspflichtige Einnahmen
77
+ 53.000 € 4 % 2.120 € höchstens 2.100 € anzusetzen 2.100 € abzüglich Zulage 175
78
+ € Mindesteigenbeitrag (§ 86 Abs. 1 Satz 2 EStG) 1.925 € Sockelbetrag (§ 86 Abs.
79
+ 1 Satz 4 EStG) 60 € maßgebend (§ 86 Abs. 1 Satz 5 EStG) 1.925 € Die von A geleisteten
80
+ Beiträge übersteigen den Mindesteigenbeitrag. Die Zulage wird nicht gekürzt.'
81
+ sentences:
82
+ - Wird die Zulage für A gekürzt, wenn die Beiträge den Mindesteigenbeitrag übersteigen?
83
+ - Was versteht man unter Sonderzahlungen des Arbeitgebers?
84
+ - Wie erfolgt die Besteuerung bei der ausgleichsberechtigten Person nach einer externen
85
+ Teilung?
86
+ model-index:
87
+ - name: German Semantic V3 BMF
88
+ results:
89
+ - task:
90
+ type: information-retrieval
91
+ name: Information Retrieval
92
+ dataset:
93
+ name: dim 768
94
+ type: dim_768
95
+ metrics:
96
+ - type: cosine_accuracy@1
97
+ value: 0.02722323049001815
98
+ name: Cosine Accuracy@1
99
+ - type: cosine_accuracy@3
100
+ value: 0.19237749546279492
101
+ name: Cosine Accuracy@3
102
+ - type: cosine_accuracy@5
103
+ value: 0.308529945553539
104
+ name: Cosine Accuracy@5
105
+ - type: cosine_accuracy@10
106
+ value: 0.5081669691470054
107
+ name: Cosine Accuracy@10
108
+ - type: cosine_precision@1
109
+ value: 0.02722323049001815
110
+ name: Cosine Precision@1
111
+ - type: cosine_precision@3
112
+ value: 0.06412583182093164
113
+ name: Cosine Precision@3
114
+ - type: cosine_precision@5
115
+ value: 0.06170598911070781
116
+ name: Cosine Precision@5
117
+ - type: cosine_precision@10
118
+ value: 0.050816696914700546
119
+ name: Cosine Precision@10
120
+ - type: cosine_recall@1
121
+ value: 0.02722323049001815
122
+ name: Cosine Recall@1
123
+ - type: cosine_recall@3
124
+ value: 0.19237749546279492
125
+ name: Cosine Recall@3
126
+ - type: cosine_recall@5
127
+ value: 0.308529945553539
128
+ name: Cosine Recall@5
129
+ - type: cosine_recall@10
130
+ value: 0.5081669691470054
131
+ name: Cosine Recall@10
132
+ - type: cosine_ndcg@10
133
+ value: 0.24120625642015497
134
+ name: Cosine Ndcg@10
135
+ - type: cosine_mrr@10
136
+ value: 0.15931423386051344
137
+ name: Cosine Mrr@10
138
+ - type: cosine_map@100
139
+ value: 0.17848852586462802
140
+ name: Cosine Map@100
141
+ - task:
142
+ type: information-retrieval
143
+ name: Information Retrieval
144
+ dataset:
145
+ name: dim 512
146
+ type: dim_512
147
+ metrics:
148
+ - type: cosine_accuracy@1
149
+ value: 0.021778584392014518
150
+ name: Cosine Accuracy@1
151
+ - type: cosine_accuracy@3
152
+ value: 0.1869328493647913
153
+ name: Cosine Accuracy@3
154
+ - type: cosine_accuracy@5
155
+ value: 0.308529945553539
156
+ name: Cosine Accuracy@5
157
+ - type: cosine_accuracy@10
158
+ value: 0.5208711433756806
159
+ name: Cosine Accuracy@10
160
+ - type: cosine_precision@1
161
+ value: 0.021778584392014518
162
+ name: Cosine Precision@1
163
+ - type: cosine_precision@3
164
+ value: 0.06231094978826376
165
+ name: Cosine Precision@3
166
+ - type: cosine_precision@5
167
+ value: 0.06170598911070781
168
+ name: Cosine Precision@5
169
+ - type: cosine_precision@10
170
+ value: 0.052087114337568054
171
+ name: Cosine Precision@10
172
+ - type: cosine_recall@1
173
+ value: 0.021778584392014518
174
+ name: Cosine Recall@1
175
+ - type: cosine_recall@3
176
+ value: 0.1869328493647913
177
+ name: Cosine Recall@3
178
+ - type: cosine_recall@5
179
+ value: 0.308529945553539
180
+ name: Cosine Recall@5
181
+ - type: cosine_recall@10
182
+ value: 0.5208711433756806
183
+ name: Cosine Recall@10
184
+ - type: cosine_ndcg@10
185
+ value: 0.24282995414753708
186
+ name: Cosine Ndcg@10
187
+ - type: cosine_mrr@10
188
+ value: 0.15777590528044255
189
+ name: Cosine Mrr@10
190
+ - type: cosine_map@100
191
+ value: 0.17621353349099725
192
+ name: Cosine Map@100
193
+ - task:
194
+ type: information-retrieval
195
+ name: Information Retrieval
196
+ dataset:
197
+ name: dim 256
198
+ type: dim_256
199
+ metrics:
200
+ - type: cosine_accuracy@1
201
+ value: 0.019963702359346643
202
+ name: Cosine Accuracy@1
203
+ - type: cosine_accuracy@3
204
+ value: 0.18148820326678766
205
+ name: Cosine Accuracy@3
206
+ - type: cosine_accuracy@5
207
+ value: 0.30490018148820325
208
+ name: Cosine Accuracy@5
209
+ - type: cosine_accuracy@10
210
+ value: 0.5245009074410163
211
+ name: Cosine Accuracy@10
212
+ - type: cosine_precision@1
213
+ value: 0.019963702359346643
214
+ name: Cosine Precision@1
215
+ - type: cosine_precision@3
216
+ value: 0.06049606775559588
217
+ name: Cosine Precision@3
218
+ - type: cosine_precision@5
219
+ value: 0.060980036297640657
220
+ name: Cosine Precision@5
221
+ - type: cosine_precision@10
222
+ value: 0.05245009074410163
223
+ name: Cosine Precision@10
224
+ - type: cosine_recall@1
225
+ value: 0.019963702359346643
226
+ name: Cosine Recall@1
227
+ - type: cosine_recall@3
228
+ value: 0.18148820326678766
229
+ name: Cosine Recall@3
230
+ - type: cosine_recall@5
231
+ value: 0.30490018148820325
232
+ name: Cosine Recall@5
233
+ - type: cosine_recall@10
234
+ value: 0.5245009074410163
235
+ name: Cosine Recall@10
236
+ - type: cosine_ndcg@10
237
+ value: 0.24230231157748117
238
+ name: Cosine Ndcg@10
239
+ - type: cosine_mrr@10
240
+ value: 0.15604888658427682
241
+ name: Cosine Mrr@10
242
+ - type: cosine_map@100
243
+ value: 0.17417213610538765
244
+ name: Cosine Map@100
245
+ - task:
246
+ type: information-retrieval
247
+ name: Information Retrieval
248
+ dataset:
249
+ name: dim 128
250
+ type: dim_128
251
+ metrics:
252
+ - type: cosine_accuracy@1
253
+ value: 0.018148820326678767
254
+ name: Cosine Accuracy@1
255
+ - type: cosine_accuracy@3
256
+ value: 0.1705989110707804
257
+ name: Cosine Accuracy@3
258
+ - type: cosine_accuracy@5
259
+ value: 0.2831215970961887
260
+ name: Cosine Accuracy@5
261
+ - type: cosine_accuracy@10
262
+ value: 0.5136116152450091
263
+ name: Cosine Accuracy@10
264
+ - type: cosine_precision@1
265
+ value: 0.018148820326678767
266
+ name: Cosine Precision@1
267
+ - type: cosine_precision@3
268
+ value: 0.056866303690260134
269
+ name: Cosine Precision@3
270
+ - type: cosine_precision@5
271
+ value: 0.056624319419237755
272
+ name: Cosine Precision@5
273
+ - type: cosine_precision@10
274
+ value: 0.0513611615245009
275
+ name: Cosine Precision@10
276
+ - type: cosine_recall@1
277
+ value: 0.018148820326678767
278
+ name: Cosine Recall@1
279
+ - type: cosine_recall@3
280
+ value: 0.1705989110707804
281
+ name: Cosine Recall@3
282
+ - type: cosine_recall@5
283
+ value: 0.2831215970961887
284
+ name: Cosine Recall@5
285
+ - type: cosine_recall@10
286
+ value: 0.5136116152450091
287
+ name: Cosine Recall@10
288
+ - type: cosine_ndcg@10
289
+ value: 0.23270161109694265
290
+ name: Cosine Ndcg@10
291
+ - type: cosine_mrr@10
292
+ value: 0.14741595367729682
293
+ name: Cosine Mrr@10
294
+ - type: cosine_map@100
295
+ value: 0.16618168136483366
296
+ name: Cosine Map@100
297
+ - task:
298
+ type: information-retrieval
299
+ name: Information Retrieval
300
+ dataset:
301
+ name: dim 64
302
+ type: dim_64
303
+ metrics:
304
+ - type: cosine_accuracy@1
305
+ value: 0.014519056261343012
306
+ name: Cosine Accuracy@1
307
+ - type: cosine_accuracy@3
308
+ value: 0.15245009074410162
309
+ name: Cosine Accuracy@3
310
+ - type: cosine_accuracy@5
311
+ value: 0.2849364791288566
312
+ name: Cosine Accuracy@5
313
+ - type: cosine_accuracy@10
314
+ value: 0.4882032667876588
315
+ name: Cosine Accuracy@10
316
+ - type: cosine_precision@1
317
+ value: 0.014519056261343012
318
+ name: Cosine Precision@1
319
+ - type: cosine_precision@3
320
+ value: 0.050816696914700546
321
+ name: Cosine Precision@3
322
+ - type: cosine_precision@5
323
+ value: 0.056987295825771334
324
+ name: Cosine Precision@5
325
+ - type: cosine_precision@10
326
+ value: 0.04882032667876588
327
+ name: Cosine Precision@10
328
+ - type: cosine_recall@1
329
+ value: 0.014519056261343012
330
+ name: Cosine Recall@1
331
+ - type: cosine_recall@3
332
+ value: 0.15245009074410162
333
+ name: Cosine Recall@3
334
+ - type: cosine_recall@5
335
+ value: 0.2849364791288566
336
+ name: Cosine Recall@5
337
+ - type: cosine_recall@10
338
+ value: 0.4882032667876588
339
+ name: Cosine Recall@10
340
+ - type: cosine_ndcg@10
341
+ value: 0.22104069496061615
342
+ name: Cosine Ndcg@10
343
+ - type: cosine_mrr@10
344
+ value: 0.13950969377466657
345
+ name: Cosine Mrr@10
346
+ - type: cosine_map@100
347
+ value: 0.15832869552609827
348
+ name: Cosine Map@100
349
+ ---
350
+
351
+ # German Semantic V3 BMF
352
+
353
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [aari1995/German_Semantic_V3](https://huggingface.co/aari1995/German_Semantic_V3). It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
354
+
355
+ ## Model Details
356
+
357
+ ### Model Description
358
+ - **Model Type:** Sentence Transformer
359
+ - **Base model:** [aari1995/German_Semantic_V3](https://huggingface.co/aari1995/German_Semantic_V3) <!-- at revision 11b76103bdf441513d7fc14fefae28c1064d3d04 -->
360
+ - **Maximum Sequence Length:** 1024 tokens
361
+ - **Output Dimensionality:** 1024 tokens
362
+ - **Similarity Function:** Cosine Similarity
363
+ <!-- - **Training Dataset:** Unknown -->
364
+ - **Language:** en
365
+ - **License:** apache-2.0
366
+
367
+ ### Model Sources
368
+
369
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
370
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
371
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
372
+
373
+ ### Full Model Architecture
374
+
375
+ ```
376
+ SentenceTransformer(
377
+ (0): Transformer({'max_seq_length': 1024, 'do_lower_case': False}) with Transformer model: BertModel
378
+ (1): Pooling({'word_embedding_dimension': 1024, '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})
379
+ )
380
+ ```
381
+
382
+ ## Usage
383
+
384
+ ### Direct Usage (Sentence Transformers)
385
+
386
+ First install the Sentence Transformers library:
387
+
388
+ ```bash
389
+ pip install -U sentence-transformers
390
+ ```
391
+
392
+ Then you can load this model and run inference.
393
+ ```python
394
+ from sentence_transformers import SentenceTransformer
395
+
396
+ # Download from the 🤗 Hub
397
+ model = SentenceTransformer("akot/german-semantic-bmf-matryoshka-512-10epochs")
398
+ # Run inference
399
+ sentences = [
400
+ '67 Abwandlung des Beispiels 1 in Rn. 66: A erhält zudem zwei Kinderzulagen für seine in den Jahren 2004 und 2005 geborenen Kinder. Beitragspflichtige Einnahmen 53.000 € 4 % 2.120 € höchstens 2.100 € anzusetzen 2.100 € abzüglich Zulage 175 € Mindesteigenbeitrag (§ 86 Abs. 1 Satz 2 EStG) 1.925 € Sockelbetrag (§ 86 Abs. 1 Satz 4 EStG) 60 € maßgebend (§ 86 Abs. 1 Satz 5 EStG) 1.925 € Die von A geleisteten Beiträge übersteigen den Mindesteigenbeitrag. Die Zulage wird nicht gekürzt.',
401
+ 'Wird die Zulage für A gekürzt, wenn die Beiträge den Mindesteigenbeitrag übersteigen?',
402
+ 'Wie erfolgt die Besteuerung bei der ausgleichsberechtigten Person nach einer externen Teilung?',
403
+ ]
404
+ embeddings = model.encode(sentences)
405
+ print(embeddings.shape)
406
+ # [3, 1024]
407
+
408
+ # Get the similarity scores for the embeddings
409
+ similarities = model.similarity(embeddings, embeddings)
410
+ print(similarities.shape)
411
+ # [3, 3]
412
+ ```
413
+
414
+ <!--
415
+ ### Direct Usage (Transformers)
416
+
417
+ <details><summary>Click to see the direct usage in Transformers</summary>
418
+
419
+ </details>
420
+ -->
421
+
422
+ <!--
423
+ ### Downstream Usage (Sentence Transformers)
424
+
425
+ You can finetune this model on your own dataset.
426
+
427
+ <details><summary>Click to expand</summary>
428
+
429
+ </details>
430
+ -->
431
+
432
+ <!--
433
+ ### Out-of-Scope Use
434
+
435
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
436
+ -->
437
+
438
+ ## Evaluation
439
+
440
+ ### Metrics
441
+
442
+ #### Information Retrieval
443
+ * Dataset: `dim_768`
444
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
445
+
446
+ | Metric | Value |
447
+ |:--------------------|:-----------|
448
+ | cosine_accuracy@1 | 0.0272 |
449
+ | cosine_accuracy@3 | 0.1924 |
450
+ | cosine_accuracy@5 | 0.3085 |
451
+ | cosine_accuracy@10 | 0.5082 |
452
+ | cosine_precision@1 | 0.0272 |
453
+ | cosine_precision@3 | 0.0641 |
454
+ | cosine_precision@5 | 0.0617 |
455
+ | cosine_precision@10 | 0.0508 |
456
+ | cosine_recall@1 | 0.0272 |
457
+ | cosine_recall@3 | 0.1924 |
458
+ | cosine_recall@5 | 0.3085 |
459
+ | cosine_recall@10 | 0.5082 |
460
+ | cosine_ndcg@10 | 0.2412 |
461
+ | cosine_mrr@10 | 0.1593 |
462
+ | **cosine_map@100** | **0.1785** |
463
+
464
+ #### Information Retrieval
465
+ * Dataset: `dim_512`
466
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
467
+
468
+ | Metric | Value |
469
+ |:--------------------|:-----------|
470
+ | cosine_accuracy@1 | 0.0218 |
471
+ | cosine_accuracy@3 | 0.1869 |
472
+ | cosine_accuracy@5 | 0.3085 |
473
+ | cosine_accuracy@10 | 0.5209 |
474
+ | cosine_precision@1 | 0.0218 |
475
+ | cosine_precision@3 | 0.0623 |
476
+ | cosine_precision@5 | 0.0617 |
477
+ | cosine_precision@10 | 0.0521 |
478
+ | cosine_recall@1 | 0.0218 |
479
+ | cosine_recall@3 | 0.1869 |
480
+ | cosine_recall@5 | 0.3085 |
481
+ | cosine_recall@10 | 0.5209 |
482
+ | cosine_ndcg@10 | 0.2428 |
483
+ | cosine_mrr@10 | 0.1578 |
484
+ | **cosine_map@100** | **0.1762** |
485
+
486
+ #### Information Retrieval
487
+ * Dataset: `dim_256`
488
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
489
+
490
+ | Metric | Value |
491
+ |:--------------------|:-----------|
492
+ | cosine_accuracy@1 | 0.02 |
493
+ | cosine_accuracy@3 | 0.1815 |
494
+ | cosine_accuracy@5 | 0.3049 |
495
+ | cosine_accuracy@10 | 0.5245 |
496
+ | cosine_precision@1 | 0.02 |
497
+ | cosine_precision@3 | 0.0605 |
498
+ | cosine_precision@5 | 0.061 |
499
+ | cosine_precision@10 | 0.0525 |
500
+ | cosine_recall@1 | 0.02 |
501
+ | cosine_recall@3 | 0.1815 |
502
+ | cosine_recall@5 | 0.3049 |
503
+ | cosine_recall@10 | 0.5245 |
504
+ | cosine_ndcg@10 | 0.2423 |
505
+ | cosine_mrr@10 | 0.156 |
506
+ | **cosine_map@100** | **0.1742** |
507
+
508
+ #### Information Retrieval
509
+ * Dataset: `dim_128`
510
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
511
+
512
+ | Metric | Value |
513
+ |:--------------------|:-----------|
514
+ | cosine_accuracy@1 | 0.0181 |
515
+ | cosine_accuracy@3 | 0.1706 |
516
+ | cosine_accuracy@5 | 0.2831 |
517
+ | cosine_accuracy@10 | 0.5136 |
518
+ | cosine_precision@1 | 0.0181 |
519
+ | cosine_precision@3 | 0.0569 |
520
+ | cosine_precision@5 | 0.0566 |
521
+ | cosine_precision@10 | 0.0514 |
522
+ | cosine_recall@1 | 0.0181 |
523
+ | cosine_recall@3 | 0.1706 |
524
+ | cosine_recall@5 | 0.2831 |
525
+ | cosine_recall@10 | 0.5136 |
526
+ | cosine_ndcg@10 | 0.2327 |
527
+ | cosine_mrr@10 | 0.1474 |
528
+ | **cosine_map@100** | **0.1662** |
529
+
530
+ #### Information Retrieval
531
+ * Dataset: `dim_64`
532
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
533
+
534
+ | Metric | Value |
535
+ |:--------------------|:-----------|
536
+ | cosine_accuracy@1 | 0.0145 |
537
+ | cosine_accuracy@3 | 0.1525 |
538
+ | cosine_accuracy@5 | 0.2849 |
539
+ | cosine_accuracy@10 | 0.4882 |
540
+ | cosine_precision@1 | 0.0145 |
541
+ | cosine_precision@3 | 0.0508 |
542
+ | cosine_precision@5 | 0.057 |
543
+ | cosine_precision@10 | 0.0488 |
544
+ | cosine_recall@1 | 0.0145 |
545
+ | cosine_recall@3 | 0.1525 |
546
+ | cosine_recall@5 | 0.2849 |
547
+ | cosine_recall@10 | 0.4882 |
548
+ | cosine_ndcg@10 | 0.221 |
549
+ | cosine_mrr@10 | 0.1395 |
550
+ | **cosine_map@100** | **0.1583** |
551
+
552
+ <!--
553
+ ## Bias, Risks and Limitations
554
+
555
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
556
+ -->
557
+
558
+ <!--
559
+ ### Recommendations
560
+
561
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
562
+ -->
563
+
564
+ ## Training Details
565
+
566
+ ### Training Dataset
567
+
568
+ #### Unnamed Dataset
569
+
570
+
571
+ * Size: 4,957 training samples
572
+ * Columns: <code>positive</code> and <code>anchor</code>
573
+ * Approximate statistics based on the first 1000 samples:
574
+ | | positive | anchor |
575
+ |:--------|:-------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
576
+ | type | string | string |
577
+ | details | <ul><li>min: 5 tokens</li><li>mean: 158.11 tokens</li><li>max: 1024 tokens</li></ul> | <ul><li>min: 10 tokens</li><li>mean: 21.11 tokens</li><li>max: 47 tokens</li></ul> |
578
+ * Samples:
579
+ | positive | anchor |
580
+ |:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------|
581
+ | <code>134 Eine Rückzahlungsverpflichtung besteht nicht für den Teil der Zulagen, der auf nach § 1 Abs. 1 Nr. 2 AltZertG angespartes gefördertes Altersvorsorgevermögen entfällt, wenn es in Form einer Hinterbliebenenrente an die dort genannten Hinterbliebenen ausgezahlt wird. Dies gilt auch für den entsprechenden Teil der Steuerermäßigung.</code> | <code>Muss man Zulagen zurückzahlen, wenn das Altersvorsorgevermögen als Hinterbliebenenrente ausgezahlt wird?</code> |
582
+ | <code>140 Beendet der Zulageberechtigte vor der vollständigen Rückzahlung des AltersvorsorgeEigenheimbetrags die Nutzung zu eigenen Wohnzwecken, wird er so behandelt, als habe er den noch nicht zurückgezahlten Betrag schädlich verwendet. Die auf den noch ausstehenden Rückzahlungsbetrag entfallenden Zulagen sowie die nach § 10a Abs. 4 EStG gesondert festgestellten Steuerermäßigungen sind zurückzuzahlen (§ 92a Abs. 3 EStG). Die im noch ausstehenden Rückzahlungsbetrag enthaltenen Zuwächse (z.B. Zinserträge und Kursgewinne) Seite 41 sind als sonstige Einkünfte zu versteuern (§ 22 Nr. 5 Satz 5 Halbsatz 1 EStG). Außerdem hat der Zulageberechtigte den Vorteil zu versteuern, der sich aus der zinslosen Nutzung des noch nicht zurückgezahlten Betrags ergibt. Zugrunde gelegt wird hierbei eine Verzinsung von 5 % (Zins und Zinseszins) für jedes volle Kalenderjahr der Nutzung (§ 22 Nr. 5 Satz 5 Halbsatz 2 EStG). Diese Folgen treten nicht ein, wenn er den noch nicht zurückgezahlten Betrag in ein Folgeobjekt investiert (§ 92a Abs. 4 Satz 3 Nr. 1 EStG) oder zugunsten eines auf seinen Namen lautenden zertifizierten Altersvorsorgevertrags einzahlt (§ 92a Abs. 4 Satz 3 Nr. 2 EStG).</code> | <code>Was geschieht steuerlich, wenn der AltersvorsorgeEigenheimbetrag nicht vollständig zurückgezahlt wird und die Immobilie nicht mehr selbst genutzt wird?</code> |
583
+ | <code>144 Die als Einkünfte nach § 22 Nr. 5 Satz 3 EStG i.V.m. § 22 Nr. 5 Satz 2 EStG zu besteuernden Beträge muss der Anbieter gem. § 94 Abs. 1 Satz 4 EStG dem Zulageberechtigten bescheinigen und im Wege des Rentenbezugsmitteilungsverfahrens (§ 22a EStG) mitteilen. Ergeben sich insoweit steuerpflichtige Einkünfte nach § 22 Nr. 5 Satz 3 EStG für einen anderen Leistungsempfänger (z. B. Erben), ist für diesen eine entsprechende Rentenbezugsmitteilung der ZfA zu übermitteln.</code> | <code>Was muss im Falle eines anderen Leistungsempfängers, wie Erben, hinsichtlich der Rentenbezugsmitteilung getan werden?</code> |
584
+ * Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
585
+ ```json
586
+ {
587
+ "loss": "MultipleNegativesRankingLoss",
588
+ "matryoshka_dims": [
589
+ 768,
590
+ 512,
591
+ 256,
592
+ 128,
593
+ 64
594
+ ],
595
+ "matryoshka_weights": [
596
+ 1,
597
+ 1,
598
+ 1,
599
+ 1,
600
+ 1
601
+ ],
602
+ "n_dims_per_step": -1
603
+ }
604
+ ```
605
+
606
+ ### Training Hyperparameters
607
+ #### Non-Default Hyperparameters
608
+
609
+ - `eval_strategy`: epoch
610
+ - `per_device_train_batch_size`: 16
611
+ - `per_device_eval_batch_size`: 16
612
+ - `gradient_accumulation_steps`: 16
613
+ - `learning_rate`: 2e-05
614
+ - `num_train_epochs`: 10
615
+ - `lr_scheduler_type`: cosine
616
+ - `warmup_ratio`: 0.1
617
+ - `bf16`: True
618
+ - `tf32`: True
619
+ - `load_best_model_at_end`: True
620
+ - `optim`: adamw_torch_fused
621
+ - `batch_sampler`: no_duplicates
622
+
623
+ #### All Hyperparameters
624
+ <details><summary>Click to expand</summary>
625
+
626
+ - `overwrite_output_dir`: False
627
+ - `do_predict`: False
628
+ - `eval_strategy`: epoch
629
+ - `prediction_loss_only`: True
630
+ - `per_device_train_batch_size`: 16
631
+ - `per_device_eval_batch_size`: 16
632
+ - `per_gpu_train_batch_size`: None
633
+ - `per_gpu_eval_batch_size`: None
634
+ - `gradient_accumulation_steps`: 16
635
+ - `eval_accumulation_steps`: None
636
+ - `learning_rate`: 2e-05
637
+ - `weight_decay`: 0.0
638
+ - `adam_beta1`: 0.9
639
+ - `adam_beta2`: 0.999
640
+ - `adam_epsilon`: 1e-08
641
+ - `max_grad_norm`: 1.0
642
+ - `num_train_epochs`: 10
643
+ - `max_steps`: -1
644
+ - `lr_scheduler_type`: cosine
645
+ - `lr_scheduler_kwargs`: {}
646
+ - `warmup_ratio`: 0.1
647
+ - `warmup_steps`: 0
648
+ - `log_level`: passive
649
+ - `log_level_replica`: warning
650
+ - `log_on_each_node`: True
651
+ - `logging_nan_inf_filter`: True
652
+ - `save_safetensors`: True
653
+ - `save_on_each_node`: False
654
+ - `save_only_model`: False
655
+ - `restore_callback_states_from_checkpoint`: False
656
+ - `no_cuda`: False
657
+ - `use_cpu`: False
658
+ - `use_mps_device`: False
659
+ - `seed`: 42
660
+ - `data_seed`: None
661
+ - `jit_mode_eval`: False
662
+ - `use_ipex`: False
663
+ - `bf16`: True
664
+ - `fp16`: False
665
+ - `fp16_opt_level`: O1
666
+ - `half_precision_backend`: auto
667
+ - `bf16_full_eval`: False
668
+ - `fp16_full_eval`: False
669
+ - `tf32`: True
670
+ - `local_rank`: 0
671
+ - `ddp_backend`: None
672
+ - `tpu_num_cores`: None
673
+ - `tpu_metrics_debug`: False
674
+ - `debug`: []
675
+ - `dataloader_drop_last`: False
676
+ - `dataloader_num_workers`: 0
677
+ - `dataloader_prefetch_factor`: None
678
+ - `past_index`: -1
679
+ - `disable_tqdm`: False
680
+ - `remove_unused_columns`: True
681
+ - `label_names`: None
682
+ - `load_best_model_at_end`: True
683
+ - `ignore_data_skip`: False
684
+ - `fsdp`: []
685
+ - `fsdp_min_num_params`: 0
686
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
687
+ - `fsdp_transformer_layer_cls_to_wrap`: None
688
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
689
+ - `deepspeed`: None
690
+ - `label_smoothing_factor`: 0.0
691
+ - `optim`: adamw_torch_fused
692
+ - `optim_args`: None
693
+ - `adafactor`: False
694
+ - `group_by_length`: False
695
+ - `length_column_name`: length
696
+ - `ddp_find_unused_parameters`: None
697
+ - `ddp_bucket_cap_mb`: None
698
+ - `ddp_broadcast_buffers`: False
699
+ - `dataloader_pin_memory`: True
700
+ - `dataloader_persistent_workers`: False
701
+ - `skip_memory_metrics`: True
702
+ - `use_legacy_prediction_loop`: False
703
+ - `push_to_hub`: False
704
+ - `resume_from_checkpoint`: None
705
+ - `hub_model_id`: None
706
+ - `hub_strategy`: every_save
707
+ - `hub_private_repo`: False
708
+ - `hub_always_push`: False
709
+ - `gradient_checkpointing`: False
710
+ - `gradient_checkpointing_kwargs`: None
711
+ - `include_inputs_for_metrics`: False
712
+ - `eval_do_concat_batches`: True
713
+ - `fp16_backend`: auto
714
+ - `push_to_hub_model_id`: None
715
+ - `push_to_hub_organization`: None
716
+ - `mp_parameters`:
717
+ - `auto_find_batch_size`: False
718
+ - `full_determinism`: False
719
+ - `torchdynamo`: None
720
+ - `ray_scope`: last
721
+ - `ddp_timeout`: 1800
722
+ - `torch_compile`: False
723
+ - `torch_compile_backend`: None
724
+ - `torch_compile_mode`: None
725
+ - `dispatch_batches`: None
726
+ - `split_batches`: None
727
+ - `include_tokens_per_second`: False
728
+ - `include_num_input_tokens_seen`: False
729
+ - `neftune_noise_alpha`: None
730
+ - `optim_target_modules`: None
731
+ - `batch_eval_metrics`: False
732
+ - `batch_sampler`: no_duplicates
733
+ - `multi_dataset_batch_sampler`: proportional
734
+
735
+ </details>
736
+
737
+ ### Training Logs
738
+ | 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 |
739
+ |:----------:|:-------:|:-------------:|:----------------------:|:----------------------:|:----------------------:|:---------------------:|:----------------------:|
740
+ | 0.5161 | 10 | 8.2406 | - | - | - | - | - |
741
+ | 0.9806 | 19 | - | 0.1125 | 0.1196 | 0.1231 | 0.0951 | 0.1231 |
742
+ | 1.0323 | 20 | 5.0545 | - | - | - | - | - |
743
+ | 1.5484 | 30 | 3.253 | - | - | - | - | - |
744
+ | 1.9613 | 38 | - | 0.1388 | 0.1423 | 0.1462 | 0.1282 | 0.1496 |
745
+ | 2.0645 | 40 | 2.3708 | - | - | - | - | - |
746
+ | 2.5806 | 50 | 1.7379 | - | - | - | - | - |
747
+ | 2.9935 | 58 | - | 0.1536 | 0.1611 | 0.1703 | 0.1409 | 0.1688 |
748
+ | 3.0968 | 60 | 1.3531 | - | - | - | - | - |
749
+ | 3.6129 | 70 | 1.1393 | - | - | - | - | - |
750
+ | 3.9742 | 77 | - | 0.1580 | 0.1667 | 0.1753 | 0.1515 | 0.1743 |
751
+ | 4.1290 | 80 | 0.8556 | - | - | - | - | - |
752
+ | 4.6452 | 90 | 0.8594 | - | - | - | - | - |
753
+ | 4.9548 | 96 | - | 0.1668 | 0.1718 | 0.1736 | 0.1588 | 0.1739 |
754
+ | 5.1613 | 100 | 0.6492 | - | - | - | - | - |
755
+ | 5.6774 | 110 | 0.6018 | - | - | - | - | - |
756
+ | 5.9871 | 116 | - | 0.1610 | 0.1714 | 0.1680 | 0.1569 | 0.1739 |
757
+ | 6.1935 | 120 | 0.4951 | - | - | - | - | - |
758
+ | 6.7097 | 130 | 0.4958 | - | - | - | - | - |
759
+ | **6.9677** | **135** | **-** | **0.1684** | **0.1742** | **0.1792** | **0.1616** | **0.1764** |
760
+ | 7.2258 | 140 | 0.4286 | - | - | - | - | - |
761
+ | 7.7419 | 150 | 0.4297 | - | - | - | - | - |
762
+ | 8.0 | 155 | - | 0.1647 | 0.1746 | 0.1777 | 0.1582 | 0.1772 |
763
+ | 8.2581 | 160 | 0.3508 | - | - | - | - | - |
764
+ | 8.7742 | 170 | 0.3937 | - | - | - | - | - |
765
+ | 8.9806 | 174 | - | 0.1652 | 0.1714 | 0.1780 | 0.1595 | 0.1743 |
766
+ | 9.2903 | 180 | 0.3621 | - | - | - | - | - |
767
+ | 9.8065 | 190 | 0.3503 | 0.1662 | 0.1742 | 0.1762 | 0.1583 | 0.1785 |
768
+
769
+ * The bold row denotes the saved checkpoint.
770
+
771
+ ### Framework Versions
772
+ - Python: 3.11.4
773
+ - Sentence Transformers: 3.0.1
774
+ - Transformers: 4.41.2
775
+ - PyTorch: 2.1.2+cu121
776
+ - Accelerate: 0.33.0
777
+ - Datasets: 2.19.1
778
+ - Tokenizers: 0.19.1
779
+
780
+ ## Citation
781
+
782
+ ### BibTeX
783
+
784
+ #### Sentence Transformers
785
+ ```bibtex
786
+ @inproceedings{reimers-2019-sentence-bert,
787
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
788
+ author = "Reimers, Nils and Gurevych, Iryna",
789
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
790
+ month = "11",
791
+ year = "2019",
792
+ publisher = "Association for Computational Linguistics",
793
+ url = "https://arxiv.org/abs/1908.10084",
794
+ }
795
+ ```
796
+
797
+ #### MatryoshkaLoss
798
+ ```bibtex
799
+ @misc{kusupati2024matryoshka,
800
+ title={Matryoshka Representation Learning},
801
+ 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},
802
+ year={2024},
803
+ eprint={2205.13147},
804
+ archivePrefix={arXiv},
805
+ primaryClass={cs.LG}
806
+ }
807
+ ```
808
+
809
+ #### MultipleNegativesRankingLoss
810
+ ```bibtex
811
+ @misc{henderson2017efficient,
812
+ title={Efficient Natural Language Response Suggestion for Smart Reply},
813
+ 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},
814
+ year={2017},
815
+ eprint={1705.00652},
816
+ archivePrefix={arXiv},
817
+ primaryClass={cs.CL}
818
+ }
819
+ ```
820
+
821
+ <!--
822
+ ## Glossary
823
+
824
+ *Clearly define terms in order to be accessible across audiences.*
825
+ -->
826
+
827
+ <!--
828
+ ## Model Card Authors
829
+
830
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
831
+ -->
832
+
833
+ <!--
834
+ ## Model Card Contact
835
+
836
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
837
+ -->
config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "aari1995/German_Semantic_V3",
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "auto_map": {
8
+ "AutoConfig": "aari1995/German_Semantic_V3--configuration_bert.JinaBertConfig",
9
+ "AutoModel": "aari1995/German_Semantic_V3--modeling_bert.JinaBertModel",
10
+ "AutoModelForMaskedLM": "aari1995/German_Semantic_V3--modeling_bert.JinaBertForMaskedLM",
11
+ "AutoModelForSequenceClassification": "aari1995/German_Semantic_V3--modeling_bert.JinaBertForSequenceClassification"
12
+ },
13
+ "classifier_dropout": null,
14
+ "emb_pooler": null,
15
+ "feed_forward_type": "original",
16
+ "hidden_act": "gelu",
17
+ "hidden_dropout_prob": 0.1,
18
+ "hidden_size": 1024,
19
+ "initializer_range": 0.02,
20
+ "intermediate_size": 4096,
21
+ "layer_norm_eps": 1e-12,
22
+ "max_position_embeddings": 8192,
23
+ "model_type": "bert",
24
+ "num_attention_heads": 16,
25
+ "num_hidden_layers": 24,
26
+ "pad_token_id": 0,
27
+ "position_embedding_type": "alibi",
28
+ "torch_dtype": "float32",
29
+ "transformers_version": "4.41.2",
30
+ "type_vocab_size": 2,
31
+ "use_cache": true,
32
+ "vocab_size": 31102
33
+ }
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:f5da74e19d1bd011d06b0914329426324ba95592a294fc198416b29adc5fd5bb
3
+ size 1374445424
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": 1024,
3
+ "do_lower_case": false
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 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "101": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "102": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "103": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "104": {
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": false,
48
+ "mask_token": "[MASK]",
49
+ "max_len": 9999999999,
50
+ "max_length": 1000,
51
+ "model_max_length": 8192,
52
+ "never_split": null,
53
+ "pad_to_multiple_of": null,
54
+ "pad_token": "[PAD]",
55
+ "pad_token_type_id": 0,
56
+ "padding_side": "right",
57
+ "sep_token": "[SEP]",
58
+ "stride": 0,
59
+ "strip_accents": false,
60
+ "tokenize_chinese_chars": true,
61
+ "tokenizer_class": "BertTokenizer",
62
+ "truncation_side": "right",
63
+ "truncation_strategy": "longest_first",
64
+ "unk_token": "[UNK]"
65
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff