ppuva1 commited on
Commit
ce496eb
·
verified ·
1 Parent(s): fbdc86f

Add new SentenceTransformer model

Browse files
Files changed (2) hide show
  1. README.md +151 -149
  2. model.safetensors +1 -1
README.md CHANGED
@@ -4,23 +4,23 @@ tags:
4
  - sentence-similarity
5
  - feature-extraction
6
  - generated_from_trainer
7
- - dataset_size:208
8
  - loss:BatchSemiHardTripletLoss
9
  base_model: BAAI/bge-base-en
10
  widget:
11
  - source_sentence: '
12
 
13
- Name : Alegro
14
 
15
- Category: Dining Services, Catering
16
 
17
- Department: Sales
18
 
19
- Location: Barcelona, Spain
20
 
21
- Amount: 432.75
22
 
23
- Card: Client Relationship Dinner
24
 
25
  Trip Name: unknown
26
 
@@ -28,23 +28,6 @@ widget:
28
  sentences:
29
  - '
30
 
31
- Name : Pardalis Digital
32
-
33
- Category: Data Analytics Platform, Professional Networking Service
34
-
35
- Department: Sales
36
-
37
- Location: Dublin, Ireland
38
-
39
- Amount: 1456.75
40
-
41
- Card: Sales Intelligence & Networking Platform
42
-
43
- Trip Name: unknown
44
-
45
- '
46
- - '
47
-
48
  Name : Nimbus Networks Inc.
49
 
50
  Category: Cloud Services, Application Hosting
@@ -62,71 +45,71 @@ widget:
62
  '
63
  - '
64
 
65
- Name : Rising Tide Solutions
66
 
67
- Category: IT Resource Management
68
 
69
- Department: Engineering
70
 
71
- Location: Amsterdam, Netherlands
72
 
73
- Amount: 1423.57
74
 
75
- Card: Cloud Transition Project
76
 
77
  Trip Name: unknown
78
 
79
  '
80
- - source_sentence: '
81
 
82
- Name : Freenet AG
83
 
84
- Category: Telecommunication Services
85
 
86
- Department: IT Operations
87
 
88
- Location: Zurich, Switzerland
89
 
90
- Amount: 2794.37
91
 
92
- Card: Infrastructure Support Services
93
 
94
  Trip Name: unknown
95
 
96
  '
97
- sentences:
98
- - '
99
 
100
- Name : True Grid Consulting
101
 
102
- Category: Workspace Optimization, Productivity Tools
103
 
104
  Department: Office Administration
105
 
106
- Location: Tokyo, Japan
107
 
108
- Amount: 1532.75
109
 
110
- Card: Enhanced Collaborative Environment
111
 
112
  Trip Name: unknown
113
 
114
  '
 
115
  - '
116
 
117
- Name : SentBe
118
 
119
- Category: Business Services,
120
 
121
- Department: Finance
122
 
123
- Location: Korea
124
 
125
- Amount: 734.99
126
 
127
- Card: Cross-Border Business Solutions
128
 
129
- Trip Name: 2024Q2-Seoul
130
 
131
  '
132
  - '
@@ -145,20 +128,37 @@ widget:
145
 
146
  Trip Name: unknown
147
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
148
  '
149
  - source_sentence: '
150
 
151
- Name : Technika Solutions
152
 
153
- Category: Networking Equipment Services, System Security
154
 
155
- Department: IT Operations
156
 
157
- Location: Zurich, Switzerland
158
 
159
- Amount: 1384.29
160
 
161
- Card: Integrated Systems Update
162
 
163
  Trip Name: unknown
164
 
@@ -166,68 +166,68 @@ widget:
166
  sentences:
167
  - '
168
 
169
- Name : Connexis Group
170
 
171
- Category: Venue Logistics Services, Corporate Membership Consultancy
172
 
173
- Department: Sales
174
 
175
- Location: Berlin, Germany
176
 
177
- Amount: 1478.55
178
 
179
- Card: International Trade Show Engagement
180
 
181
  Trip Name: unknown
182
 
183
  '
184
  - '
185
 
186
- Name : SkillAdvance Academy
187
 
188
- Category: Online Learning Platform, Professional Development
189
 
190
- Department: Engineering
191
 
192
- Location: Austin, TX
193
 
194
- Amount: 1875.67
195
 
196
- Card: Continuous Improvement Initiative
197
 
198
  Trip Name: unknown
199
 
200
  '
201
  - '
202
 
203
- Name : Apex Global Solutions
204
 
205
- Category: Risk Management Services, Consulting
206
 
207
- Department: Legal
208
 
209
- Location: Sydney, Australia
210
 
211
- Amount: 1750.45
212
 
213
- Card: Annual Compliance Review
214
 
215
  Trip Name: unknown
216
 
217
  '
218
  - source_sentence: '
219
 
220
- Name : Global Wellness Network
221
 
222
- Category: Corporate Wellness Programs, Employee Engagement
223
 
224
- Department: HR
225
 
226
- Location: Berlin, Germany
227
 
228
- Amount: 1285.75
229
 
230
- Card: Wellness and Engagement Program
231
 
232
  Trip Name: unknown
233
 
@@ -235,68 +235,68 @@ widget:
235
  sentences:
236
  - '
237
 
238
- Name : Analytix Global Solutions
239
 
240
- Category: Business Intelligence Services, Regulatory Compliance Tools
241
 
242
- Department: Finance
243
 
244
  Location: London, UK
245
 
246
- Amount: 1323.67
247
 
248
- Card: Financial Compliance Enhancement
249
 
250
  Trip Name: unknown
251
 
252
  '
253
  - '
254
 
255
- Name : InfiniTech Systems
256
 
257
- Category: Technical Hardware Management, Software Solutions
258
 
259
- Department: IT Operations
260
 
261
- Location: New York, USA
262
 
263
- Amount: 1099.47
264
 
265
- Card: Integrated Support Package
266
 
267
  Trip Name: unknown
268
 
269
  '
270
  - '
271
 
272
- Name : Veritas Insights
273
 
274
- Category: Consulting, Recruitment Services
275
 
276
- Department: HR
277
 
278
- Location: Amsterdam, Netherlands
279
 
280
- Amount: 2890.75
281
 
282
- Card: Talent Acquisition Initiative
283
 
284
  Trip Name: unknown
285
 
286
  '
287
  - source_sentence: '
288
 
289
- Name : Viacom Solutions
290
 
291
- Category: Telecom Hardware, Network Architecture
292
 
293
- Department: Engineering
294
 
295
- Location: Tokyo, Japan
296
 
297
- Amount: 1450.67
298
 
299
- Card: Global Network Optimization Project
300
 
301
  Trip Name: unknown
302
 
@@ -304,51 +304,51 @@ widget:
304
  sentences:
305
  - '
306
 
307
- Name : InfoCircle Global
308
 
309
- Category: Business Intelligence Tools, Enterprise Solutions
310
 
311
- Department: Finance
312
 
313
- Location: Singapore
314
 
315
- Amount: 1897.34
316
 
317
- Card: Enterprise Intelligence Subscription
318
 
319
  Trip Name: unknown
320
 
321
  '
322
  - '
323
 
324
- Name : Kepler Dynamics
325
 
326
- Category: Strategic Consultancy, Tech Solutions
327
 
328
- Department: Finance
329
 
330
- Location: Zurich, Switzerland
331
 
332
- Amount: 2375.88
333
 
334
- Card: Integration Strategy Review
335
 
336
  Trip Name: unknown
337
 
338
  '
339
  - '
340
 
341
- Name : GlobalCert Compliance Ltd.
342
 
343
- Category: Legal Consulting, Regulatory Services
344
 
345
  Department: Compliance
346
 
347
- Location: Boston, MA
348
 
349
- Amount: 1342.67
350
 
351
- Card: Annual Regulatory Certification Program
352
 
353
  Trip Name: unknown
354
 
@@ -368,7 +368,7 @@ model-index:
368
  type: bge-base-en-train
369
  metrics:
370
  - type: cosine_accuracy
371
- value: 0.8557692170143127
372
  name: Cosine Accuracy
373
  - task:
374
  type: triplet
@@ -378,7 +378,7 @@ model-index:
378
  type: bge-base-en-eval
379
  metrics:
380
  - type: cosine_accuracy
381
- value: 0.4848484992980957
382
  name: Cosine Accuracy
383
  ---
384
 
@@ -432,9 +432,9 @@ from sentence_transformers import SentenceTransformer
432
  model = SentenceTransformer("ppuva1/finetuned-bge-base-en")
433
  # Run inference
434
  sentences = [
435
- '\nName : Viacom Solutions\nCategory: Telecom Hardware, Network Architecture\nDepartment: Engineering\nLocation: Tokyo, Japan\nAmount: 1450.67\nCard: Global Network Optimization Project\nTrip Name: unknown\n',
436
- '\nName : GlobalCert Compliance Ltd.\nCategory: Legal Consulting, Regulatory Services\nDepartment: Compliance\nLocation: Boston, MA\nAmount: 1342.67\nCard: Annual Regulatory Certification Program\nTrip Name: unknown\n',
437
- '\nName : Kepler Dynamics\nCategory: Strategic Consultancy, Tech Solutions\nDepartment: Finance\nLocation: Zurich, Switzerland\nAmount: 2375.88\nCard: Integration Strategy Review\nTrip Name: unknown\n',
438
  ]
439
  embeddings = model.encode(sentences)
440
  print(embeddings.shape)
@@ -481,7 +481,7 @@ You can finetune this model on your own dataset.
481
 
482
  | Metric | bge-base-en-train | bge-base-en-eval |
483
  |:--------------------|:------------------|:-----------------|
484
- | **cosine_accuracy** | **0.8558** | **0.4848** |
485
 
486
  <!--
487
  ## Bias, Risks and Limitations
@@ -501,38 +501,38 @@ You can finetune this model on your own dataset.
501
 
502
  #### Unnamed Dataset
503
 
504
- * Size: 208 training samples
505
  * Columns: <code>sentence</code> and <code>label</code>
506
- * Approximate statistics based on the first 208 samples:
507
  | | sentence | label |
508
  |:--------|:-----------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
509
  | type | string | int |
510
- | details | <ul><li>min: 33 tokens</li><li>mean: 39.89 tokens</li><li>max: 49 tokens</li></ul> | <ul><li>0: ~4.81%</li><li>1: ~2.40%</li><li>2: ~4.81%</li><li>3: ~3.37%</li><li>4: ~2.88%</li><li>5: ~4.81%</li><li>6: ~4.33%</li><li>7: ~2.88%</li><li>8: ~2.88%</li><li>9: ~6.73%</li><li>10: ~5.77%</li><li>11: ~2.88%</li><li>12: ~2.40%</li><li>13: ~5.77%</li><li>14: ~4.81%</li><li>15: ~4.33%</li><li>16: ~5.29%</li><li>17: ~2.40%</li><li>18: ~3.85%</li><li>19: ~2.40%</li><li>20: ~3.37%</li><li>21: ~3.85%</li><li>22: ~1.92%</li><li>23: ~1.44%</li><li>24: ~4.33%</li><li>25: ~3.85%</li><li>26: ~1.44%</li></ul> |
511
  * Samples:
512
- | sentence | label |
513
- |:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------|
514
- | <code><br>Name : Wong & Lim<br>Category: Technical Equipment Services, Facility Services<br>Department: Office Administration<br>Location: Berlin, Germany<br>Amount: 458.29<br>Card: Monthly Equipment Care Program<br>Trip Name: unknown<br></code> | <code>0</code> |
515
- | <code><br>Name : InterGlobal Tech<br>Category: Business Software Solutions, Data Processing Services<br>Department: Marketing<br>Location: New York, NY<br>Amount: 1249.95<br>Card: Marketing Automation Tools<br>Trip Name: unknown<br></code> | <code>1</code> |
516
- | <code><br>Name : Garrison Transport Group<br>Category: Transportation Services, Logistics Management<br>Department: Sales<br>Location: New York, NY<br>Amount: 317.45<br>Card: Client Acquisition Trip<br>Trip Name: Q3-Sales-Pitch<br></code> | <code>2</code> |
517
  * Loss: [<code>BatchSemiHardTripletLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#batchsemihardtripletloss)
518
 
519
  ### Evaluation Dataset
520
 
521
  #### Unnamed Dataset
522
 
523
- * Size: 52 evaluation samples
524
  * Columns: <code>sentence</code> and <code>label</code>
525
- * Approximate statistics based on the first 52 samples:
526
- | | sentence | label |
527
- |:--------|:-----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
528
- | type | string | int |
529
- | details | <ul><li>min: 32 tokens</li><li>mean: 39.19 tokens</li><li>max: 46 tokens</li></ul> | <ul><li>0: ~3.85%</li><li>2: ~1.92%</li><li>3: ~3.85%</li><li>4: ~1.92%</li><li>5: ~7.69%</li><li>6: ~3.85%</li><li>7: ~3.85%</li><li>8: ~3.85%</li><li>9: ~1.92%</li><li>10: ~3.85%</li><li>11: ~3.85%</li><li>12: ~1.92%</li><li>13: ~1.92%</li><li>14: ~1.92%</li><li>15: ~3.85%</li><li>16: ~1.92%</li><li>17: ~9.62%</li><li>18: ~5.77%</li><li>19: ~5.77%</li><li>20: ~5.77%</li><li>22: ~5.77%</li><li>23: ~1.92%</li><li>24: ~5.77%</li><li>25: ~1.92%</li><li>26: ~5.77%</li></ul> |
530
  * Samples:
531
- | sentence | label |
532
- |:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:----------------|
533
- | <code><br>Name : Freenet AG<br>Category: Telecommunication Services<br>Department: IT Operations<br>Location: Zurich, Switzerland<br>Amount: 2794.37<br>Card: Infrastructure Support Services<br>Trip Name: unknown<br></code> | <code>20</code> |
534
- | <code><br>Name : Analytix Global Solutions<br>Category: Business Intelligence Services, Regulatory Compliance Tools<br>Department: Finance<br>Location: London, UK<br>Amount: 1323.67<br>Card: Financial Compliance Enhancement<br>Trip Name: unknown<br></code> | <code>15</code> |
535
- | <code><br>Name : Allianz<br>Category: Insurance Services, Financial Services<br>Department: Finance<br>Location: New York, NY<br>Amount: 2547.39<br>Card: Quarterly Coverage Evaluation<br>Trip Name: unknown<br></code> | <code>6</code> |
536
  * Loss: [<code>BatchSemiHardTripletLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#batchsemihardtripletloss)
537
 
538
  ### Training Hyperparameters
@@ -668,9 +668,11 @@ You can finetune this model on your own dataset.
668
  </details>
669
 
670
  ### Training Logs
671
- | Epoch | Step | bge-base-en-train_cosine_accuracy | bge-base-en-eval_cosine_accuracy |
672
- |:-----:|:----:|:---------------------------------:|:--------------------------------:|
673
- | -1 | -1 | 0.8558 | 0.4848 |
 
 
674
 
675
 
676
  ### Framework Versions
 
4
  - sentence-similarity
5
  - feature-extraction
6
  - generated_from_trainer
7
+ - dataset_size:416
8
  - loss:BatchSemiHardTripletLoss
9
  base_model: BAAI/bge-base-en
10
  widget:
11
  - source_sentence: '
12
 
13
+ Name : CloudMetric Solutions
14
 
15
+ Category: Data Analytics, Virtual Infrastructure Management
16
 
17
+ Department: Engineering
18
 
19
+ Location: Toronto, Canada
20
 
21
+ Amount: 1644.75
22
 
23
+ Card: Real-Time Resource Monitoring
24
 
25
  Trip Name: unknown
26
 
 
28
  sentences:
29
  - '
30
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
31
  Name : Nimbus Networks Inc.
32
 
33
  Category: Cloud Services, Application Hosting
 
45
  '
46
  - '
47
 
48
+ Name : Allianz
49
 
50
+ Category: Insurance Services, Financial Services
51
 
52
+ Department: Finance
53
 
54
+ Location: New York, NY
55
 
56
+ Amount: 2547.39
57
 
58
+ Card: Quarterly Coverage Evaluation
59
 
60
  Trip Name: unknown
61
 
62
  '
63
+ - '
64
 
65
+ Name : Connexis Group
66
 
67
+ Category: Venue Logistics Services, Corporate Membership Consultancy
68
 
69
+ Department: Sales
70
 
71
+ Location: Berlin, Germany
72
 
73
+ Amount: 1478.55
74
 
75
+ Card: International Trade Show Engagement
76
 
77
  Trip Name: unknown
78
 
79
  '
80
+ - source_sentence: '
 
81
 
82
+ Name : BuroPro Services
83
 
84
+ Category: Facilities Management, Maintenance Solutions
85
 
86
  Department: Office Administration
87
 
88
+ Location: Berlin, Germany
89
 
90
+ Amount: 879.99
91
 
92
+ Card: Monthly Equipment Oversight
93
 
94
  Trip Name: unknown
95
 
96
  '
97
+ sentences:
98
  - '
99
 
100
+ Name : SynthioSolutions Global
101
 
102
+ Category: Technology Consulting, Research Services
103
 
104
+ Department: Research & Development
105
 
106
+ Location: Singapore
107
 
108
+ Amount: 1342.67
109
 
110
+ Card: Advanced Data Integration Project
111
 
112
+ Trip Name: unknown
113
 
114
  '
115
  - '
 
128
 
129
  Trip Name: unknown
130
 
131
+ '
132
+ - '
133
+
134
+ Name : City Shuttle Services
135
+
136
+ Category: Transportation, Logistics
137
+
138
+ Department: Sales
139
+
140
+ Location: San Francisco, CA
141
+
142
+ Amount: 85.0
143
+
144
+ Card: Sales Team Travel Fund
145
+
146
+ Trip Name: Client Meeting in Bay Area
147
+
148
  '
149
  - source_sentence: '
150
 
151
+ Name : SkillAdvance Academy
152
 
153
+ Category: Online Learning Platform, Professional Development
154
 
155
+ Department: Engineering
156
 
157
+ Location: Austin, TX
158
 
159
+ Amount: 1875.67
160
 
161
+ Card: Continuous Improvement Initiative
162
 
163
  Trip Name: unknown
164
 
 
166
  sentences:
167
  - '
168
 
169
+ Name : ComplyTech Solutions
170
 
171
+ Category: Regulatory Software, Consultancy Services
172
 
173
+ Department: Compliance
174
 
175
+ Location: Brussels, Belgium
176
 
177
+ Amount: 1095.45
178
 
179
+ Card: Regulatory Compliance Optimization Plan
180
 
181
  Trip Name: unknown
182
 
183
  '
184
  - '
185
 
186
+ Name : AlphaTech Solutions
187
 
188
+ Category: Computer & Electronics Retail
189
 
190
+ Department: Research & Development
191
 
192
+ Location: Toronto, Canada
193
 
194
+ Amount: 1599.99
195
 
196
+ Card: Innovative Hardware Acquisition
197
 
198
  Trip Name: unknown
199
 
200
  '
201
  - '
202
 
203
+ Name : Craft Gate Systems
204
 
205
+ Category: Payment Processing Gateway, Data Analytics Software
206
 
207
+ Department: Finance
208
 
209
+ Location: Austin, TX
210
 
211
+ Amount: 1132.58
212
 
213
+ Card: Quarterly Revenue Analysis
214
 
215
  Trip Name: unknown
216
 
217
  '
218
  - source_sentence: '
219
 
220
+ Name : Rising Tide Solutions
221
 
222
+ Category: IT Resource Management
223
 
224
+ Department: Engineering
225
 
226
+ Location: Amsterdam, Netherlands
227
 
228
+ Amount: 1423.57
229
 
230
+ Card: Cloud Transition Project
231
 
232
  Trip Name: unknown
233
 
 
235
  sentences:
236
  - '
237
 
238
+ Name : GigaTrend
239
 
240
+ Category: Data Services, Cloud Software Solutions
241
 
242
+ Department: Research & Development
243
 
244
  Location: London, UK
245
 
246
+ Amount: 1345.67
247
 
248
+ Card: Data-Driven Innovation Project
249
 
250
  Trip Name: unknown
251
 
252
  '
253
  - '
254
 
255
+ Name : Apex Innovations Group
256
 
257
+ Category: Business Consulting, Training Services
258
 
259
+ Department: Executive
260
 
261
+ Location: Sydney, Australia
262
 
263
+ Amount: 1575.34
264
 
265
+ Card: Leadership Development Program
266
 
267
  Trip Name: unknown
268
 
269
  '
270
  - '
271
 
272
+ Name : Aegis Risk Consultants
273
 
274
+ Category: Executive Risk Management, Enterprise Solutions
275
 
276
+ Department: Legal
277
 
278
+ Location: London, UK
279
 
280
+ Amount: 1743.56
281
 
282
+ Card: Leadership Liability Initiative
283
 
284
  Trip Name: unknown
285
 
286
  '
287
  - source_sentence: '
288
 
289
+ Name : Allegro Integrations
290
 
291
+ Category: Payment Processing Solutions, Financial Technology Services
292
 
293
+ Department: Finance
294
 
295
+ Location: Dublin, Ireland
296
 
297
+ Amount: 1298.75
298
 
299
+ Card: Bi-annual Financial Systems Audit
300
 
301
  Trip Name: unknown
302
 
 
304
  sentences:
305
  - '
306
 
307
+ Name : Banyan Tree Pte Ltd
308
 
309
+ Category: General Contractors - Residential and Commercial
310
 
311
+ Department: Office Administration
312
 
313
+ Location: Houston, TX
314
 
315
+ Amount: 987.65
316
 
317
+ Card: Operational Infrastructure Management
318
 
319
  Trip Name: unknown
320
 
321
  '
322
  - '
323
 
324
+ Name : InsightWave Research
325
 
326
+ Category: Business Intelligence Consultations, Market Expansion Strategy Services
327
 
328
+ Department: Marketing
329
 
330
+ Location: Tokyo, Japan
331
 
332
+ Amount: 2034.67
333
 
334
+ Card: Global Market Insights Program
335
 
336
  Trip Name: unknown
337
 
338
  '
339
  - '
340
 
341
+ Name : ComplyTech Solutions
342
 
343
+ Category: Regulatory Software, Consultancy Services
344
 
345
  Department: Compliance
346
 
347
+ Location: Brussels, Belgium
348
 
349
+ Amount: 1095.45
350
 
351
+ Card: Regulatory Compliance Optimization Plan
352
 
353
  Trip Name: unknown
354
 
 
368
  type: bge-base-en-train
369
  metrics:
370
  - type: cosine_accuracy
371
+ value: 0.4759615361690521
372
  name: Cosine Accuracy
373
  - task:
374
  type: triplet
 
378
  type: bge-base-en-eval
379
  metrics:
380
  - type: cosine_accuracy
381
+ value: 0.0
382
  name: Cosine Accuracy
383
  ---
384
 
 
432
  model = SentenceTransformer("ppuva1/finetuned-bge-base-en")
433
  # Run inference
434
  sentences = [
435
+ '\nName : Allegro Integrations\nCategory: Payment Processing Solutions, Financial Technology Services\nDepartment: Finance\nLocation: Dublin, Ireland\nAmount: 1298.75\nCard: Bi-annual Financial Systems Audit\nTrip Name: unknown\n',
436
+ '\nName : Banyan Tree Pte Ltd\nCategory: General Contractors - Residential and Commercial\nDepartment: Office Administration\nLocation: Houston, TX\nAmount: 987.65\nCard: Operational Infrastructure Management\nTrip Name: unknown\n',
437
+ '\nName : ComplyTech Solutions\nCategory: Regulatory Software, Consultancy Services\nDepartment: Compliance\nLocation: Brussels, Belgium\nAmount: 1095.45\nCard: Regulatory Compliance Optimization Plan\nTrip Name: unknown\n',
438
  ]
439
  embeddings = model.encode(sentences)
440
  print(embeddings.shape)
 
481
 
482
  | Metric | bge-base-en-train | bge-base-en-eval |
483
  |:--------------------|:------------------|:-----------------|
484
+ | **cosine_accuracy** | **0.476** | **0.0** |
485
 
486
  <!--
487
  ## Bias, Risks and Limitations
 
501
 
502
  #### Unnamed Dataset
503
 
504
+ * Size: 416 training samples
505
  * Columns: <code>sentence</code> and <code>label</code>
506
+ * Approximate statistics based on the first 416 samples:
507
  | | sentence | label |
508
  |:--------|:-----------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
509
  | type | string | int |
510
+ | details | <ul><li>min: 32 tokens</li><li>mean: 39.99 tokens</li><li>max: 49 tokens</li></ul> | <ul><li>0: ~3.12%</li><li>1: ~3.12%</li><li>2: ~3.85%</li><li>3: ~4.81%</li><li>4: ~2.16%</li><li>5: ~4.33%</li><li>6: ~4.57%</li><li>7: ~3.85%</li><li>8: ~5.05%</li><li>9: ~4.09%</li><li>10: ~2.88%</li><li>11: ~4.33%</li><li>12: ~2.16%</li><li>13: ~4.09%</li><li>14: ~3.61%</li><li>15: ~5.77%</li><li>16: ~3.12%</li><li>17: ~6.01%</li><li>18: ~5.05%</li><li>19: ~2.64%</li><li>20: ~3.37%</li><li>21: ~2.88%</li><li>22: ~4.57%</li><li>23: ~2.64%</li><li>24: ~2.64%</li><li>25: ~3.85%</li><li>26: ~1.44%</li></ul> |
511
  * Samples:
512
+ | sentence | label |
513
+ |:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------|
514
+ | <code><br>Name : InnovaThink Global<br>Category: Management Consultancy, Technical Training Services<br>Department: HR<br>Location: Zurich, Switzerland<br>Amount: 1675.32<br>Card: Innovation and Efficiency Program<br>Trip Name: unknown<br></code> | <code>0</code> |
515
+ | <code><br>Name : Global Wellness Network<br>Category: Corporate Wellness Programs, Employee Engagement<br>Department: HR<br>Location: Berlin, Germany<br>Amount: 1285.75<br>Card: Wellness and Engagement Program<br>Trip Name: unknown<br></code> | <code>1</code> |
516
+ | <code><br>Name : Wong & Lim<br>Category: Technical Equipment Services, Facility Services<br>Department: Office Administration<br>Location: Berlin, Germany<br>Amount: 458.29<br>Card: Monthly Equipment Care Program<br>Trip Name: unknown<br></code> | <code>2</code> |
517
  * Loss: [<code>BatchSemiHardTripletLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#batchsemihardtripletloss)
518
 
519
  ### Evaluation Dataset
520
 
521
  #### Unnamed Dataset
522
 
523
+ * Size: 104 evaluation samples
524
  * Columns: <code>sentence</code> and <code>label</code>
525
+ * Approximate statistics based on the first 104 samples:
526
+ | | sentence | label |
527
+ |:--------|:-----------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
528
+ | type | string | int |
529
+ | details | <ul><li>min: 32 tokens</li><li>mean: 39.19 tokens</li><li>max: 46 tokens</li></ul> | <ul><li>0: ~1.92%</li><li>1: ~0.96%</li><li>2: ~4.81%</li><li>3: ~1.92%</li><li>5: ~5.77%</li><li>6: ~7.69%</li><li>7: ~4.81%</li><li>8: ~3.85%</li><li>9: ~5.77%</li><li>10: ~2.88%</li><li>11: ~4.81%</li><li>12: ~2.88%</li><li>13: ~1.92%</li><li>14: ~2.88%</li><li>15: ~0.96%</li><li>16: ~1.92%</li><li>17: ~3.85%</li><li>18: ~4.81%</li><li>19: ~3.85%</li><li>20: ~1.92%</li><li>21: ~0.96%</li><li>22: ~5.77%</li><li>23: ~7.69%</li><li>24: ~7.69%</li><li>25: ~4.81%</li><li>26: ~2.88%</li></ul> |
530
  * Samples:
531
+ | sentence | label |
532
+ |:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:----------------|
533
+ | <code><br>Name : Aegis Risk Consultants<br>Category: Executive Risk Management, Enterprise Solutions<br>Department: Legal<br>Location: London, UK<br>Amount: 1743.56<br>Card: Leadership Liability Initiative<br>Trip Name: unknown<br></code> | <code>11</code> |
534
+ | <code><br>Name : Vinobia Lounge<br>Category: Culinary Experiences, Networking Venues<br>Department: Marketing<br>Location: Dallas, TX<br>Amount: 651.58<br>Card: Innovative Marketing Strategies<br>Trip Name: Annual Marketing Event<br></code> | <code>8</code> |
535
+ | <code><br>Name : Freenet AG<br>Category: Telecommunication Services<br>Department: IT Operations<br>Location: Zurich, Switzerland<br>Amount: 2794.37<br>Card: Infrastructure Support Services<br>Trip Name: unknown<br></code> | <code>25</code> |
536
  * Loss: [<code>BatchSemiHardTripletLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#batchsemihardtripletloss)
537
 
538
  ### Training Hyperparameters
 
668
  </details>
669
 
670
  ### Training Logs
671
+ | Epoch | Step | Training Loss | Validation Loss | bge-base-en-train_cosine_accuracy | bge-base-en-eval_cosine_accuracy |
672
+ |:------:|:----:|:-------------:|:---------------:|:---------------------------------:|:--------------------------------:|
673
+ | -1 | -1 | - | - | 0.8510 | - |
674
+ | 3.8462 | 100 | 4.9979 | 5.0174 | 0.4760 | - |
675
+ | -1 | -1 | - | - | - | 0.0 |
676
 
677
 
678
  ### Framework Versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a47186f8526175cd887439afbc9998126d26374f7042a8ce8cfe1a1882911933
3
  size 437951328
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:898bb5668ad27dc6ca8570e6e3589dec57af7f79f4e35dbebed0c57142369f52
3
  size 437951328