tsss1 commited on
Commit
5f6eeda
·
verified ·
1 Parent(s): e6093b7

Add new SentenceTransformer model

Browse files
.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": 768,
3
+ "pooling_mode_cls_token": false,
4
+ "pooling_mode_mean_tokens": true,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
README.md ADDED
@@ -0,0 +1,1118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: apache-2.0
5
+ tags:
6
+ - sentence-transformers
7
+ - sentence-similarity
8
+ - feature-extraction
9
+ - generated_from_trainer
10
+ - dataset_size:357
11
+ - loss:MatryoshkaLoss
12
+ - loss:MultipleNegativesRankingLoss
13
+ base_model: nomic-ai/nomic-embed-text-v2-moe
14
+ widget:
15
+ - source_sentence: 'Last updated: September 4, 2024
16
+
17
+
18
+ This guide will show you how to use the split tunneling feature on the ExpressVPN
19
+ apps for Windows and Mac.
20
+
21
+
22
+ Important: The split tunneling feature is not available on macOS 11 and above.
23
+ Follow this guide to find out which version of macOS you are using.
24
+
25
+ Note: Split tunneling is also not currently available on the ExpressVPN app for
26
+ Windows 11 on ARM64.'
27
+ sentences:
28
+ - How to use Kodi with ExpressVPN on iOS
29
+ - How to use the split tunneling feature
30
+ - What’s new in ExpressVPN for Linux?
31
+ - source_sentence: 'Last updated: July 5, 2024
32
+
33
+
34
+ Need a VPN or DNS for your device?
35
+
36
+ Get ExpressVPN Now
37
+
38
+
39
+ Password Health is a feature in ExpressVPN Keys designed to help improve your
40
+ overall account security. It offers an assessment of your password security through
41
+ a security score, as well as tips and guidance on how to improve it, protecting
42
+ you against password hacks.
43
+
44
+
45
+ Your security score is calculated based on the strength of your passwords, whether
46
+ you use the same password more than once, and whether the website URLs you stored
47
+ are secure. Your score will improve as you resolve these issues.
48
+
49
+
50
+ Your logins and passwords are assessed locally on your device—they never leave
51
+ your device. Learn more about the privacy of Password Health.
52
+
53
+
54
+ To access Password Health, add at least one login, then tap on the number icon
55
+ at the top-right of the Keys screen.
56
+
57
+
58
+ How to improve your security score?
59
+
60
+
61
+ Learn more about the security issues that can be affecting your security score
62
+ and ways to improve it below.
63
+
64
+
65
+ Weak passwords
66
+
67
+
68
+ Weak passwords can be easy to guess or crack easily by both humans and computers,
69
+ making them vulnerable to password theft or compromise. They include common dictionary
70
+ words, such as “apple” or “boy” or words that are easy to guess, such as place
71
+ names or the names of your partner.
72
+
73
+
74
+ To keep all your accounts safe, generate strong and unique passwords in ExpressVPN
75
+ Keys.
76
+
77
+
78
+ Reused passwords
79
+
80
+
81
+ A reused password is a password you use across multiple services or accounts.
82
+ If your password is compromised, attackers may gain unauthorized access to all
83
+ these accounts with the compromised password. The more accounts you have using
84
+ the same password, the higher the risks involved.
85
+
86
+
87
+ To keep all your accounts secure, use strong and unique passwords for all your
88
+ accounts.
89
+
90
+
91
+ Logins using an unsecure URL
92
+
93
+
94
+ Unsecure URLs start with http:// instead of https://. Any data passing between
95
+ you and the website is not encrypted and can be read by third parties, including
96
+ your passwords, usernames, email address, and credit card numbers.
97
+
98
+
99
+ To keep all your accounts safe, use only https:// in your URLs:
100
+
101
+
102
+ On the Keys login information screen, tap Help Me Fix at the top.
103
+
104
+ Tap Use HTTPS.
105
+
106
+ Exposed passwords
107
+
108
+
109
+ A password is exposed if it has been found in a list of passwords online, often
110
+ as part of data breaches. Attackers often attempt to sign in to services by trying
111
+ all known exposed passwords available on the dark web. The more accounts you have
112
+ using the same exposed password, the more vulnerable you are to being hacked.
113
+
114
+
115
+ Data breaches potentially put your passwords at risk and are being reported at
116
+ an accelerated rate, sometimes multiple times a week. To minimize the risk of
117
+ exposed passwords, use strong and unique passwords for all your accounts and update
118
+ any exposed password as soon as ExpressVPN Keys alerts you to do so.
119
+
120
+
121
+ Two-factor authentication not enabled
122
+
123
+
124
+ Two-factor authentication (2FA) adds an extra layer of protection to your online
125
+ accounts, preventing unauthorized access even if your passwords are compromised.
126
+
127
+
128
+ To keep your accounts safe, you should enable 2FA where possible. ExpressVPN Keys
129
+ lets you add 2FA codes to compatible accounts on the ExpressVPN app for Android
130
+ and iOS. Once set up, Keys can easily generate 2FA codes—also known as time-based
131
+ one-time passwords (TOTPs)—for websites or apps with 2FA enabled.
132
+
133
+
134
+ Important: You can only view and copy 2FA codes on the ExpressVPN Keys browser
135
+ extension. You will be able to add 2FA codes on the browser extension at a later
136
+ date.
137
+
138
+
139
+ Need help? Contact the ExpressVPN Support Team for immediate assistance.
140
+
141
+
142
+ Back to top
143
+
144
+
145
+ Was this article helpful?
146
+
147
+ Yes No'
148
+ sentences:
149
+ - what is batch size in one line
150
+ - What is password health?
151
+ - How to set up ExpressVPN on Amazon Fire tablet
152
+ - source_sentence: "Thanks for the clarification that I can provide information on\
153
+ \ ExpressVPN. I can provide some general information on ExpressVPN, or help with\
154
+ \ troubleshooting common issues. I can provide explanations on how to install\
155
+ \ and set up ExpressVPN on various devices. \n\nSome potential areas to explore\
156
+ \ could be:\n\n1. Troubleshooting common issues with ExpressVPN, such as connection\
157
+ \ problems, slow speeds, or DNS leaks.\n2. Providing step-by-step instructions\
158
+ \ on how to install and set up ExpressVPN on different devices, such as Windows,\
159
+ \ macOS, Android, iOS, or routers.\n3. Explaining ExpressVPN's features, such\
160
+ \ as split tunneling, kill switch, and server locations.\n4. Discussing ExpressVPN's\
161
+ \ pricing plans, refund policy, and customer support.\n\nWhat would you like to\
162
+ \ know about ExpressVPN, or what type of issue would you like to troubleshoot?"
163
+ sentences:
164
+ - "What are the features of ExpressVPN?\n\nExpressVPN is known for its high-speed\
165
+ \ performance and wide range of servers. Some of its key features include:\n\n\
166
+ 1. 128-bit AES encryption, \n2. 256-bit encryption,\n3. 128-bit Blowfish encryption,\n\
167
+ 4. 256-bit encryption,\n5. AES encryption,\n6. AES-256,\n7. AES 256-bit encryption.\n\
168
+ \nNote: The encryption methods are not as stated. \n\n"
169
+ - I can provide information on ExpressVPN. What do you need help with?
170
+ - "I'm trying to use the ExpressVPN app on my Windows 10 computer, but the VPN connection\
171
+ \ isn't working. I've checked the account credentials, and they seem to be correct.\
172
+ \ I've also tried restarting the Windows 10 computer and restarting the ExpressVPN\
173
+ \ app. \n\n"
174
+ - source_sentence: 'Last updated: November 7, 2024
175
+
176
+
177
+ Need a VPN or DNS for your device?
178
+
179
+ Get ExpressVPN Now
180
+
181
+
182
+ This guide will show you how to reset your recovery code for your ExpressVPN Keys
183
+ account. You should reset your recovery code if you have lost it or believe it
184
+ may have been compromised.
185
+
186
+
187
+ To reset your recovery code, you will need your primary password. You can reset
188
+ your recovery code using the ExpressVPN app for Android or the ExpressVPN Keys
189
+ Chrome extension. This functionality is not available in the ExpressVPN app for
190
+ iOS.
191
+
192
+
193
+ ExpressVPN for Android
194
+
195
+ In the ExpressVPN app for Android, tap Options.
196
+
197
+ Tap Other Settings > Security > Reset your recovery code.
198
+
199
+ Tap RESET RECOVERY CODE.
200
+
201
+ Enter your primary password.
202
+
203
+ Store your new recovery code in a safe place.
204
+
205
+ ExpressVPN Keys Chrome extension
206
+
207
+ In the ExpressVPN Keys Chrome extension, click Options.
208
+
209
+ Go to Settings > Security > Reset Recovery Code.
210
+
211
+ Click Reset Recovery Code.
212
+
213
+ Enter your primary password.
214
+
215
+ Store your new recovery code in a safe place.
216
+
217
+
218
+ If you have lost both your primary password and your recovery code, contact the
219
+ ExpressVPN Support Team to reset your password manager account.
220
+
221
+
222
+ Back to top
223
+
224
+
225
+ Was this article helpful?
226
+
227
+ Yes No'
228
+ sentences:
229
+ - How is my ExpressVPN Keys data managed?
230
+ - How to reset your recovery code
231
+ - How to set up ExpressVPN on ASUSTOR with OpenVPN
232
+ - source_sentence: 'Last updated: October 21, 2024
233
+
234
+
235
+ This guide is for users who are having issues streaming Max (formerly HBO Max)
236
+ while connected to the VPN.
237
+
238
+
239
+ To comply with the Max Terms of Use and ExpressVPN Terms of Service, you should
240
+ connect to a server location that matches the country where you are currently
241
+ located.
242
+
243
+
244
+ Jump to…
245
+
246
+
247
+ 1. Change to a different VPN server location
248
+
249
+ 2. Sign out of the Max app, then sign in again
250
+
251
+ 3. Watch Max using your browser
252
+
253
+ 4. Contact ExpressVPN Support
254
+
255
+
256
+ 1. Change to a different VPN server location
257
+
258
+
259
+ If you are a U.S. user having issues streaming Max, try changing to these VPN
260
+ server locations in the following order:
261
+
262
+
263
+ USA – San Francisco
264
+
265
+ USA – Washington DC
266
+
267
+ USA – New York
268
+
269
+ USA – Los Angeles – 1
270
+
271
+
272
+ Below are instructions for changing your VPN server location on:
273
+
274
+
275
+ Windows
276
+
277
+ Mac
278
+
279
+ iOS
280
+
281
+ Android
282
+
283
+ Android TV
284
+
285
+ Apple TV
286
+
287
+ Linux
288
+
289
+ Routers
290
+
291
+ If you are streaming via the Max app, you should force-close it and reopen it
292
+ each time you change location. Below are instructions for force-closing an app
293
+ on:iOS: Swipe up from the bottom of the homescreen, keeping your finger pressed
294
+ until app previews appear at left. Swipe to find the Max app preview, then swipe
295
+ up to close the app.
296
+
297
+
298
+ Android: On your Android device, open your multitasking interface. The way to
299
+ do this varies depending on your device:
300
+
301
+
302
+ If your device has three icons at the bottom of the screen, tap either the three
303
+ vertical lines icon or the square icon.
304
+
305
+ If your device features a single horizontal line at the bottom of the screen,
306
+ swipe up from the bottom to the middle of the screen, hold for a second, then
307
+ release.
308
+
309
+
310
+ Next, swipe to find the Max app preview, then swipe to force-close the app. The
311
+ direction you need to swipe will vary depending on your device.
312
+
313
+
314
+ Android TV: Go to Settings, select Apps, and scroll to find the Max app. Select
315
+ the app, then select Force Stop.
316
+
317
+
318
+ Fire TV/Fire Stick: Go to Settings, select Applications, select Manage Installed
319
+ Applications. Scroll to find the Max app. Select the app, then select Force Stop.
320
+
321
+
322
+ Apple TV: Double-click the TV icon on your remote to see the apps currently running.
323
+ Swipe to find the Max app preview, then swipe up to close the app.
324
+
325
+
326
+ If you are a non-U.S. user having issues streaming Max, proceed to the next step.
327
+
328
+
329
+ Need help? Contact the ExpressVPN Support Team for immediate assistance.
330
+
331
+
332
+ Back to top
333
+
334
+
335
+ 2. Sign out of the Max app, then sign in again
336
+
337
+
338
+ If you are using the Max app, sign out of it, restart your device, and then sign
339
+ back in.
340
+
341
+
342
+ Need help? Contact the ExpressVPN Support Team for immediate assistance.
343
+
344
+
345
+ Back to top
346
+
347
+
348
+ 3. Watch Max on your browser
349
+
350
+
351
+ Try streaming Max via your browser by going to https://www.max.com/login and signing
352
+ in with your Max account details.
353
+
354
+
355
+ If you are having issues streaming Max from your browser while connected to the
356
+ VPN:
357
+
358
+
359
+ Get the ExpressVPN browser extension (available for Windows, Mac, and Linux).
360
+ To use the browser extension, you must also have the ExpressVPN app installed
361
+ on your computer.
362
+
363
+ U.S. users should try connecting to these server locations in the following order:
364
+
365
+ USA – San Francisco
366
+
367
+ USA – Washington DC
368
+
369
+ USA – New York
370
+
371
+ USA – Los Angeles – 1
372
+
373
+
374
+ Non-U.S. users should proceed to the next step.
375
+
376
+
377
+ Try using a different browser. The ExpressVPN browser extension is available on
378
+ Windows, Mac, and Linux, and it works with Chrome, Firefox, Vivaldi, Chromium,
379
+ Brave, and Microsoft Edge. The ExpressVPN app must also be installed.
380
+
381
+
382
+ Need help? Contact the ExpressVPN Support Team for immediate assistance.
383
+
384
+
385
+ Back to top
386
+
387
+
388
+ 4. Contact Support
389
+
390
+
391
+ If you are still unable to stream Max while connected to the VPN, contact the
392
+ ExpressVPN Support Team.
393
+
394
+
395
+ Back to top
396
+
397
+
398
+ ExpressVPN is optimized to work with Max so you can enjoy online privacy and security
399
+ all the time, without the VPN interfering. It should never be used as a means
400
+ of copyright circumvention, which is strictly against our Terms of Service. As
401
+ we cannot see or control what you do while connected to our VPN, you are responsible
402
+ at all times for complying with our terms, the Max Terms of Use, and any applicable
403
+ laws. Compliance requires you to be located in the U.S. while streaming Max with
404
+ ExpressVPN.
405
+
406
+ Was this article helpful?
407
+
408
+ Yes No'
409
+ sentences:
410
+ - Troubleshooting steps for streaming Max
411
+ - "I can help you with various questions and issues related to ExpressVPN. What\
412
+ \ do you need assistance with: \n\n"
413
+ - How to choose the appropriate Wi-Fi settings on a router running ExpressVPN
414
+ pipeline_tag: sentence-similarity
415
+ library_name: sentence-transformers
416
+ metrics:
417
+ - cosine_accuracy@1
418
+ - cosine_accuracy@3
419
+ - cosine_accuracy@5
420
+ - cosine_accuracy@10
421
+ - cosine_precision@1
422
+ - cosine_precision@3
423
+ - cosine_precision@5
424
+ - cosine_precision@10
425
+ - cosine_recall@1
426
+ - cosine_recall@3
427
+ - cosine_recall@5
428
+ - cosine_recall@10
429
+ - cosine_ndcg@10
430
+ - cosine_mrr@10
431
+ - cosine_map@100
432
+ model-index:
433
+ - name: ModernBERT Embed base Legal Matryoshka
434
+ results:
435
+ - task:
436
+ type: information-retrieval
437
+ name: Information Retrieval
438
+ dataset:
439
+ name: dim 768
440
+ type: dim_768
441
+ metrics:
442
+ - type: cosine_accuracy@1
443
+ value: 0.6
444
+ name: Cosine Accuracy@1
445
+ - type: cosine_accuracy@3
446
+ value: 0.775
447
+ name: Cosine Accuracy@3
448
+ - type: cosine_accuracy@5
449
+ value: 0.825
450
+ name: Cosine Accuracy@5
451
+ - type: cosine_accuracy@10
452
+ value: 0.875
453
+ name: Cosine Accuracy@10
454
+ - type: cosine_precision@1
455
+ value: 0.6
456
+ name: Cosine Precision@1
457
+ - type: cosine_precision@3
458
+ value: 0.26666666666666666
459
+ name: Cosine Precision@3
460
+ - type: cosine_precision@5
461
+ value: 0.17
462
+ name: Cosine Precision@5
463
+ - type: cosine_precision@10
464
+ value: 0.09
465
+ name: Cosine Precision@10
466
+ - type: cosine_recall@1
467
+ value: 0.5875
468
+ name: Cosine Recall@1
469
+ - type: cosine_recall@3
470
+ value: 0.775
471
+ name: Cosine Recall@3
472
+ - type: cosine_recall@5
473
+ value: 0.825
474
+ name: Cosine Recall@5
475
+ - type: cosine_recall@10
476
+ value: 0.875
477
+ name: Cosine Recall@10
478
+ - type: cosine_ndcg@10
479
+ value: 0.7464836502947895
480
+ name: Cosine Ndcg@10
481
+ - type: cosine_mrr@10
482
+ value: 0.7041666666666667
483
+ name: Cosine Mrr@10
484
+ - type: cosine_map@100
485
+ value: 0.7103005547084493
486
+ name: Cosine Map@100
487
+ - task:
488
+ type: information-retrieval
489
+ name: Information Retrieval
490
+ dataset:
491
+ name: dim 512
492
+ type: dim_512
493
+ metrics:
494
+ - type: cosine_accuracy@1
495
+ value: 0.6
496
+ name: Cosine Accuracy@1
497
+ - type: cosine_accuracy@3
498
+ value: 0.775
499
+ name: Cosine Accuracy@3
500
+ - type: cosine_accuracy@5
501
+ value: 0.875
502
+ name: Cosine Accuracy@5
503
+ - type: cosine_accuracy@10
504
+ value: 0.9
505
+ name: Cosine Accuracy@10
506
+ - type: cosine_precision@1
507
+ value: 0.6
508
+ name: Cosine Precision@1
509
+ - type: cosine_precision@3
510
+ value: 0.26666666666666666
511
+ name: Cosine Precision@3
512
+ - type: cosine_precision@5
513
+ value: 0.18
514
+ name: Cosine Precision@5
515
+ - type: cosine_precision@10
516
+ value: 0.0925
517
+ name: Cosine Precision@10
518
+ - type: cosine_recall@1
519
+ value: 0.5875
520
+ name: Cosine Recall@1
521
+ - type: cosine_recall@3
522
+ value: 0.775
523
+ name: Cosine Recall@3
524
+ - type: cosine_recall@5
525
+ value: 0.875
526
+ name: Cosine Recall@5
527
+ - type: cosine_recall@10
528
+ value: 0.9
529
+ name: Cosine Recall@10
530
+ - type: cosine_ndcg@10
531
+ value: 0.7552425519590626
532
+ name: Cosine Ndcg@10
533
+ - type: cosine_mrr@10
534
+ value: 0.7083333333333333
535
+ name: Cosine Mrr@10
536
+ - type: cosine_map@100
537
+ value: 0.7123785914060513
538
+ name: Cosine Map@100
539
+ - task:
540
+ type: information-retrieval
541
+ name: Information Retrieval
542
+ dataset:
543
+ name: dim 256
544
+ type: dim_256
545
+ metrics:
546
+ - type: cosine_accuracy@1
547
+ value: 0.625
548
+ name: Cosine Accuracy@1
549
+ - type: cosine_accuracy@3
550
+ value: 0.775
551
+ name: Cosine Accuracy@3
552
+ - type: cosine_accuracy@5
553
+ value: 0.85
554
+ name: Cosine Accuracy@5
555
+ - type: cosine_accuracy@10
556
+ value: 0.875
557
+ name: Cosine Accuracy@10
558
+ - type: cosine_precision@1
559
+ value: 0.625
560
+ name: Cosine Precision@1
561
+ - type: cosine_precision@3
562
+ value: 0.26666666666666666
563
+ name: Cosine Precision@3
564
+ - type: cosine_precision@5
565
+ value: 0.175
566
+ name: Cosine Precision@5
567
+ - type: cosine_precision@10
568
+ value: 0.09
569
+ name: Cosine Precision@10
570
+ - type: cosine_recall@1
571
+ value: 0.6125
572
+ name: Cosine Recall@1
573
+ - type: cosine_recall@3
574
+ value: 0.775
575
+ name: Cosine Recall@3
576
+ - type: cosine_recall@5
577
+ value: 0.85
578
+ name: Cosine Recall@5
579
+ - type: cosine_recall@10
580
+ value: 0.875
581
+ name: Cosine Recall@10
582
+ - type: cosine_ndcg@10
583
+ value: 0.751012115577437
584
+ name: Cosine Ndcg@10
585
+ - type: cosine_mrr@10
586
+ value: 0.7108333333333333
587
+ name: Cosine Mrr@10
588
+ - type: cosine_map@100
589
+ value: 0.7160762727610553
590
+ name: Cosine Map@100
591
+ - task:
592
+ type: information-retrieval
593
+ name: Information Retrieval
594
+ dataset:
595
+ name: dim 128
596
+ type: dim_128
597
+ metrics:
598
+ - type: cosine_accuracy@1
599
+ value: 0.625
600
+ name: Cosine Accuracy@1
601
+ - type: cosine_accuracy@3
602
+ value: 0.725
603
+ name: Cosine Accuracy@3
604
+ - type: cosine_accuracy@5
605
+ value: 0.775
606
+ name: Cosine Accuracy@5
607
+ - type: cosine_accuracy@10
608
+ value: 0.875
609
+ name: Cosine Accuracy@10
610
+ - type: cosine_precision@1
611
+ value: 0.625
612
+ name: Cosine Precision@1
613
+ - type: cosine_precision@3
614
+ value: 0.25
615
+ name: Cosine Precision@3
616
+ - type: cosine_precision@5
617
+ value: 0.16
618
+ name: Cosine Precision@5
619
+ - type: cosine_precision@10
620
+ value: 0.09
621
+ name: Cosine Precision@10
622
+ - type: cosine_recall@1
623
+ value: 0.6125
624
+ name: Cosine Recall@1
625
+ - type: cosine_recall@3
626
+ value: 0.725
627
+ name: Cosine Recall@3
628
+ - type: cosine_recall@5
629
+ value: 0.775
630
+ name: Cosine Recall@5
631
+ - type: cosine_recall@10
632
+ value: 0.875
633
+ name: Cosine Recall@10
634
+ - type: cosine_ndcg@10
635
+ value: 0.7406583468855168
636
+ name: Cosine Ndcg@10
637
+ - type: cosine_mrr@10
638
+ value: 0.6992361111111112
639
+ name: Cosine Mrr@10
640
+ - type: cosine_map@100
641
+ value: 0.7050551531801532
642
+ name: Cosine Map@100
643
+ - task:
644
+ type: information-retrieval
645
+ name: Information Retrieval
646
+ dataset:
647
+ name: dim 64
648
+ type: dim_64
649
+ metrics:
650
+ - type: cosine_accuracy@1
651
+ value: 0.625
652
+ name: Cosine Accuracy@1
653
+ - type: cosine_accuracy@3
654
+ value: 0.7
655
+ name: Cosine Accuracy@3
656
+ - type: cosine_accuracy@5
657
+ value: 0.75
658
+ name: Cosine Accuracy@5
659
+ - type: cosine_accuracy@10
660
+ value: 0.8
661
+ name: Cosine Accuracy@10
662
+ - type: cosine_precision@1
663
+ value: 0.625
664
+ name: Cosine Precision@1
665
+ - type: cosine_precision@3
666
+ value: 0.24166666666666664
667
+ name: Cosine Precision@3
668
+ - type: cosine_precision@5
669
+ value: 0.15500000000000003
670
+ name: Cosine Precision@5
671
+ - type: cosine_precision@10
672
+ value: 0.08249999999999999
673
+ name: Cosine Precision@10
674
+ - type: cosine_recall@1
675
+ value: 0.6125
676
+ name: Cosine Recall@1
677
+ - type: cosine_recall@3
678
+ value: 0.7
679
+ name: Cosine Recall@3
680
+ - type: cosine_recall@5
681
+ value: 0.75
682
+ name: Cosine Recall@5
683
+ - type: cosine_recall@10
684
+ value: 0.8
685
+ name: Cosine Recall@10
686
+ - type: cosine_ndcg@10
687
+ value: 0.7078094654337568
688
+ name: Cosine Ndcg@10
689
+ - type: cosine_mrr@10
690
+ value: 0.6793055555555555
691
+ name: Cosine Mrr@10
692
+ - type: cosine_map@100
693
+ value: 0.6907221236895149
694
+ name: Cosine Map@100
695
+ ---
696
+
697
+ # ModernBERT Embed base Legal Matryoshka
698
+
699
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [nomic-ai/nomic-embed-text-v2-moe](https://huggingface.co/nomic-ai/nomic-embed-text-v2-moe) on the json dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
700
+
701
+ ## Model Details
702
+
703
+ ### Model Description
704
+ - **Model Type:** Sentence Transformer
705
+ - **Base model:** [nomic-ai/nomic-embed-text-v2-moe](https://huggingface.co/nomic-ai/nomic-embed-text-v2-moe) <!-- at revision f6a8873b415144a69ffc529ec1e234d1e00ee765 -->
706
+ - **Maximum Sequence Length:** 512 tokens
707
+ - **Output Dimensionality:** 768 dimensions
708
+ - **Similarity Function:** Cosine Similarity
709
+ - **Training Dataset:**
710
+ - json
711
+ - **Language:** en
712
+ - **License:** apache-2.0
713
+
714
+ ### Model Sources
715
+
716
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
717
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
718
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
719
+
720
+ ### Full Model Architecture
721
+
722
+ ```
723
+ SentenceTransformer(
724
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: NomicBertModel
725
+ (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
726
+ (2): Normalize()
727
+ )
728
+ ```
729
+
730
+ ## Usage
731
+
732
+ ### Direct Usage (Sentence Transformers)
733
+
734
+ First install the Sentence Transformers library:
735
+
736
+ ```bash
737
+ pip install -U sentence-transformers
738
+ ```
739
+
740
+ Then you can load this model and run inference.
741
+ ```python
742
+ from sentence_transformers import SentenceTransformer
743
+
744
+ # Download from the 🤗 Hub
745
+ model = SentenceTransformer("tsss1/expressvpn_embeddingmodel")
746
+ # Run inference
747
+ sentences = [
748
+ 'Last updated: October 21, 2024\n\nThis guide is for users who are having issues streaming Max (formerly HBO Max) while connected to the VPN.\n\nTo comply with the Max Terms of Use and ExpressVPN Terms of Service, you should connect to a server location that matches the country where you are currently located.\n\nJump to…\n\n1. Change to a different VPN server location\n2. Sign out of the Max app, then sign in again\n3. Watch Max using your browser\n4. Contact ExpressVPN Support\n\n1. Change to a different VPN server location\n\nIf you are a U.S. user having issues streaming Max, try changing to these VPN server locations in the following order:\n\nUSA – San Francisco\nUSA – Washington DC\nUSA – New York\nUSA – Los Angeles – 1\n\nBelow are instructions for changing your VPN server location on:\n\nWindows\nMac\niOS\nAndroid\nAndroid TV\nApple TV\nLinux\nRouters\nIf you are streaming via the Max app, you should force-close it and reopen it each time you change location. Below are instructions for force-closing an app on:iOS: Swipe up from the bottom of the homescreen, keeping your finger pressed until app previews appear at left. Swipe to find the Max app preview, then swipe up to close the app.\n\nAndroid: On your Android device, open your multitasking interface. The way to do this varies depending on your device:\n\nIf your device has three icons at the bottom of the screen, tap either the three vertical lines icon or the square icon.\nIf your device features a single horizontal line at the bottom of the screen, swipe up from the bottom to the middle of the screen, hold for a second, then release.\n\nNext, swipe to find the Max app preview, then swipe to force-close the app. The direction you need to swipe will vary depending on your device.\n\nAndroid TV: Go to Settings, select Apps, and scroll to find the Max app. Select the app, then select Force Stop.\n\nFire TV/Fire Stick: Go to Settings, select Applications, select Manage Installed Applications. Scroll to find the Max app. Select the app, then select Force Stop.\n\nApple TV: Double-click the TV icon on your remote to see the apps currently running. Swipe to find the Max app preview, then swipe up to close the app.\n\nIf you are a non-U.S. user having issues streaming Max, proceed to the next step.\n\nNeed help?\xa0Contact the ExpressVPN Support Team for immediate assistance.\n\nBack to top\n\n2. Sign out of the Max app, then sign in again\n\nIf you are using the Max app, sign out of it, restart your device, and then sign back in.\n\nNeed help?\xa0Contact the ExpressVPN Support Team for immediate assistance.\n\nBack to top\n\n3. Watch Max on your browser\n\nTry streaming Max via your browser by going to https://www.max.com/login and signing in with your Max account details.\n\nIf you are having issues streaming Max from your browser while connected to the VPN:\n\nGet the ExpressVPN browser extension (available for Windows, Mac, and Linux). To use the browser extension, you must also have the ExpressVPN app installed on your computer.\nU.S. users should try connecting to these server locations in the following order:\nUSA – San Francisco\nUSA – Washington DC\nUSA – New York\nUSA – Los Angeles – 1\n\nNon-U.S. users should proceed to the next step.\n\nTry using a different browser. The ExpressVPN browser extension is available on Windows, Mac, and Linux, and it works with Chrome, Firefox, Vivaldi, Chromium, Brave, and Microsoft Edge. The ExpressVPN app must also be installed.\n\nNeed help?\xa0Contact the ExpressVPN Support Team for immediate assistance.\n\nBack to top\n\n4. Contact Support\n\nIf you are still unable to stream Max while connected to the VPN, contact the ExpressVPN Support Team.\n\nBack to top\n\nExpressVPN is optimized to work with Max so you can enjoy online privacy and security all the time, without the VPN interfering. It should never be used as a means of copyright circumvention, which is strictly against our Terms of Service. As we cannot see or control what you do while connected to our VPN, you are responsible at all times for complying with our terms, the Max Terms of Use, and any applicable laws. Compliance requires you to be located in the U.S. while streaming Max with ExpressVPN.\nWas this article helpful?\nYes No',
749
+ 'Troubleshooting steps for streaming Max',
750
+ 'I can help you with various questions and issues related to ExpressVPN. What do you need assistance with: \n\n',
751
+ ]
752
+ embeddings = model.encode(sentences)
753
+ print(embeddings.shape)
754
+ # [3, 768]
755
+
756
+ # Get the similarity scores for the embeddings
757
+ similarities = model.similarity(embeddings, embeddings)
758
+ print(similarities.shape)
759
+ # [3, 3]
760
+ ```
761
+
762
+ <!--
763
+ ### Direct Usage (Transformers)
764
+
765
+ <details><summary>Click to see the direct usage in Transformers</summary>
766
+
767
+ </details>
768
+ -->
769
+
770
+ <!--
771
+ ### Downstream Usage (Sentence Transformers)
772
+
773
+ You can finetune this model on your own dataset.
774
+
775
+ <details><summary>Click to expand</summary>
776
+
777
+ </details>
778
+ -->
779
+
780
+ <!--
781
+ ### Out-of-Scope Use
782
+
783
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
784
+ -->
785
+
786
+ ## Evaluation
787
+
788
+ ### Metrics
789
+
790
+ #### Information Retrieval
791
+
792
+ * Datasets: `dim_768`, `dim_512`, `dim_256`, `dim_128` and `dim_64`
793
+ * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
794
+
795
+ | Metric | dim_768 | dim_512 | dim_256 | dim_128 | dim_64 |
796
+ |:--------------------|:-----------|:-----------|:----------|:-----------|:-----------|
797
+ | cosine_accuracy@1 | 0.6 | 0.6 | 0.625 | 0.625 | 0.625 |
798
+ | cosine_accuracy@3 | 0.775 | 0.775 | 0.775 | 0.725 | 0.7 |
799
+ | cosine_accuracy@5 | 0.825 | 0.875 | 0.85 | 0.775 | 0.75 |
800
+ | cosine_accuracy@10 | 0.875 | 0.9 | 0.875 | 0.875 | 0.8 |
801
+ | cosine_precision@1 | 0.6 | 0.6 | 0.625 | 0.625 | 0.625 |
802
+ | cosine_precision@3 | 0.2667 | 0.2667 | 0.2667 | 0.25 | 0.2417 |
803
+ | cosine_precision@5 | 0.17 | 0.18 | 0.175 | 0.16 | 0.155 |
804
+ | cosine_precision@10 | 0.09 | 0.0925 | 0.09 | 0.09 | 0.0825 |
805
+ | cosine_recall@1 | 0.5875 | 0.5875 | 0.6125 | 0.6125 | 0.6125 |
806
+ | cosine_recall@3 | 0.775 | 0.775 | 0.775 | 0.725 | 0.7 |
807
+ | cosine_recall@5 | 0.825 | 0.875 | 0.85 | 0.775 | 0.75 |
808
+ | cosine_recall@10 | 0.875 | 0.9 | 0.875 | 0.875 | 0.8 |
809
+ | **cosine_ndcg@10** | **0.7465** | **0.7552** | **0.751** | **0.7407** | **0.7078** |
810
+ | cosine_mrr@10 | 0.7042 | 0.7083 | 0.7108 | 0.6992 | 0.6793 |
811
+ | cosine_map@100 | 0.7103 | 0.7124 | 0.7161 | 0.7051 | 0.6907 |
812
+
813
+ <!--
814
+ ## Bias, Risks and Limitations
815
+
816
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
817
+ -->
818
+
819
+ <!--
820
+ ### Recommendations
821
+
822
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
823
+ -->
824
+
825
+ ## Training Details
826
+
827
+ ### Training Dataset
828
+
829
+ #### json
830
+
831
+ * Dataset: json
832
+ * Size: 357 training samples
833
+ * Columns: <code>positive</code> and <code>anchor</code>
834
+ * Approximate statistics based on the first 357 samples:
835
+ | | positive | anchor |
836
+ |:--------|:-------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
837
+ | type | string | string |
838
+ | details | <ul><li>min: 21 tokens</li><li>mean: 322.47 tokens</li><li>max: 512 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 32.25 tokens</li><li>max: 512 tokens</li></ul> |
839
+ * Samples:
840
+ | positive | anchor |
841
+ |:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------|
842
+ | <code>I'd like to discuss common issues that users face when using ExpressVPN. <br><br>1. Slow speeds and connectivity issues.<br>2. Difficulty in setting up ExpressVPN on various devices such as routers, smart TVs, and gaming consoles.<br>3. Issues with unblocking geo-restricted content on popular streaming services like Netflix, Hulu, and BBC iPlayer.<br>4. Troubleshooting failed connections and unable to connect to a VPN server.<br><br>Which one of these topics would you like to discuss further, or is there something else you'd like to bring up?</code> | <code>I'd be happy to help with any questions or concerns you have about ExpressVPN. What would you like to know or discuss?</code> |
843
+ | <code>I'd like to provide information about ExpressVPN, but I think it would be more helpful to get some assistance from you. <br><br>I'd like to know more about the process of setting up ExpressVPN on a router. Could you explain the general steps to follow and any potential issues that users may encounter during the setup process? Additionally, are there any specific router models that are known to be compatible with ExpressVPN?</code> | <code>I can help you with any question you have about ExpressVPN. What is it that you need help with?</code> |
844
+ | <code>Last updated: January 11, 2023<br><br>Important: If your ExpressVPN free trial or subscription was initiated via the iOS App Store, refer to this guide.<br><br>This guide will explain how to get or extend an ExpressVPN subscription for iOS users who did not get a free trial or subscription via the App Store.<br><br>Note: Upgrades and renewals are not currently available within the ExpressVPN app for iOS.</code> | <code>ExpressVPN iOS free trial or subscription expiring</code> |
845
+ * Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
846
+ ```json
847
+ {
848
+ "loss": "MultipleNegativesRankingLoss",
849
+ "matryoshka_dims": [
850
+ 768,
851
+ 512,
852
+ 256,
853
+ 128,
854
+ 64
855
+ ],
856
+ "matryoshka_weights": [
857
+ 1,
858
+ 1,
859
+ 1,
860
+ 1,
861
+ 1
862
+ ],
863
+ "n_dims_per_step": -1
864
+ }
865
+ ```
866
+
867
+ ### Training Hyperparameters
868
+ #### Non-Default Hyperparameters
869
+
870
+ - `eval_strategy`: epoch
871
+ - `per_device_train_batch_size`: 2
872
+ - `per_device_eval_batch_size`: 2
873
+ - `gradient_accumulation_steps`: 4
874
+ - `learning_rate`: 2e-05
875
+ - `num_train_epochs`: 7
876
+ - `lr_scheduler_type`: cosine
877
+ - `warmup_ratio`: 0.1
878
+ - `bf16`: True
879
+ - `tf32`: False
880
+ - `load_best_model_at_end`: True
881
+ - `optim`: adamw_torch_fused
882
+ - `batch_sampler`: no_duplicates
883
+
884
+ #### All Hyperparameters
885
+ <details><summary>Click to expand</summary>
886
+
887
+ - `overwrite_output_dir`: False
888
+ - `do_predict`: False
889
+ - `eval_strategy`: epoch
890
+ - `prediction_loss_only`: True
891
+ - `per_device_train_batch_size`: 2
892
+ - `per_device_eval_batch_size`: 2
893
+ - `per_gpu_train_batch_size`: None
894
+ - `per_gpu_eval_batch_size`: None
895
+ - `gradient_accumulation_steps`: 4
896
+ - `eval_accumulation_steps`: None
897
+ - `torch_empty_cache_steps`: None
898
+ - `learning_rate`: 2e-05
899
+ - `weight_decay`: 0.0
900
+ - `adam_beta1`: 0.9
901
+ - `adam_beta2`: 0.999
902
+ - `adam_epsilon`: 1e-08
903
+ - `max_grad_norm`: 1.0
904
+ - `num_train_epochs`: 7
905
+ - `max_steps`: -1
906
+ - `lr_scheduler_type`: cosine
907
+ - `lr_scheduler_kwargs`: {}
908
+ - `warmup_ratio`: 0.1
909
+ - `warmup_steps`: 0
910
+ - `log_level`: passive
911
+ - `log_level_replica`: warning
912
+ - `log_on_each_node`: True
913
+ - `logging_nan_inf_filter`: True
914
+ - `save_safetensors`: True
915
+ - `save_on_each_node`: False
916
+ - `save_only_model`: False
917
+ - `restore_callback_states_from_checkpoint`: False
918
+ - `no_cuda`: False
919
+ - `use_cpu`: False
920
+ - `use_mps_device`: False
921
+ - `seed`: 42
922
+ - `data_seed`: None
923
+ - `jit_mode_eval`: False
924
+ - `use_ipex`: False
925
+ - `bf16`: True
926
+ - `fp16`: False
927
+ - `fp16_opt_level`: O1
928
+ - `half_precision_backend`: auto
929
+ - `bf16_full_eval`: False
930
+ - `fp16_full_eval`: False
931
+ - `tf32`: False
932
+ - `local_rank`: 0
933
+ - `ddp_backend`: None
934
+ - `tpu_num_cores`: None
935
+ - `tpu_metrics_debug`: False
936
+ - `debug`: []
937
+ - `dataloader_drop_last`: False
938
+ - `dataloader_num_workers`: 0
939
+ - `dataloader_prefetch_factor`: None
940
+ - `past_index`: -1
941
+ - `disable_tqdm`: False
942
+ - `remove_unused_columns`: True
943
+ - `label_names`: None
944
+ - `load_best_model_at_end`: True
945
+ - `ignore_data_skip`: False
946
+ - `fsdp`: []
947
+ - `fsdp_min_num_params`: 0
948
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
949
+ - `fsdp_transformer_layer_cls_to_wrap`: None
950
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
951
+ - `deepspeed`: None
952
+ - `label_smoothing_factor`: 0.0
953
+ - `optim`: adamw_torch_fused
954
+ - `optim_args`: None
955
+ - `adafactor`: False
956
+ - `group_by_length`: False
957
+ - `length_column_name`: length
958
+ - `ddp_find_unused_parameters`: None
959
+ - `ddp_bucket_cap_mb`: None
960
+ - `ddp_broadcast_buffers`: False
961
+ - `dataloader_pin_memory`: True
962
+ - `dataloader_persistent_workers`: False
963
+ - `skip_memory_metrics`: True
964
+ - `use_legacy_prediction_loop`: False
965
+ - `push_to_hub`: False
966
+ - `resume_from_checkpoint`: None
967
+ - `hub_model_id`: None
968
+ - `hub_strategy`: every_save
969
+ - `hub_private_repo`: None
970
+ - `hub_always_push`: False
971
+ - `gradient_checkpointing`: False
972
+ - `gradient_checkpointing_kwargs`: None
973
+ - `include_inputs_for_metrics`: False
974
+ - `include_for_metrics`: []
975
+ - `eval_do_concat_batches`: True
976
+ - `fp16_backend`: auto
977
+ - `push_to_hub_model_id`: None
978
+ - `push_to_hub_organization`: None
979
+ - `mp_parameters`:
980
+ - `auto_find_batch_size`: False
981
+ - `full_determinism`: False
982
+ - `torchdynamo`: None
983
+ - `ray_scope`: last
984
+ - `ddp_timeout`: 1800
985
+ - `torch_compile`: False
986
+ - `torch_compile_backend`: None
987
+ - `torch_compile_mode`: None
988
+ - `dispatch_batches`: None
989
+ - `split_batches`: None
990
+ - `include_tokens_per_second`: False
991
+ - `include_num_input_tokens_seen`: False
992
+ - `neftune_noise_alpha`: None
993
+ - `optim_target_modules`: None
994
+ - `batch_eval_metrics`: False
995
+ - `eval_on_start`: False
996
+ - `use_liger_kernel`: False
997
+ - `eval_use_gather_object`: False
998
+ - `average_tokens_across_devices`: False
999
+ - `prompts`: None
1000
+ - `batch_sampler`: no_duplicates
1001
+ - `multi_dataset_batch_sampler`: proportional
1002
+
1003
+ </details>
1004
+
1005
+ ### Training Logs
1006
+ | Epoch | Step | Training Loss | dim_768_cosine_ndcg@10 | dim_512_cosine_ndcg@10 | dim_256_cosine_ndcg@10 | dim_128_cosine_ndcg@10 | dim_64_cosine_ndcg@10 |
1007
+ |:----------:|:------:|:-------------:|:----------------------:|:----------------------:|:----------------------:|:----------------------:|:---------------------:|
1008
+ | 0.2235 | 10 | 2.9921 | - | - | - | - | - |
1009
+ | 0.4469 | 20 | 0.9824 | - | - | - | - | - |
1010
+ | 0.6704 | 30 | 0.6762 | - | - | - | - | - |
1011
+ | 0.8939 | 40 | 0.0133 | - | - | - | - | - |
1012
+ | 0.9832 | 44 | - | 0.7669 | 0.7701 | - | - | - |
1013
+ | 0.2235 | 10 | 0.0179 | - | - | - | - | - |
1014
+ | 0.4469 | 20 | 0.2714 | - | - | - | - | - |
1015
+ | 0.6704 | 30 | 0.0104 | - | - | - | - | - |
1016
+ | 0.8939 | 40 | 0.0015 | - | - | - | - | - |
1017
+ | 0.9832 | 44 | - | 0.7442 | 0.7594 | 0.7465 | 0.7149 | 0.7046 |
1018
+ | 1.1341 | 50 | 0.2207 | - | - | - | - | - |
1019
+ | 1.3575 | 60 | 0.48 | - | - | - | - | - |
1020
+ | 1.5810 | 70 | 0.003 | - | - | - | - | - |
1021
+ | 1.8045 | 80 | 0.2985 | - | - | - | - | - |
1022
+ | **1.9832** | **88** | **-** | **0.7751** | **0.774** | **0.7821** | **0.7746** | **0.7365** |
1023
+ | 2.0447 | 90 | 0.0168 | - | - | - | - | - |
1024
+ | 2.2682 | 100 | 0.0698 | - | - | - | - | - |
1025
+ | 2.4916 | 110 | 0.0054 | - | - | - | - | - |
1026
+ | 2.7151 | 120 | 0.0112 | - | - | - | - | - |
1027
+ | 2.9385 | 130 | 0.0031 | - | - | - | - | - |
1028
+ | 2.9832 | 132 | - | 0.7569 | 0.7537 | 0.7565 | 0.7588 | 0.7251 |
1029
+ | 3.1788 | 140 | 0.1794 | - | - | - | - | - |
1030
+ | 3.4022 | 150 | 0.3266 | - | - | - | - | - |
1031
+ | 3.6257 | 160 | 0.0006 | - | - | - | - | - |
1032
+ | 3.8492 | 170 | 0.0003 | - | - | - | - | - |
1033
+ | 3.9832 | 176 | - | 0.7491 | 0.7613 | 0.7526 | 0.7513 | 0.7206 |
1034
+ | 4.0894 | 180 | 0.2622 | - | - | - | - | - |
1035
+ | 4.3128 | 190 | 0.0004 | - | - | - | - | - |
1036
+ | 4.5363 | 200 | 0.0392 | - | - | - | - | - |
1037
+ | 4.7598 | 210 | 0.3312 | - | - | - | - | - |
1038
+ | 4.9832 | 220 | 0.0021 | 0.7548 | 0.7527 | 0.7466 | 0.7568 | 0.7101 |
1039
+ | 5.2235 | 230 | 0.7593 | - | - | - | - | - |
1040
+ | 5.4469 | 240 | 0.0004 | - | - | - | - | - |
1041
+ | 5.6704 | 250 | 0.0003 | - | - | - | - | - |
1042
+ | 5.8939 | 260 | 0.0154 | - | - | - | - | - |
1043
+ | 5.9832 | 264 | - | 0.7498 | 0.7545 | 0.7510 | 0.7407 | 0.7147 |
1044
+ | 6.1341 | 270 | 0.0162 | - | - | - | - | - |
1045
+ | 6.3575 | 280 | 0.447 | - | - | - | - | - |
1046
+ | 6.5810 | 290 | 0.001 | - | - | - | - | - |
1047
+ | 6.8045 | 300 | 0.1628 | - | - | - | - | - |
1048
+ | 6.9832 | 308 | - | 0.7465 | 0.7552 | 0.7510 | 0.7407 | 0.7078 |
1049
+
1050
+ * The bold row denotes the saved checkpoint.
1051
+
1052
+ ### Framework Versions
1053
+ - Python: 3.11.11
1054
+ - Sentence Transformers: 3.4.1
1055
+ - Transformers: 4.48.3
1056
+ - PyTorch: 2.3.1+cu121
1057
+ - Accelerate: 1.3.0
1058
+ - Datasets: 3.3.2
1059
+ - Tokenizers: 0.21.0
1060
+
1061
+ ## Citation
1062
+
1063
+ ### BibTeX
1064
+
1065
+ #### Sentence Transformers
1066
+ ```bibtex
1067
+ @inproceedings{reimers-2019-sentence-bert,
1068
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
1069
+ author = "Reimers, Nils and Gurevych, Iryna",
1070
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
1071
+ month = "11",
1072
+ year = "2019",
1073
+ publisher = "Association for Computational Linguistics",
1074
+ url = "https://arxiv.org/abs/1908.10084",
1075
+ }
1076
+ ```
1077
+
1078
+ #### MatryoshkaLoss
1079
+ ```bibtex
1080
+ @misc{kusupati2024matryoshka,
1081
+ title={Matryoshka Representation Learning},
1082
+ 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},
1083
+ year={2024},
1084
+ eprint={2205.13147},
1085
+ archivePrefix={arXiv},
1086
+ primaryClass={cs.LG}
1087
+ }
1088
+ ```
1089
+
1090
+ #### MultipleNegativesRankingLoss
1091
+ ```bibtex
1092
+ @misc{henderson2017efficient,
1093
+ title={Efficient Natural Language Response Suggestion for Smart Reply},
1094
+ 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},
1095
+ year={2017},
1096
+ eprint={1705.00652},
1097
+ archivePrefix={arXiv},
1098
+ primaryClass={cs.CL}
1099
+ }
1100
+ ```
1101
+
1102
+ <!--
1103
+ ## Glossary
1104
+
1105
+ *Clearly define terms in order to be accessible across audiences.*
1106
+ -->
1107
+
1108
+ <!--
1109
+ ## Model Card Authors
1110
+
1111
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
1112
+ -->
1113
+
1114
+ <!--
1115
+ ## Model Card Contact
1116
+
1117
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
1118
+ -->
config.json ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "nomic-ai/nomic-embed-text-v2-moe",
3
+ "activation_function": "gelu",
4
+ "add_pooling_layer": false,
5
+ "architectures": [
6
+ "NomicBertModel"
7
+ ],
8
+ "attn_pdrop": 0.0,
9
+ "auto_map": {
10
+ "AutoConfig": "nomic-ai/nomic-bert-2048--configuration_hf_nomic_bert.NomicBertConfig",
11
+ "AutoModel": "nomic-ai/nomic-bert-2048--modeling_hf_nomic_bert.NomicBertModel",
12
+ "AutoModelForMaskedLM": "nomic-ai/nomic-bert-2048--modeling_hf_nomic_bert.NomicBertForPreTraining",
13
+ "AutoModelForMultipleChoice": "nomic-ai/nomic-bert-2048--modeling_hf_nomic_bert.NomicBertForMultipleChoice",
14
+ "AutoModelForQuestionAnswering": "nomic-ai/nomic-bert-2048--modeling_hf_nomic_bert.NomicBertForQuestionAnswering",
15
+ "AutoModelForSequenceClassification": "nomic-ai/nomic-bert-2048--modeling_hf_nomic_bert.NomicBertForSequenceClassification",
16
+ "AutoModelForTokenClassification": "nomic-ai/nomic-bert-2048--modeling_hf_nomic_bert.NomicBertForTokenClassification"
17
+ },
18
+ "bos_token_id": null,
19
+ "causal": false,
20
+ "dense_seq_output": true,
21
+ "embd_pdrop": 0.1,
22
+ "eos_token_id": null,
23
+ "expert_choice_router": false,
24
+ "ffn_div": 1,
25
+ "fused_bias_fc": true,
26
+ "fused_dropout_add_ln": true,
27
+ "initializer_range": 0.02,
28
+ "layer_norm_epsilon": 1e-05,
29
+ "max_trained_positions": 2048,
30
+ "mlp_fc1_bias": true,
31
+ "mlp_fc2_bias": true,
32
+ "model_type": "nomic_bert",
33
+ "moe_every_n_layers": 2,
34
+ "moe_impl": "megablocks",
35
+ "moe_normalize_expert_weights": false,
36
+ "moe_resid_pdrop": 0.0,
37
+ "moe_top_k": 2,
38
+ "n_embd": 768,
39
+ "n_head": 12,
40
+ "n_inner": 3072,
41
+ "n_layer": 12,
42
+ "n_positions": 2048,
43
+ "num_experts": 8,
44
+ "num_shared_experts": 0,
45
+ "pad_token_id": 1,
46
+ "pad_vocab_size_multiple": 64,
47
+ "parallel_block": false,
48
+ "parallel_block_tied_norm": false,
49
+ "prenorm": false,
50
+ "qkv_proj_bias": true,
51
+ "reorder_and_upcast_attn": false,
52
+ "resid_pdrop": 0.0,
53
+ "rotary_emb_base": 10000,
54
+ "rotary_emb_fraction": 1.0,
55
+ "rotary_emb_interleaved": false,
56
+ "rotary_emb_scale_base": null,
57
+ "rotary_scaling_factor": null,
58
+ "router_aux_loss_coef": 0.1,
59
+ "scale_attn_by_inverse_layer_idx": false,
60
+ "scale_attn_weights": true,
61
+ "summary_activation": null,
62
+ "summary_first_dropout": 0.1,
63
+ "summary_proj_to_labels": true,
64
+ "summary_type": "cls_index",
65
+ "summary_use_proj": true,
66
+ "torch_dtype": "float32",
67
+ "transformers_version": "4.48.3",
68
+ "type_vocab_size": 1,
69
+ "use_cache": true,
70
+ "use_flash_attn": true,
71
+ "use_rms_norm": null,
72
+ "use_xentropy": true,
73
+ "vocab_size": 250048
74
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "3.4.1",
4
+ "transformers": "4.48.3",
5
+ "pytorch": "2.3.1+cu121"
6
+ },
7
+ "prompts": {
8
+ "query": "search_query: ",
9
+ "passage": "search_document: ",
10
+ "Classification": "classification: ",
11
+ "MultilabelClassification": "classification: ",
12
+ "Clustering": "clustering: ",
13
+ "PairClassification": "classification: ",
14
+ "STS": "classification: ",
15
+ "Summarization": "classification: ",
16
+ "Speed": "search_document: "
17
+ },
18
+ "default_prompt_name": null,
19
+ "similarity_fn_name": "cosine"
20
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a4f10932b923ce457752d06578d441424e1457e8935c16e9dd5530103e12ef0f
3
+ size 1901187232
modules.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ },
14
+ {
15
+ "idx": 2,
16
+ "name": "2",
17
+ "path": "2_Normalize",
18
+ "type": "sentence_transformers.models.Normalize"
19
+ }
20
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": false
4
+ }
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
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 sha256:883b037111086fd4dfebbbc9b7cee11e1517b5e0c0514879478661440f137085
3
+ size 17082987
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
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizer",
54
+ "unk_token": "<unk>"
55
+ }