Hugging Face's logo

haophancs
/
bge-m3-financial-matryoshka

Sentence Similarity
sentence-transformers
ONNX
Safetensors
English
xlm-roberta
feature-extraction
Generated from Trainer
dataset_size:6300
loss:MatryoshkaLoss
loss:MultipleNegativesRankingLoss
Eval Results
text-embeddings-inference
Inference Endpoints
Model card Files Community
haophancs commited on
Commit
adf982e
•
1 Parent(s): 3610f00

Add new SentenceTransformer model.

Browse files
Files changed (11) hide show
  1. .gitattributes +1 -0
  2. 1_Pooling/config.json +10 -0
  3. README.md +1046 -0
  4. config.json +28 -0
  5. config_sentence_transformers.json +10 -0
  6. model.safetensors +3 -0
  7. modules.json +20 -0
  8. sentence_bert_config.json +4 -0
  9. special_tokens_map.json +51 -0
  10. tokenizer.json +3 -0
  11. tokenizer_config.json +55 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 1024,
3
+ "pooling_mode_cls_token": 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 +1,1046 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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:6300
12
+ - loss:MatryoshkaLoss
13
+ - loss:MultipleNegativesRankingLoss
14
+ base_model: BAAI/bge-m3
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: The consolidated financial statements and accompanying notes listed
34
+ in Part IV, Item 15(a)(1) of this Annual Report on Form 10-K.
35
+ sentences:
36
+ - How much total space does an average The Home Depot store encompass including
37
+ its garden area?
38
+ - What section of the Annual Report on Form 10-K contains the consolidated financial
39
+ statements and accompanying notes?
40
+ - What types of competitive factors does Garmin believe are important in its markets?
41
+ - source_sentence: Item 3. Legal Proceedings, which covers litigation and regulatory
42
+ matters, refers to Note 12 – Commitments and Contingencies for more detailed information
43
+ within the Consolidated Financial Statements.
44
+ sentences:
45
+ - What pages contain the Financial Statements and Supplementary Data in IBM’s 2023
46
+ Annual Report to Stockholders?
47
+ - In which note can further details on Legal Proceedings be found within the Consolidated
48
+ Financial Statements?
49
+ - What is the title of Item 8 in the document?
50
+ - source_sentence: Net Revenues for the Entertainment segment were $659.3 million
51
+ in 2023.
52
+ sentences:
53
+ - What were the net revenues for the Entertainment segment in 2023?
54
+ - How much net cash was provided by operating activities in 2023?
55
+ - What was the net income reported for the fiscal year ending in August 2023?
56
+ - source_sentence: 'The capital allocation program focuses on three objectives: (1)
57
+ grow our business at an average target ROIC-adjusted rate of 20% or greater; (2)
58
+ maintain a strong investment-grade balance sheet, including a target average automotive
59
+ cash balance of $18.0 billion; and (3) after the first two objectives are met,
60
+ return available cash to shareholders.'
61
+ sentences:
62
+ - Why is ICE Mortgage Technology subject to the examination by the Federal Financial
63
+ Institutions Examination Council (FFIEC) and its member agencies?
64
+ - What type of regulations do U.S. automobiles need to comply with under the National
65
+ Highway Traffic Safety Administration?
66
+ - What are the three objectives of the capital allocation program referenced?
67
+ - source_sentence: As of January 28, 2024 the net carrying value of our inventories
68
+ was $1.3 billion, which included provisions for obsolete and damaged inventory
69
+ of $139.7 million.
70
+ sentences:
71
+ - What is the status of the company's inventory as of January 28, 2024, in terms
72
+ of its valuation and provisions for obsolescence?
73
+ - What is the relationship between the ESG goals and the long-term growth strategy?
74
+ - What were the financial impacts of Ford's investments in Rivian and Argo in the
75
+ year 2022?
76
+ pipeline_tag: sentence-similarity
77
+ model-index:
78
+ - name: BGE-M3 Financial Matryoshka
79
+ results:
80
+ - task:
81
+ type: information-retrieval
82
+ name: Information Retrieval
83
+ dataset:
84
+ name: dim 1024
85
+ type: dim_1024
86
+ metrics:
87
+ - type: cosine_accuracy@1
88
+ value: 0.7171428571428572
89
+ name: Cosine Accuracy@1
90
+ - type: cosine_accuracy@3
91
+ value: 0.8314285714285714
92
+ name: Cosine Accuracy@3
93
+ - type: cosine_accuracy@5
94
+ value: 0.87
95
+ name: Cosine Accuracy@5
96
+ - type: cosine_accuracy@10
97
+ value: 0.9142857142857143
98
+ name: Cosine Accuracy@10
99
+ - type: cosine_precision@1
100
+ value: 0.7171428571428572
101
+ name: Cosine Precision@1
102
+ - type: cosine_precision@3
103
+ value: 0.27714285714285714
104
+ name: Cosine Precision@3
105
+ - type: cosine_precision@5
106
+ value: 0.174
107
+ name: Cosine Precision@5
108
+ - type: cosine_precision@10
109
+ value: 0.09142857142857141
110
+ name: Cosine Precision@10
111
+ - type: cosine_recall@1
112
+ value: 0.7171428571428572
113
+ name: Cosine Recall@1
114
+ - type: cosine_recall@3
115
+ value: 0.8314285714285714
116
+ name: Cosine Recall@3
117
+ - type: cosine_recall@5
118
+ value: 0.87
119
+ name: Cosine Recall@5
120
+ - type: cosine_recall@10
121
+ value: 0.9142857142857143
122
+ name: Cosine Recall@10
123
+ - type: cosine_ndcg@10
124
+ value: 0.8152097277196483
125
+ name: Cosine Ndcg@10
126
+ - type: cosine_mrr@10
127
+ value: 0.7835873015873015
128
+ name: Cosine Mrr@10
129
+ - type: cosine_map@100
130
+ value: 0.7867088346410263
131
+ name: Cosine Map@100
132
+ - task:
133
+ type: information-retrieval
134
+ name: Information Retrieval
135
+ dataset:
136
+ name: dim 768
137
+ type: dim_768
138
+ metrics:
139
+ - type: cosine_accuracy@1
140
+ value: 0.7128571428571429
141
+ name: Cosine Accuracy@1
142
+ - type: cosine_accuracy@3
143
+ value: 0.8342857142857143
144
+ name: Cosine Accuracy@3
145
+ - type: cosine_accuracy@5
146
+ value: 0.8657142857142858
147
+ name: Cosine Accuracy@5
148
+ - type: cosine_accuracy@10
149
+ value: 0.91
150
+ name: Cosine Accuracy@10
151
+ - type: cosine_precision@1
152
+ value: 0.7128571428571429
153
+ name: Cosine Precision@1
154
+ - type: cosine_precision@3
155
+ value: 0.2780952380952381
156
+ name: Cosine Precision@3
157
+ - type: cosine_precision@5
158
+ value: 0.17314285714285713
159
+ name: Cosine Precision@5
160
+ - type: cosine_precision@10
161
+ value: 0.09099999999999998
162
+ name: Cosine Precision@10
163
+ - type: cosine_recall@1
164
+ value: 0.7128571428571429
165
+ name: Cosine Recall@1
166
+ - type: cosine_recall@3
167
+ value: 0.8342857142857143
168
+ name: Cosine Recall@3
169
+ - type: cosine_recall@5
170
+ value: 0.8657142857142858
171
+ name: Cosine Recall@5
172
+ - type: cosine_recall@10
173
+ value: 0.91
174
+ name: Cosine Recall@10
175
+ - type: cosine_ndcg@10
176
+ value: 0.8122143155463835
177
+ name: Cosine Ndcg@10
178
+ - type: cosine_mrr@10
179
+ value: 0.7808730158730155
180
+ name: Cosine Mrr@10
181
+ - type: cosine_map@100
182
+ value: 0.7843065190190194
183
+ name: Cosine Map@100
184
+ - task:
185
+ type: information-retrieval
186
+ name: Information Retrieval
187
+ dataset:
188
+ name: dim 512
189
+ type: dim_512
190
+ metrics:
191
+ - type: cosine_accuracy@1
192
+ value: 0.7114285714285714
193
+ name: Cosine Accuracy@1
194
+ - type: cosine_accuracy@3
195
+ value: 0.8357142857142857
196
+ name: Cosine Accuracy@3
197
+ - type: cosine_accuracy@5
198
+ value: 0.8642857142857143
199
+ name: Cosine Accuracy@5
200
+ - type: cosine_accuracy@10
201
+ value: 0.91
202
+ name: Cosine Accuracy@10
203
+ - type: cosine_precision@1
204
+ value: 0.7114285714285714
205
+ name: Cosine Precision@1
206
+ - type: cosine_precision@3
207
+ value: 0.2785714285714286
208
+ name: Cosine Precision@3
209
+ - type: cosine_precision@5
210
+ value: 0.17285714285714285
211
+ name: Cosine Precision@5
212
+ - type: cosine_precision@10
213
+ value: 0.09099999999999998
214
+ name: Cosine Precision@10
215
+ - type: cosine_recall@1
216
+ value: 0.7114285714285714
217
+ name: Cosine Recall@1
218
+ - type: cosine_recall@3
219
+ value: 0.8357142857142857
220
+ name: Cosine Recall@3
221
+ - type: cosine_recall@5
222
+ value: 0.8642857142857143
223
+ name: Cosine Recall@5
224
+ - type: cosine_recall@10
225
+ value: 0.91
226
+ name: Cosine Recall@10
227
+ - type: cosine_ndcg@10
228
+ value: 0.8109635546819154
229
+ name: Cosine Ndcg@10
230
+ - type: cosine_mrr@10
231
+ value: 0.7792959183673466
232
+ name: Cosine Mrr@10
233
+ - type: cosine_map@100
234
+ value: 0.782703758965192
235
+ name: Cosine Map@100
236
+ - task:
237
+ type: information-retrieval
238
+ name: Information Retrieval
239
+ dataset:
240
+ name: dim 384
241
+ type: dim_384
242
+ metrics:
243
+ - type: cosine_accuracy@1
244
+ value: 0.7142857142857143
245
+ name: Cosine Accuracy@1
246
+ - type: cosine_accuracy@3
247
+ value: 0.8328571428571429
248
+ name: Cosine Accuracy@3
249
+ - type: cosine_accuracy@5
250
+ value: 0.8628571428571429
251
+ name: Cosine Accuracy@5
252
+ - type: cosine_accuracy@10
253
+ value: 0.9128571428571428
254
+ name: Cosine Accuracy@10
255
+ - type: cosine_precision@1
256
+ value: 0.7142857142857143
257
+ name: Cosine Precision@1
258
+ - type: cosine_precision@3
259
+ value: 0.2776190476190476
260
+ name: Cosine Precision@3
261
+ - type: cosine_precision@5
262
+ value: 0.17257142857142854
263
+ name: Cosine Precision@5
264
+ - type: cosine_precision@10
265
+ value: 0.09128571428571428
266
+ name: Cosine Precision@10
267
+ - type: cosine_recall@1
268
+ value: 0.7142857142857143
269
+ name: Cosine Recall@1
270
+ - type: cosine_recall@3
271
+ value: 0.8328571428571429
272
+ name: Cosine Recall@3
273
+ - type: cosine_recall@5
274
+ value: 0.8628571428571429
275
+ name: Cosine Recall@5
276
+ - type: cosine_recall@10
277
+ value: 0.9128571428571428
278
+ name: Cosine Recall@10
279
+ - type: cosine_ndcg@10
280
+ value: 0.8125530857386527
281
+ name: Cosine Ndcg@10
282
+ - type: cosine_mrr@10
283
+ value: 0.7806292517006799
284
+ name: Cosine Mrr@10
285
+ - type: cosine_map@100
286
+ value: 0.7837508100457361
287
+ name: Cosine Map@100
288
+ ---
289
+
290
+ # BGE-M3 Financial Matryoshka
291
+
292
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [BAAI/bge-m3](https://huggingface.co/BAAI/bge-m3). 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.
293
+
294
+ ## Model Details
295
+
296
+ ### Model Description
297
+ - **Model Type:** Sentence Transformer
298
+ - **Base model:** [BAAI/bge-m3](https://huggingface.co/BAAI/bge-m3) <!-- at revision babcf60cae0a1f438d7ade582983d4ba462303c2 -->
299
+ - **Maximum Sequence Length:** 8192 tokens
300
+ - **Output Dimensionality:** 1024 tokens
301
+ - **Similarity Function:** Cosine Similarity
302
+ <!-- - **Training Dataset:** Unknown -->
303
+ - **Language:** en
304
+ - **License:** apache-2.0
305
+
306
+ ### Model Sources
307
+
308
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
309
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
310
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
311
+
312
+ ### Full Model Architecture
313
+
314
+ ```
315
+ SentenceTransformer(
316
+ (0): Transformer({'max_seq_length': 8192, 'do_lower_case': False}) with Transformer model: XLMRobertaModel
317
+ (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})
318
+ (2): Normalize()
319
+ )
320
+ ```
321
+
322
+ ## Usage
323
+
324
+ ### Direct Usage (Sentence Transformers)
325
+
326
+ First install the Sentence Transformers library:
327
+
328
+ ```bash
329
+ pip install -U sentence-transformers
330
+ ```
331
+
332
+ Then you can load this model and run inference.
333
+ ```python
334
+ from sentence_transformers import SentenceTransformer
335
+
336
+ # Download from the 🤗 Hub
337
+ model = SentenceTransformer("haophancs/bge-m3-financial-matryoshka")
338
+ # Run inference
339
+ sentences = [
340
+ 'As of January 28, 2024 the net carrying value of our inventories was $1.3 billion, which included provisions for obsolete and damaged inventory of $139.7 million.',
341
+ "What is the status of the company's inventory as of January 28, 2024, in terms of its valuation and provisions for obsolescence?",
342
+ 'What is the relationship between the ESG goals and the long-term growth strategy?',
343
+ ]
344
+ embeddings = model.encode(sentences)
345
+ print(embeddings.shape)
346
+ # [3, 1024]
347
+
348
+ # Get the similarity scores for the embeddings
349
+ similarities = model.similarity(embeddings, embeddings)
350
+ print(similarities.shape)
351
+ # [3, 3]
352
+ ```
353
+
354
+ <!--
355
+ ### Direct Usage (Transformers)
356
+
357
+ <details><summary>Click to see the direct usage in Transformers</summary>
358
+
359
+ </details>
360
+ -->
361
+
362
+ <!--
363
+ ### Downstream Usage (Sentence Transformers)
364
+
365
+ You can finetune this model on your own dataset.
366
+
367
+ <details><summary>Click to expand</summary>
368
+
369
+ </details>
370
+ -->
371
+
372
+ <!--
373
+ ### Out-of-Scope Use
374
+
375
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
376
+ -->
377
+
378
+ ## Evaluation
379
+
380
+ ### Metrics
381
+
382
+ #### Information Retrieval
383
+ * Dataset: `dim_1024`
384
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
385
+
386
+ | Metric | Value |
387
+ |:--------------------|:-----------|
388
+ | cosine_accuracy@1 | 0.7171 |
389
+ | cosine_accuracy@3 | 0.8314 |
390
+ | cosine_accuracy@5 | 0.87 |
391
+ | cosine_accuracy@10 | 0.9143 |
392
+ | cosine_precision@1 | 0.7171 |
393
+ | cosine_precision@3 | 0.2771 |
394
+ | cosine_precision@5 | 0.174 |
395
+ | cosine_precision@10 | 0.0914 |
396
+ | cosine_recall@1 | 0.7171 |
397
+ | cosine_recall@3 | 0.8314 |
398
+ | cosine_recall@5 | 0.87 |
399
+ | cosine_recall@10 | 0.9143 |
400
+ | cosine_ndcg@10 | 0.8152 |
401
+ | cosine_mrr@10 | 0.7836 |
402
+ | **cosine_map@100** | **0.7867** |
403
+
404
+ #### Information Retrieval
405
+ * Dataset: `dim_768`
406
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
407
+
408
+ | Metric | Value |
409
+ |:--------------------|:-----------|
410
+ | cosine_accuracy@1 | 0.7129 |
411
+ | cosine_accuracy@3 | 0.8343 |
412
+ | cosine_accuracy@5 | 0.8657 |
413
+ | cosine_accuracy@10 | 0.91 |
414
+ | cosine_precision@1 | 0.7129 |
415
+ | cosine_precision@3 | 0.2781 |
416
+ | cosine_precision@5 | 0.1731 |
417
+ | cosine_precision@10 | 0.091 |
418
+ | cosine_recall@1 | 0.7129 |
419
+ | cosine_recall@3 | 0.8343 |
420
+ | cosine_recall@5 | 0.8657 |
421
+ | cosine_recall@10 | 0.91 |
422
+ | cosine_ndcg@10 | 0.8122 |
423
+ | cosine_mrr@10 | 0.7809 |
424
+ | **cosine_map@100** | **0.7843** |
425
+
426
+ #### Information Retrieval
427
+ * Dataset: `dim_512`
428
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
429
+
430
+ | Metric | Value |
431
+ |:--------------------|:-----------|
432
+ | cosine_accuracy@1 | 0.7114 |
433
+ | cosine_accuracy@3 | 0.8357 |
434
+ | cosine_accuracy@5 | 0.8643 |
435
+ | cosine_accuracy@10 | 0.91 |
436
+ | cosine_precision@1 | 0.7114 |
437
+ | cosine_precision@3 | 0.2786 |
438
+ | cosine_precision@5 | 0.1729 |
439
+ | cosine_precision@10 | 0.091 |
440
+ | cosine_recall@1 | 0.7114 |
441
+ | cosine_recall@3 | 0.8357 |
442
+ | cosine_recall@5 | 0.8643 |
443
+ | cosine_recall@10 | 0.91 |
444
+ | cosine_ndcg@10 | 0.811 |
445
+ | cosine_mrr@10 | 0.7793 |
446
+ | **cosine_map@100** | **0.7827** |
447
+
448
+ #### Information Retrieval
449
+ * Dataset: `dim_384`
450
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
451
+
452
+ | Metric | Value |
453
+ |:--------------------|:-----------|
454
+ | cosine_accuracy@1 | 0.7143 |
455
+ | cosine_accuracy@3 | 0.8329 |
456
+ | cosine_accuracy@5 | 0.8629 |
457
+ | cosine_accuracy@10 | 0.9129 |
458
+ | cosine_precision@1 | 0.7143 |
459
+ | cosine_precision@3 | 0.2776 |
460
+ | cosine_precision@5 | 0.1726 |
461
+ | cosine_precision@10 | 0.0913 |
462
+ | cosine_recall@1 | 0.7143 |
463
+ | cosine_recall@3 | 0.8329 |
464
+ | cosine_recall@5 | 0.8629 |
465
+ | cosine_recall@10 | 0.9129 |
466
+ | cosine_ndcg@10 | 0.8126 |
467
+ | cosine_mrr@10 | 0.7806 |
468
+ | **cosine_map@100** | **0.7838** |
469
+
470
+ <!--
471
+ ## Bias, Risks and Limitations
472
+
473
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
474
+ -->
475
+
476
+ <!--
477
+ ### Recommendations
478
+
479
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
480
+ -->
481
+
482
+ ## Training Details
483
+
484
+ ### Training Dataset
485
+
486
+ #### Unnamed Dataset
487
+
488
+
489
+ * Size: 6,300 training samples
490
+ * Columns: <code>positive</code> and <code>anchor</code>
491
+ * Approximate statistics based on the first 1000 samples:
492
+ | | positive | anchor |
493
+ |:--------|:-------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
494
+ | type | string | string |
495
+ | details | <ul><li>min: 11 tokens</li><li>mean: 51.97 tokens</li><li>max: 1146 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 21.63 tokens</li><li>max: 47 tokens</li></ul> |
496
+ * Samples:
497
+ | positive | anchor |
498
+ |:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------|
499
+ | <code>From fiscal year 2022 to 2023, the cost of revenue as a percentage of total net revenue decreased by 3 percent.</code> | <code>What was the percentage change in cost of revenue as a percentage of total net revenue from fiscal year 2022 to 2023?</code> |
500
+ | <code> •Operating income increased $321 million, or 2%, to $18.1 billion versus year ago due to the increase in net sales, partially offset by a modest decrease in operating margin.</code> | <code>What factors contributed to the increase in operating income for Procter & Gamble in 2023?</code> |
501
+ | <code>market specific brands including 'Aurrera,' 'Lider,' and 'PhonePe.'</code> | <code>What specific brands does Walmart International market?</code> |
502
+ * Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
503
+ ```json
504
+ {
505
+ "loss": "MultipleNegativesRankingLoss",
506
+ "matryoshka_dims": [
507
+ 1024,
508
+ 768,
509
+ 512,
510
+ 384
511
+ ],
512
+ "matryoshka_weights": [
513
+ 1,
514
+ 1,
515
+ 1,
516
+ 1
517
+ ],
518
+ "n_dims_per_step": -1
519
+ }
520
+ ```
521
+
522
+ ### Training Hyperparameters
523
+ #### Non-Default Hyperparameters
524
+
525
+ - `eval_strategy`: epoch
526
+ - `per_device_train_batch_size`: 4
527
+ - `per_device_eval_batch_size`: 2
528
+ - `gradient_accumulation_steps`: 2
529
+ - `learning_rate`: 2e-05
530
+ - `num_train_epochs`: 4
531
+ - `lr_scheduler_type`: cosine
532
+ - `warmup_ratio`: 0.1
533
+ - `bf16`: True
534
+ - `tf32`: True
535
+ - `load_best_model_at_end`: True
536
+ - `optim`: adamw_torch_fused
537
+ - `batch_sampler`: no_duplicates
538
+
539
+ #### All Hyperparameters
540
+ <details><summary>Click to expand</summary>
541
+
542
+ - `overwrite_output_dir`: False
543
+ - `do_predict`: False
544
+ - `eval_strategy`: epoch
545
+ - `prediction_loss_only`: True
546
+ - `per_device_train_batch_size`: 4
547
+ - `per_device_eval_batch_size`: 2
548
+ - `per_gpu_train_batch_size`: None
549
+ - `per_gpu_eval_batch_size`: None
550
+ - `gradient_accumulation_steps`: 2
551
+ - `eval_accumulation_steps`: None
552
+ - `learning_rate`: 2e-05
553
+ - `weight_decay`: 0.0
554
+ - `adam_beta1`: 0.9
555
+ - `adam_beta2`: 0.999
556
+ - `adam_epsilon`: 1e-08
557
+ - `max_grad_norm`: 1.0
558
+ - `num_train_epochs`: 4
559
+ - `max_steps`: -1
560
+ - `lr_scheduler_type`: cosine
561
+ - `lr_scheduler_kwargs`: {}
562
+ - `warmup_ratio`: 0.1
563
+ - `warmup_steps`: 0
564
+ - `log_level`: passive
565
+ - `log_level_replica`: warning
566
+ - `log_on_each_node`: True
567
+ - `logging_nan_inf_filter`: True
568
+ - `save_safetensors`: True
569
+ - `save_on_each_node`: False
570
+ - `save_only_model`: False
571
+ - `restore_callback_states_from_checkpoint`: False
572
+ - `no_cuda`: False
573
+ - `use_cpu`: False
574
+ - `use_mps_device`: False
575
+ - `seed`: 42
576
+ - `data_seed`: None
577
+ - `jit_mode_eval`: False
578
+ - `use_ipex`: False
579
+ - `bf16`: True
580
+ - `fp16`: False
581
+ - `fp16_opt_level`: O1
582
+ - `half_precision_backend`: auto
583
+ - `bf16_full_eval`: False
584
+ - `fp16_full_eval`: False
585
+ - `tf32`: True
586
+ - `local_rank`: 0
587
+ - `ddp_backend`: None
588
+ - `tpu_num_cores`: None
589
+ - `tpu_metrics_debug`: False
590
+ - `debug`: []
591
+ - `dataloader_drop_last`: False
592
+ - `dataloader_num_workers`: 0
593
+ - `dataloader_prefetch_factor`: None
594
+ - `past_index`: -1
595
+ - `disable_tqdm`: False
596
+ - `remove_unused_columns`: True
597
+ - `label_names`: None
598
+ - `load_best_model_at_end`: True
599
+ - `ignore_data_skip`: False
600
+ - `fsdp`: []
601
+ - `fsdp_min_num_params`: 0
602
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
603
+ - `fsdp_transformer_layer_cls_to_wrap`: None
604
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
605
+ - `deepspeed`: None
606
+ - `label_smoothing_factor`: 0.0
607
+ - `optim`: adamw_torch_fused
608
+ - `optim_args`: None
609
+ - `adafactor`: False
610
+ - `group_by_length`: False
611
+ - `length_column_name`: length
612
+ - `ddp_find_unused_parameters`: None
613
+ - `ddp_bucket_cap_mb`: None
614
+ - `ddp_broadcast_buffers`: False
615
+ - `dataloader_pin_memory`: True
616
+ - `dataloader_persistent_workers`: False
617
+ - `skip_memory_metrics`: True
618
+ - `use_legacy_prediction_loop`: False
619
+ - `push_to_hub`: False
620
+ - `resume_from_checkpoint`: None
621
+ - `hub_model_id`: None
622
+ - `hub_strategy`: every_save
623
+ - `hub_private_repo`: False
624
+ - `hub_always_push`: False
625
+ - `gradient_checkpointing`: False
626
+ - `gradient_checkpointing_kwargs`: None
627
+ - `include_inputs_for_metrics`: False
628
+ - `eval_do_concat_batches`: True
629
+ - `fp16_backend`: auto
630
+ - `push_to_hub_model_id`: None
631
+ - `push_to_hub_organization`: None
632
+ - `mp_parameters`:
633
+ - `auto_find_batch_size`: False
634
+ - `full_determinism`: False
635
+ - `torchdynamo`: None
636
+ - `ray_scope`: last
637
+ - `ddp_timeout`: 1800
638
+ - `torch_compile`: False
639
+ - `torch_compile_backend`: None
640
+ - `torch_compile_mode`: None
641
+ - `dispatch_batches`: None
642
+ - `split_batches`: None
643
+ - `include_tokens_per_second`: False
644
+ - `include_num_input_tokens_seen`: False
645
+ - `neftune_noise_alpha`: None
646
+ - `optim_target_modules`: None
647
+ - `batch_eval_metrics`: False
648
+ - `batch_sampler`: no_duplicates
649
+ - `multi_dataset_batch_sampler`: proportional
650
+
651
+ </details>
652
+
653
+ ### Training Logs
654
+ <details><summary>Click to expand</summary>
655
+
656
+ | Epoch | Step | Training Loss | dim_1024_cosine_map@100 | dim_384_cosine_map@100 | dim_512_cosine_map@100 | dim_768_cosine_map@100 |
657
+ |:----------:|:--------:|:-------------:|:-----------------------:|:----------------------:|:----------------------:|:----------------------:|
658
+ | 0.0127 | 10 | 0.2059 | - | - | - | - |
659
+ | 0.0254 | 20 | 0.2612 | - | - | - | - |
660
+ | 0.0381 | 30 | 0.0873 | - | - | - | - |
661
+ | 0.0508 | 40 | 0.1352 | - | - | - | - |
662
+ | 0.0635 | 50 | 0.156 | - | - | - | - |
663
+ | 0.0762 | 60 | 0.0407 | - | - | - | - |
664
+ | 0.0889 | 70 | 0.09 | - | - | - | - |
665
+ | 0.1016 | 80 | 0.027 | - | - | - | - |
666
+ | 0.1143 | 90 | 0.0978 | - | - | - | - |
667
+ | 0.1270 | 100 | 0.0105 | - | - | - | - |
668
+ | 0.1397 | 110 | 0.0402 | - | - | - | - |
669
+ | 0.1524 | 120 | 0.0745 | - | - | - | - |
670
+ | 0.1651 | 130 | 0.0655 | - | - | - | - |
671
+ | 0.1778 | 140 | 0.0075 | - | - | - | - |
672
+ | 0.1905 | 150 | 0.0141 | - | - | - | - |
673
+ | 0.2032 | 160 | 0.0615 | - | - | - | - |
674
+ | 0.2159 | 170 | 0.0029 | - | - | - | - |
675
+ | 0.2286 | 180 | 0.0269 | - | - | - | - |
676
+ | 0.2413 | 190 | 0.0724 | - | - | - | - |
677
+ | 0.2540 | 200 | 0.0218 | - | - | - | - |
678
+ | 0.2667 | 210 | 0.0027 | - | - | - | - |
679
+ | 0.2794 | 220 | 0.007 | - | - | - | - |
680
+ | 0.2921 | 230 | 0.0814 | - | - | - | - |
681
+ | 0.3048 | 240 | 0.0326 | - | - | - | - |
682
+ | 0.3175 | 250 | 0.0061 | - | - | - | - |
683
+ | 0.3302 | 260 | 0.0471 | - | - | - | - |
684
+ | 0.3429 | 270 | 0.0115 | - | - | - | - |
685
+ | 0.3556 | 280 | 0.0021 | - | - | - | - |
686
+ | 0.3683 | 290 | 0.0975 | - | - | - | - |
687
+ | 0.3810 | 300 | 0.0572 | - | - | - | - |
688
+ | 0.3937 | 310 | 0.0125 | - | - | - | - |
689
+ | 0.4063 | 320 | 0.04 | - | - | - | - |
690
+ | 0.4190 | 330 | 0.0023 | - | - | - | - |
691
+ | 0.4317 | 340 | 0.0121 | - | - | - | - |
692
+ | 0.4444 | 350 | 0.0116 | - | - | - | - |
693
+ | 0.4571 | 360 | 0.0059 | - | - | - | - |
694
+ | 0.4698 | 370 | 0.0217 | - | - | - | - |
695
+ | 0.4825 | 380 | 0.0294 | - | - | - | - |
696
+ | 0.4952 | 390 | 0.1102 | - | - | - | - |
697
+ | 0.5079 | 400 | 0.0103 | - | - | - | - |
698
+ | 0.5206 | 410 | 0.0023 | - | - | - | - |
699
+ | 0.5333 | 420 | 0.0157 | - | - | - | - |
700
+ | 0.5460 | 430 | 0.0805 | - | - | - | - |
701
+ | 0.5587 | 440 | 0.0168 | - | - | - | - |
702
+ | 0.5714 | 450 | 0.1279 | - | - | - | - |
703
+ | 0.5841 | 460 | 0.2012 | - | - | - | - |
704
+ | 0.5968 | 470 | 0.0436 | - | - | - | - |
705
+ | 0.6095 | 480 | 0.0204 | - | - | - | - |
706
+ | 0.6222 | 490 | 0.0097 | - | - | - | - |
707
+ | 0.6349 | 500 | 0.0013 | - | - | - | - |
708
+ | 0.6476 | 510 | 0.0042 | - | - | - | - |
709
+ | 0.6603 | 520 | 0.0034 | - | - | - | - |
710
+ | 0.6730 | 530 | 0.0226 | - | - | - | - |
711
+ | 0.6857 | 540 | 0.0267 | - | - | - | - |
712
+ | 0.6984 | 550 | 0.0007 | - | - | - | - |
713
+ | 0.7111 | 560 | 0.0766 | - | - | - | - |
714
+ | 0.7238 | 570 | 0.2174 | - | - | - | - |
715
+ | 0.7365 | 580 | 0.0089 | - | - | - | - |
716
+ | 0.7492 | 590 | 0.0794 | - | - | - | - |
717
+ | 0.7619 | 600 | 0.0031 | - | - | - | - |
718
+ | 0.7746 | 610 | 0.0499 | - | - | - | - |
719
+ | 0.7873 | 620 | 0.0105 | - | - | - | - |
720
+ | 0.8 | 630 | 0.0097 | - | - | - | - |
721
+ | 0.8127 | 640 | 0.0028 | - | - | - | - |
722
+ | 0.8254 | 650 | 0.0029 | - | - | - | - |
723
+ | 0.8381 | 660 | 0.1811 | - | - | - | - |
724
+ | 0.8508 | 670 | 0.064 | - | - | - | - |
725
+ | 0.8635 | 680 | 0.0139 | - | - | - | - |
726
+ | 0.8762 | 690 | 0.055 | - | - | - | - |
727
+ | 0.8889 | 700 | 0.0013 | - | - | - | - |
728
+ | 0.9016 | 710 | 0.0402 | - | - | - | - |
729
+ | 0.9143 | 720 | 0.0824 | - | - | - | - |
730
+ | 0.9270 | 730 | 0.03 | - | - | - | - |
731
+ | 0.9397 | 740 | 0.0337 | - | - | - | - |
732
+ | 0.9524 | 750 | 0.1192 | - | - | - | - |
733
+ | 0.9651 | 760 | 0.0039 | - | - | - | - |
734
+ | 0.9778 | 770 | 0.004 | - | - | - | - |
735
+ | 0.9905 | 780 | 0.1413 | - | - | - | - |
736
+ | 0.9994 | 787 | - | 0.7851 | 0.7794 | 0.7822 | 0.7863 |
737
+ | 1.0032 | 790 | 0.019 | - | - | - | - |
738
+ | 1.0159 | 800 | 0.0587 | - | - | - | - |
739
+ | 1.0286 | 810 | 0.0186 | - | - | - | - |
740
+ | 1.0413 | 820 | 0.0018 | - | - | - | - |
741
+ | 1.0540 | 830 | 0.0631 | - | - | - | - |
742
+ | 1.0667 | 840 | 0.0127 | - | - | - | - |
743
+ | 1.0794 | 850 | 0.0037 | - | - | - | - |
744
+ | 1.0921 | 860 | 0.0029 | - | - | - | - |
745
+ | 1.1048 | 870 | 0.1437 | - | - | - | - |
746
+ | 1.1175 | 880 | 0.0015 | - | - | - | - |
747
+ | 1.1302 | 890 | 0.0024 | - | - | - | - |
748
+ | 1.1429 | 900 | 0.0133 | - | - | - | - |
749
+ | 1.1556 | 910 | 0.0245 | - | - | - | - |
750
+ | 1.1683 | 920 | 0.0017 | - | - | - | - |
751
+ | 1.1810 | 930 | 0.0007 | - | - | - | - |
752
+ | 1.1937 | 940 | 0.002 | - | - | - | - |
753
+ | 1.2063 | 950 | 0.0044 | - | - | - | - |
754
+ | 1.2190 | 960 | 0.0009 | - | - | - | - |
755
+ | 1.2317 | 970 | 0.01 | - | - | - | - |
756
+ | 1.2444 | 980 | 0.0026 | - | - | - | - |
757
+ | 1.2571 | 990 | 0.0017 | - | - | - | - |
758
+ | 1.2698 | 1000 | 0.0014 | - | - | - | - |
759
+ | 1.2825 | 1010 | 0.0009 | - | - | - | - |
760
+ | 1.2952 | 1020 | 0.0829 | - | - | - | - |
761
+ | 1.3079 | 1030 | 0.0011 | - | - | - | - |
762
+ | 1.3206 | 1040 | 0.012 | - | - | - | - |
763
+ | 1.3333 | 1050 | 0.0019 | - | - | - | - |
764
+ | 1.3460 | 1060 | 0.0007 | - | - | - | - |
765
+ | 1.3587 | 1070 | 0.0141 | - | - | - | - |
766
+ | 1.3714 | 1080 | 0.0003 | - | - | - | - |
767
+ | 1.3841 | 1090 | 0.001 | - | - | - | - |
768
+ | 1.3968 | 1100 | 0.0005 | - | - | - | - |
769
+ | 1.4095 | 1110 | 0.0031 | - | - | - | - |
770
+ | 1.4222 | 1120 | 0.0004 | - | - | - | - |
771
+ | 1.4349 | 1130 | 0.0054 | - | - | - | - |
772
+ | 1.4476 | 1140 | 0.0003 | - | - | - | - |
773
+ | 1.4603 | 1150 | 0.0007 | - | - | - | - |
774
+ | 1.4730 | 1160 | 0.0009 | - | - | - | - |
775
+ | 1.4857 | 1170 | 0.001 | - | - | - | - |
776
+ | 1.4984 | 1180 | 0.0006 | - | - | - | - |
777
+ | 1.5111 | 1190 | 0.0046 | - | - | - | - |
778
+ | 1.5238 | 1200 | 0.0003 | - | - | - | - |
779
+ | 1.5365 | 1210 | 0.0002 | - | - | - | - |
780
+ | 1.5492 | 1220 | 0.004 | - | - | - | - |
781
+ | 1.5619 | 1230 | 0.0017 | - | - | - | - |
782
+ | 1.5746 | 1240 | 0.0003 | - | - | - | - |
783
+ | 1.5873 | 1250 | 0.0027 | - | - | - | - |
784
+ | 1.6 | 1260 | 0.1134 | - | - | - | - |
785
+ | 1.6127 | 1270 | 0.0007 | - | - | - | - |
786
+ | 1.6254 | 1280 | 0.0005 | - | - | - | - |
787
+ | 1.6381 | 1290 | 0.0008 | - | - | - | - |
788
+ | 1.6508 | 1300 | 0.0001 | - | - | - | - |
789
+ | 1.6635 | 1310 | 0.0023 | - | - | - | - |
790
+ | 1.6762 | 1320 | 0.0005 | - | - | - | - |
791
+ | 1.6889 | 1330 | 0.0004 | - | - | - | - |
792
+ | 1.7016 | 1340 | 0.0003 | - | - | - | - |
793
+ | 1.7143 | 1350 | 0.0347 | - | - | - | - |
794
+ | 1.7270 | 1360 | 0.0339 | - | - | - | - |
795
+ | 1.7397 | 1370 | 0.0003 | - | - | - | - |
796
+ | 1.7524 | 1380 | 0.0005 | - | - | - | - |
797
+ | 1.7651 | 1390 | 0.0002 | - | - | - | - |
798
+ | 1.7778 | 1400 | 0.0031 | - | - | - | - |
799
+ | 1.7905 | 1410 | 0.0002 | - | - | - | - |
800
+ | 1.8032 | 1420 | 0.0012 | - | - | - | - |
801
+ | 1.8159 | 1430 | 0.0002 | - | - | - | - |
802
+ | 1.8286 | 1440 | 0.0002 | - | - | - | - |
803
+ | 1.8413 | 1450 | 0.0004 | - | - | - | - |
804
+ | 1.8540 | 1460 | 0.011 | - | - | - | - |
805
+ | 1.8667 | 1470 | 0.0824 | - | - | - | - |
806
+ | 1.8794 | 1480 | 0.0003 | - | - | - | - |
807
+ | 1.8921 | 1490 | 0.0004 | - | - | - | - |
808
+ | 1.9048 | 1500 | 0.0006 | - | - | - | - |
809
+ | 1.9175 | 1510 | 0.015 | - | - | - | - |
810
+ | 1.9302 | 1520 | 0.0004 | - | - | - | - |
811
+ | 1.9429 | 1530 | 0.0004 | - | - | - | - |
812
+ | 1.9556 | 1540 | 0.0011 | - | - | - | - |
813
+ | 1.9683 | 1550 | 0.0003 | - | - | - | - |
814
+ | 1.9810 | 1560 | 0.0006 | - | - | - | - |
815
+ | 1.9937 | 1570 | 0.0042 | - | - | - | - |
816
+ | 2.0 | 1575 | - | 0.7862 | 0.7855 | 0.7852 | 0.7878 |
817
+ | 2.0063 | 1580 | 0.0005 | - | - | - | - |
818
+ | 2.0190 | 1590 | 0.002 | - | - | - | - |
819
+ | 2.0317 | 1600 | 0.0013 | - | - | - | - |
820
+ | 2.0444 | 1610 | 0.0002 | - | - | - | - |
821
+ | 2.0571 | 1620 | 0.0035 | - | - | - | - |
822
+ | 2.0698 | 1630 | 0.0004 | - | - | - | - |
823
+ | 2.0825 | 1640 | 0.0002 | - | - | - | - |
824
+ | 2.0952 | 1650 | 0.0032 | - | - | - | - |
825
+ | 2.1079 | 1660 | 0.0916 | - | - | - | - |
826
+ | 2.1206 | 1670 | 0.0002 | - | - | - | - |
827
+ | 2.1333 | 1680 | 0.0006 | - | - | - | - |
828
+ | 2.1460 | 1690 | 0.0002 | - | - | - | - |
829
+ | 2.1587 | 1700 | 0.0003 | - | - | - | - |
830
+ | 2.1714 | 1710 | 0.0001 | - | - | - | - |
831
+ | 2.1841 | 1720 | 0.0001 | - | - | - | - |
832
+ | 2.1968 | 1730 | 0.0004 | - | - | - | - |
833
+ | 2.2095 | 1740 | 0.0004 | - | - | - | - |
834
+ | 2.2222 | 1750 | 0.0001 | - | - | - | - |
835
+ | 2.2349 | 1760 | 0.0002 | - | - | - | - |
836
+ | 2.2476 | 1770 | 0.0007 | - | - | - | - |
837
+ | 2.2603 | 1780 | 0.0001 | - | - | - | - |
838
+ | 2.2730 | 1790 | 0.0002 | - | - | - | - |
839
+ | 2.2857 | 1800 | 0.0004 | - | - | - | - |
840
+ | 2.2984 | 1810 | 0.0711 | - | - | - | - |
841
+ | 2.3111 | 1820 | 0.0001 | - | - | - | - |
842
+ | 2.3238 | 1830 | 0.0005 | - | - | - | - |
843
+ | 2.3365 | 1840 | 0.0004 | - | - | - | - |
844
+ | 2.3492 | 1850 | 0.0001 | - | - | - | - |
845
+ | 2.3619 | 1860 | 0.0005 | - | - | - | - |
846
+ | 2.3746 | 1870 | 0.0003 | - | - | - | - |
847
+ | 2.3873 | 1880 | 0.0001 | - | - | - | - |
848
+ | 2.4 | 1890 | 0.0002 | - | - | - | - |
849
+ | 2.4127 | 1900 | 0.0001 | - | - | - | - |
850
+ | 2.4254 | 1910 | 0.0002 | - | - | - | - |
851
+ | 2.4381 | 1920 | 0.0002 | - | - | - | - |
852
+ | 2.4508 | 1930 | 0.0002 | - | - | - | - |
853
+ | 2.4635 | 1940 | 0.0004 | - | - | - | - |
854
+ | 2.4762 | 1950 | 0.0001 | - | - | - | - |
855
+ | 2.4889 | 1960 | 0.0002 | - | - | - | - |
856
+ | 2.5016 | 1970 | 0.0002 | - | - | - | - |
857
+ | 2.5143 | 1980 | 0.0001 | - | - | - | - |
858
+ | 2.5270 | 1990 | 0.0001 | - | - | - | - |
859
+ | 2.5397 | 2000 | 0.0002 | - | - | - | - |
860
+ | 2.5524 | 2010 | 0.0023 | - | - | - | - |
861
+ | 2.5651 | 2020 | 0.0002 | - | - | - | - |
862
+ | 2.5778 | 2030 | 0.0001 | - | - | - | - |
863
+ | 2.5905 | 2040 | 0.0003 | - | - | - | - |
864
+ | 2.6032 | 2050 | 0.0003 | - | - | - | - |
865
+ | 2.6159 | 2060 | 0.0002 | - | - | - | - |
866
+ | 2.6286 | 2070 | 0.0001 | - | - | - | - |
867
+ | 2.6413 | 2080 | 0.0 | - | - | - | - |
868
+ | 2.6540 | 2090 | 0.0001 | - | - | - | - |
869
+ | 2.6667 | 2100 | 0.0001 | - | - | - | - |
870
+ | 2.6794 | 2110 | 0.0001 | - | - | - | - |
871
+ | 2.6921 | 2120 | 0.0001 | - | - | - | - |
872
+ | 2.7048 | 2130 | 0.0001 | - | - | - | - |
873
+ | 2.7175 | 2140 | 0.0048 | - | - | - | - |
874
+ | 2.7302 | 2150 | 0.0005 | - | - | - | - |
875
+ | 2.7429 | 2160 | 0.0001 | - | - | - | - |
876
+ | 2.7556 | 2170 | 0.0001 | - | - | - | - |
877
+ | 2.7683 | 2180 | 0.0001 | - | - | - | - |
878
+ | 2.7810 | 2190 | 0.0001 | - | - | - | - |
879
+ | 2.7937 | 2200 | 0.0001 | - | - | - | - |
880
+ | 2.8063 | 2210 | 0.0001 | - | - | - | - |
881
+ | 2.8190 | 2220 | 0.0001 | - | - | - | - |
882
+ | 2.8317 | 2230 | 0.0002 | - | - | - | - |
883
+ | 2.8444 | 2240 | 0.0036 | - | - | - | - |
884
+ | 2.8571 | 2250 | 0.0001 | - | - | - | - |
885
+ | 2.8698 | 2260 | 0.0368 | - | - | - | - |
886
+ | 2.8825 | 2270 | 0.0003 | - | - | - | - |
887
+ | 2.8952 | 2280 | 0.0002 | - | - | - | - |
888
+ | 2.9079 | 2290 | 0.0001 | - | - | - | - |
889
+ | 2.9206 | 2300 | 0.0005 | - | - | - | - |
890
+ | 2.9333 | 2310 | 0.0001 | - | - | - | - |
891
+ | 2.9460 | 2320 | 0.0001 | - | - | - | - |
892
+ | 2.9587 | 2330 | 0.0003 | - | - | - | - |
893
+ | 2.9714 | 2340 | 0.0001 | - | - | - | - |
894
+ | 2.9841 | 2350 | 0.0001 | - | - | - | - |
895
+ | 2.9968 | 2360 | 0.0002 | - | - | - | - |
896
+ | **2.9994** | **2362** | **-** | **0.7864** | **0.7805** | **0.7838** | **0.7852** |
897
+ | 3.0095 | 2370 | 0.0025 | - | - | - | - |
898
+ | 3.0222 | 2380 | 0.0002 | - | - | - | - |
899
+ | 3.0349 | 2390 | 0.0001 | - | - | - | - |
900
+ | 3.0476 | 2400 | 0.0001 | - | - | - | - |
901
+ | 3.0603 | 2410 | 0.0001 | - | - | - | - |
902
+ | 3.0730 | 2420 | 0.0001 | - | - | - | - |
903
+ | 3.0857 | 2430 | 0.0001 | - | - | - | - |
904
+ | 3.0984 | 2440 | 0.0002 | - | - | - | - |
905
+ | 3.1111 | 2450 | 0.0116 | - | - | - | - |
906
+ | 3.1238 | 2460 | 0.0002 | - | - | - | - |
907
+ | 3.1365 | 2470 | 0.0001 | - | - | - | - |
908
+ | 3.1492 | 2480 | 0.0001 | - | - | - | - |
909
+ | 3.1619 | 2490 | 0.0001 | - | - | - | - |
910
+ | 3.1746 | 2500 | 0.0001 | - | - | - | - |
911
+ | 3.1873 | 2510 | 0.0001 | - | - | - | - |
912
+ | 3.2 | 2520 | 0.0001 | - | - | - | - |
913
+ | 3.2127 | 2530 | 0.0001 | - | - | - | - |
914
+ | 3.2254 | 2540 | 0.0001 | - | - | - | - |
915
+ | 3.2381 | 2550 | 0.0002 | - | - | - | - |
916
+ | 3.2508 | 2560 | 0.0001 | - | - | - | - |
917
+ | 3.2635 | 2570 | 0.0001 | - | - | - | - |
918
+ | 3.2762 | 2580 | 0.0001 | - | - | - | - |
919
+ | 3.2889 | 2590 | 0.0001 | - | - | - | - |
920
+ | 3.3016 | 2600 | 0.063 | - | - | - | - |
921
+ | 3.3143 | 2610 | 0.0001 | - | - | - | - |
922
+ | 3.3270 | 2620 | 0.0001 | - | - | - | - |
923
+ | 3.3397 | 2630 | 0.0001 | - | - | - | - |
924
+ | 3.3524 | 2640 | 0.0001 | - | - | - | - |
925
+ | 3.3651 | 2650 | 0.0002 | - | - | - | - |
926
+ | 3.3778 | 2660 | 0.0001 | - | - | - | - |
927
+ | 3.3905 | 2670 | 0.0001 | - | - | - | - |
928
+ | 3.4032 | 2680 | 0.0001 | - | - | - | - |
929
+ | 3.4159 | 2690 | 0.0001 | - | - | - | - |
930
+ | 3.4286 | 2700 | 0.0001 | - | - | - | - |
931
+ | 3.4413 | 2710 | 0.0001 | - | - | - | - |
932
+ | 3.4540 | 2720 | 0.0002 | - | - | - | - |
933
+ | 3.4667 | 2730 | 0.0001 | - | - | - | - |
934
+ | 3.4794 | 2740 | 0.0001 | - | - | - | - |
935
+ | 3.4921 | 2750 | 0.0001 | - | - | - | - |
936
+ | 3.5048 | 2760 | 0.0001 | - | - | - | - |
937
+ | 3.5175 | 2770 | 0.0002 | - | - | - | - |
938
+ | 3.5302 | 2780 | 0.0001 | - | - | - | - |
939
+ | 3.5429 | 2790 | 0.0001 | - | - | - | - |
940
+ | 3.5556 | 2800 | 0.0001 | - | - | - | - |
941
+ | 3.5683 | 2810 | 0.0001 | - | - | - | - |
942
+ | 3.5810 | 2820 | 0.0001 | - | - | - | - |
943
+ | 3.5937 | 2830 | 0.0001 | - | - | - | - |
944
+ | 3.6063 | 2840 | 0.0001 | - | - | - | - |
945
+ | 3.6190 | 2850 | 0.0 | - | - | - | - |
946
+ | 3.6317 | 2860 | 0.0001 | - | - | - | - |
947
+ | 3.6444 | 2870 | 0.0001 | - | - | - | - |
948
+ | 3.6571 | 2880 | 0.0001 | - | - | - | - |
949
+ | 3.6698 | 2890 | 0.0001 | - | - | - | - |
950
+ | 3.6825 | 2900 | 0.0001 | - | - | - | - |
951
+ | 3.6952 | 2910 | 0.0001 | - | - | - | - |
952
+ | 3.7079 | 2920 | 0.0001 | - | - | - | - |
953
+ | 3.7206 | 2930 | 0.0003 | - | - | - | - |
954
+ | 3.7333 | 2940 | 0.0001 | - | - | - | - |
955
+ | 3.7460 | 2950 | 0.0001 | - | - | - | - |
956
+ | 3.7587 | 2960 | 0.0001 | - | - | - | - |
957
+ | 3.7714 | 2970 | 0.0002 | - | - | - | - |
958
+ | 3.7841 | 2980 | 0.0001 | - | - | - | - |
959
+ | 3.7968 | 2990 | 0.0001 | - | - | - | - |
960
+ | 3.8095 | 3000 | 0.0001 | - | - | - | - |
961
+ | 3.8222 | 3010 | 0.0001 | - | - | - | - |
962
+ | 3.8349 | 3020 | 0.0002 | - | - | - | - |
963
+ | 3.8476 | 3030 | 0.0001 | - | - | - | - |
964
+ | 3.8603 | 3040 | 0.0001 | - | - | - | - |
965
+ | 3.8730 | 3050 | 0.0214 | - | - | - | - |
966
+ | 3.8857 | 3060 | 0.0001 | - | - | - | - |
967
+ | 3.8984 | 3070 | 0.0001 | - | - | - | - |
968
+ | 3.9111 | 3080 | 0.0001 | - | - | - | - |
969
+ | 3.9238 | 3090 | 0.0001 | - | - | - | - |
970
+ | 3.9365 | 3100 | 0.0001 | - | - | - | - |
971
+ | 3.9492 | 3110 | 0.0001 | - | - | - | - |
972
+ | 3.9619 | 3120 | 0.0001 | - | - | - | - |
973
+ | 3.9746 | 3130 | 0.0001 | - | - | - | - |
974
+ | 3.9873 | 3140 | 0.0001 | - | - | - | - |
975
+ | 3.9975 | 3148 | - | 0.7867 | 0.7838 | 0.7827 | 0.7843 |
976
+
977
+ * The bold row denotes the saved checkpoint.
978
+ </details>
979
+
980
+ ### Framework Versions
981
+ - Python: 3.12.2
982
+ - Sentence Transformers: 3.0.1
983
+ - Transformers: 4.41.2
984
+ - PyTorch: 2.2.0+cu121
985
+ - Accelerate: 0.31.0
986
+ - Datasets: 2.19.1
987
+ - Tokenizers: 0.19.1
988
+
989
+ ## Citation
990
+
991
+ ### BibTeX
992
+
993
+ #### Sentence Transformers
994
+ ```bibtex
995
+ @inproceedings{reimers-2019-sentence-bert,
996
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
997
+ author = "Reimers, Nils and Gurevych, Iryna",
998
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
999
+ month = "11",
1000
+ year = "2019",
1001
+ publisher = "Association for Computational Linguistics",
1002
+ url = "https://arxiv.org/abs/1908.10084",
1003
+ }
1004
+ ```
1005
+
1006
+ #### MatryoshkaLoss
1007
+ ```bibtex
1008
+ @misc{kusupati2024matryoshka,
1009
+ title={Matryoshka Representation Learning},
1010
+ 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},
1011
+ year={2024},
1012
+ eprint={2205.13147},
1013
+ archivePrefix={arXiv},
1014
+ primaryClass={cs.LG}
1015
+ }
1016
+ ```
1017
+
1018
+ #### MultipleNegativesRankingLoss
1019
+ ```bibtex
1020
+ @misc{henderson2017efficient,
1021
+ title={Efficient Natural Language Response Suggestion for Smart Reply},
1022
+ 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},
1023
+ year={2017},
1024
+ eprint={1705.00652},
1025
+ archivePrefix={arXiv},
1026
+ primaryClass={cs.CL}
1027
+ }
1028
+ ```
1029
+
1030
+ <!--
1031
+ ## Glossary
1032
+
1033
+ *Clearly define terms in order to be accessible across audiences.*
1034
+ -->
1035
+
1036
+ <!--
1037
+ ## Model Card Authors
1038
+
1039
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
1040
+ -->
1041
+
1042
+ <!--
1043
+ ## Model Card Contact
1044
+
1045
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
1046
+ -->
config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "BAAI/bge-m3",
3
+ "architectures": [
4
+ "XLMRobertaModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 4096,
15
+ "layer_norm_eps": 1e-05,
16
+ "max_position_embeddings": 8194,
17
+ "model_type": "xlm-roberta",
18
+ "num_attention_heads": 16,
19
+ "num_hidden_layers": 24,
20
+ "output_past": true,
21
+ "pad_token_id": 1,
22
+ "position_embedding_type": "absolute",
23
+ "torch_dtype": "float32",
24
+ "transformers_version": "4.41.2",
25
+ "type_vocab_size": 1,
26
+ "use_cache": true,
27
+ "vocab_size": 250002
28
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "3.0.1",
4
+ "transformers": "4.41.2",
5
+ "pytorch": "2.2.0+cu121"
6
+ },
7
+ "prompts": {},
8
+ "default_prompt_name": null,
9
+ "similarity_fn_name": null
10
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81846a61ec24981f9e5f74b84d9a77c27d63c8e1bb9bd20409ab2aaacd068c7c
3
+ size 2271064456
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": 8192,
3
+ "do_lower_case": false
4
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": true,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e4f7e21bec3fb0044ca0bb2d50eb5d4d8c596273c422baef84466d2c73748b9c
3
+ size 17083053
tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "250001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": true,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "mask_token": "<mask>",
49
+ "model_max_length": 8192,
50
+ "pad_token": "<pad>",
51
+ "sep_token": "</s>",
52
+ "sp_model_kwargs": {},
53
+ "tokenizer_class": "XLMRobertaTokenizer",
54
+ "unk_token": "<unk>"
55
+ }