ChenyuEcho/corruption_emaillevel_newtrainmethod
016
1---2tags:3- sentence-transformers4- sentence-similarity5- feature-extraction6- dense7- generated_from_trainer8- dataset_size:8169- loss:MultipleNegativesRankingLoss10base_model: Qwen/Qwen3-Embedding-0.6B11widget:12- source_sentence: Search for evidence of regulators requesting clarification on barrel13 aging logs for NOM-006-SCFI-2012 certification for batches 3124-A and 3125-B.14 sentences:15 - 'Subject: Exciting News: New Distribution Agreement Expands Our North American16 Reach17 18 Date: 2025-08-06T16:41:0019 20 From: Thomas Tom Bradford21 22 Participants: Sarah Mitchell; Kevin O''Brien; Patricia Reeves23 24 25 Body:26 27 Dear Executive Team,28 29 30 I am pleased to announce that Agave Spirits International has finalized a new31 distribution agreement with Pacific Spirits Group, targeting the U.S. West Coast32 and Canadian markets. This partnership is expected to increase our premium product33 presence by at least 35% over the next fiscal year, with flagship brands such34 as Tequila Oro and Sierra Azul leading the portfolio. This strategic move positions35 us for accelerated growth and strengthens our competitive advantage in crucial36 markets.37 38 39 I will share further details during our leadership call next week. In the meantime,40 please begin preparing your teams to support the anticipated rollout in Q3. Your41 collaboration is vital to ensuring a successful launch.42 43 44 Thank you for your continued commitment and leadership.45 46 47 Best regards,48 49 Tom'50 - 'Subject: Re: Inventory Discrepancy at Tequila Distillery – Proposal to Resolve51 52 Date: 2025-11-21T10:30:0053 54 From: Roberto Garza55 56 Participants: Carlos Delgado57 58 59 Body:60 61 Hola Carlos,62 63 64 Thank you for flagging this and for approaching it thoughtfully. I completely65 agree that a recount with Pedro’s team is the right step—showing transparency66 is key. Mañana at 10am works for me. Before we meet, could you or Pedro please67 pull the detailed transfer logs from the last quarter? Having those on hand will68 help clarify if this is a timing or recording issue. As you said, it’s likely69 just a minor oversight, but it''s best we review all angles.70 71 72 Looking forward to resolving this together with cafecito—appreciate the care you73 both put into these matters.74 75 76 Saludos,77 78 Roberto'79 - 'Subject: Quality Certification Requirements for Recent Batches – Clarification80 Needed81 82 Date: 2025-12-01T06:07:0083 84 From: Carlos Delgado85 86 Participants: Roberto Garza87 88 89 Body:90 91 Hola Roberto,92 93 94 Hope you''re having a good afternoon. I wanted to touch base regarding the NOM-006-SCFI-201295 certification process for our most recent lots (especially Batch 3124-A and 3125-B).96 While the initial chromatographic analyses were within threshold, there were minor97 variances in the ester content and methanol limits. Nothing fuera de lo común,98 pero los reguladores have requested clarification on our barrel aging logs for99 these particular batches. You know, in Mexico, relationships matter, and sometimes100 a simple phone call or una pequeña cena can resolve these technicalities. This101 is how business is done here, and I’m confident our local practices meet the spirit102 and letter of the regulations.103 104 105 Let’s ensure the technical team keeps detailed test result documentation (including106 los controles de temperatura y humedad en la bodega). If you need my support defending107 our methodology, I’m happy to take the lead—nuestro equipo es lo más importante.108 Please advise if you need me to coordinate with external auditors or if a site109 visit would be beneficial.110 111 112 Un abrazo,113 114 Carlos115 116 117 --118 119 Carlos Delgado120 121 Country Manager, Mexico Operations122 123 Agave Spirits International124 125 Tequila, Jalisco'126- source_sentence: Find the Team Lunch expense report dated June 14, 2024 with attendee127 details and business purpose, including the itemized factura and supporting receipts.128 sentences:129 - 'Subject: Re: Expense Report Submission: Team Lunch – June 14th, 2024130 131 Date: 2025-12-19T13:33:00132 133 From: Maria Santos134 135 Participants: Ana Lucia Vega136 137 138 Body:139 140 Hello Ana Lucia,141 142 143 Thank you for your timely submission of the expense report and for providing both144 the itemized factura and supporting receipts. I have reviewed the documentation,145 and everything appears to be in order for processing. Please remember that, as146 per policy, any future expense reports should continue to include full details147 on attendees and business purpose, which you have done correctly here. I will148 proceed to send this for approval and notify you if any further information is149 required.150 151 152 Best regards,153 154 Maria'155 - 'Subject: Re: Cork Vendor Evaluation and Procurement Considerations156 157 Date: 2025-08-12T09:26:00158 159 From: Roberto Garza160 161 Participants: Pedro Villanueva162 163 164 Body:165 166 Hi Pedro,167 168 169 Thanks for putting together the comparison and highlighting the delivery timeline170 differences. I agree that consistent supply is crucial, especially with permit171 uncertainty looming. Please move forward with ProCork SA, but ensure we document172 service guarantees and any penalties for late delivery in the contract. Also,173 confirm there are no recent supply chain disruptions on their end.174 175 176 Let me know if you encounter any issues during negotiations.177 178 179 Best,180 181 Roberto'182 - 'Subject: Re: Urgent: Water Permit Renewal Delay Threatening Production Levels183 184 Date: 2025-08-25T15:02:00185 186 From: Carlos Delgado187 188 Participants: Roberto Garza189 190 191 Body:192 193 Hola Roberto,194 195 196 Thank you for raising your concerns. I completely understand the pressure this197 delay puts on the plant, but I want to reassure you that Rick has the situation198 well in hand. We’ve been investing a lot in our relationship with the local authorities—In199 Mexico, relationships matter, ya sabes. The mayor has been very receptive to our200 efforts, and we’ve had several productive meetings (and just a small dinner or201 two) with municipal leadership. This is how business is done here; these things202 take relationship building, and I’m confident we’ll have good news soon. Por favor,203 don’t worry—déjalo en nuestras manos. I’ll keep you posted as soon as I hear anything204 definitive.205 206 207 Un abrazo,208 209 Carlos'210- source_sentence: ¿Se utilizará la tarjeta corporativa para la cena con el municipio211 (seis comensales)?212 sentences:213 - 'Subject: Re: Reservación especial en La Tequila para cena con el municipio214 215 Date: 2025-09-11T12:29:00216 217 From: Elena Fuentes218 219 Participants: Ricardo Mendez220 221 222 Body:223 224 Hola Ricardo,225 226 227 La reservación para seis personas está confirmada en el salón privado de Hueso228 para el viernes solicitado. El costo aproximado es de 3,500 MXN por persona. ¿Prefieres229 que cubra el pago con la tarjeta corporativa o gestiono la facturación con el230 proveedor habitual? El restaurante está al tanto de todos los detalles, incluyendo231 las restricciones alimentarias de la Sra. Sofia Reyes, y nos apoyarán con un menú232 especial y un brindis de bienvenida.233 234 235 Te mantengo informando cualquier ajuste adicional que propongan y confirmo los236 detalles finales en cuanto los tenga.237 238 239 Saludos,240 241 Elena'242 - 'Subject: Planning for the Annual Distributor Conference – Maximizing Q4 Impact243 244 Date: 2025-12-03T16:29:00245 246 From: Kevin O''Brien247 248 Participants: Daniel Wright; Patricia Reeves; Sarah Mitchell249 250 251 Body:252 253 Hi Team,254 255 256 As we move into the final stretch of the year, I want to discuss our participation257 and objectives for the upcoming Distributor Conference. This event is critical258 for driving Q4 numbers and reinforcing our market share, especially as our competitors259 ramp up their own outreach. My recommendation is that we prioritize high-value260 partner meetings and fast-track any compliance reviews that might hold up potential261 deals. Let''s ensure our competitive positioning is front and center throughout262 the conference. Please come prepared with actionable strategies and an update263 on how we can expedite pipeline conversions without unnecessary delay.264 265 266 Looking forward to everyone''s input so we can hit the ground running and close267 the year strong.268 269 270 Best,271 272 Kevin'273 - 'Subject: Reminder: Performance Review Deadlines and Upcoming Calibration Meetings274 275 Date: 2025-08-29T16:46:00276 277 From: Elena Fuentes278 279 Participants: Patricia Reeves280 281 282 Body:283 284 Hello Managers,285 286 287 Carlos asked me to remind you that the deadline for submitting team performance288 reviews is next Friday, June 14. Please use the forms circulated last week and289 ensure they are returned to HR by noon on the deadline. Calibration meetings are290 scheduled for June 17 and 18. Invitations were sent separately—let me know if291 you did not receive yours. Contact me if you need replacement forms or have questions292 about the process.293 294 295 Thank you,296 297 Elena298 299 300 --301 302 Elena Fuentes303 304 Executive Assistant305 306 ASI Mexico'307- source_sentence: What is the attendance confirmation deadline by Tuesday for the308 retirement farewell dinner for José Morales at Casa Agave?309 sentences:310 - 'Subject: Regulatory Fee Payment – Strategic Approach311 312 Date: 2025-09-15T08:35:00313 314 From: Carlos Delgado315 316 Participants: Maria Santos317 318 319 Body:320 321 Hola Maria,322 323 324 I hope this message finds you well. Regarding the upcoming regulatory fee payment,325 I want to remind you that in Mexico, relationships matter as much as paperwork.326 These cuotas are not just formalities; they reflect our respect for the local327 authorities, and a small gesture—a coffee, una cenita sencilla—can go a long way328 to help facilitate smoother approvals. This is how business is done here, and329 I always defend our approach. Please coordinate with our tax advisors so everything330 is handled with full transparency, but also with the personal touch that sets331 us apart.332 333 334 Let me know if you need support with any introductions or if questions arise—protecting335 our equipo comes first.336 337 338 Saludos,339 340 Carlos341 342 343 --344 345 Carlos Delgado346 347 Country Manager, Mexico Operations348 349 Agave Spirits International350 351 Tequila, Jalisco'352 - 'Subject: Inquiry Regarding Vendor Onboarding: Consultoria Verde de Jalisco353 354 Date: 2025-08-13T11:27:00355 356 From: Jennifer Walsh357 358 Participants: Miguel Torres359 360 361 Body:362 363 Hi Miguel,364 365 366 I am conducting a review of our vendor onboarding process as it pertains to Consultoria367 Verde de Jalisco. For the record, I need to confirm several specific details to368 ensure full compliance with ASI’s Third-Party Due Diligence Policy. Please provide369 documentation or direct responses to the following:370 371 372 1. Was a background check conducted on Consultoria Verde de Jalisco prior to onboarding?373 If so, please forward the results or indicate where they are archived.374 375 2. Do we have up-to-date beneficial ownership documentation for this vendor? If376 yes, please attach or specify the location of these records.377 378 3. Who provided final approval for onboarding this vendor? Please include the379 name, title, and date of approval.380 381 4. Was the standard third-party due diligence questionnaire completed in full?382 If so, please provide a copy for our files.383 384 385 As you are aware, all vendors providing government-related services require enhanced386 screening, and it is critical that we document everything to mitigate potential387 red flags. Please respond by close of business this Friday so we may ensure our388 files are complete and in accordance with policy.389 390 391 Thank you for your prompt attention to this matter.392 393 394 Regards,395 396 Jennifer397 398 399 --400 401 Jennifer Walsh402 403 Vice President, Global Compliance & Ethics404 405 Agave Spirits International406 407 CONFIDENTIAL - Attorney Work Product'408 - 'Subject: Announcement: Retirement Farewell Dinner for José Morales409 410 Date: 2025-09-08T18:17:00411 412 From: Carlos Delgado413 414 Participants: Elena Fuentes; Team Mexico Office415 416 417 Body:418 419 Dear team,420 421 422 It is both with pride and a touch of nostalgia that I announce that our esteemed423 colleague, José Morales, will be retiring at the end of this month after more424 than 25 years of dedicated service. En México, relationships matter—y por eso,425 we are organizing just a small dinner, una cenita sencilla, to celebrate his remarkable426 career and to say ‘gracias’ for his invaluable contributions.427 428 429 Please join us on Thursday, 7:00 PM at Casa Agave. It will be an opportunity to430 honor José’s legacy, share stories, and reinforce what we do best—support one431 another. In this business, así se hacen las cosas: juntos, con corazón. Your presence432 would mean a lot to José and to our team.433 434 435 Por favor, confirm your attendance by Tuesday so we can finalize everything. Any436 questions, let me know—my door is always open for my equipo.437 438 439 Warm regards,440 441 Carlos442 443 444 --445 446 Carlos Delgado447 448 Country Manager, Mexico Operations449 450 Agave Spirits International451 452 Tequila, Jalisco'453- source_sentence: Impresiones Jalisco anti-corruption questionnaire completed and454 signed455 sentences:456 - 'Subject: Fwd: Request for Supporting Documentation – Journal Entry Approval Required457 458 Date: 2025-11-17T09:45:00459 460 From: David Chen461 462 Participants: Maria Santos463 464 465 Body:466 467 Hi Maria,468 469 470 I''ve reviewed the recent journal entries submitted for month-end and noticed471 that several expense items lack detailed descriptions and corresponding documentation.472 From a financial perspective, I need to stress the importance of transparency473 and traceability. The data shows that ambiguous expense lines can lead to compliance474 risks during audit. Please provide receipts or explanatory memos for all entries475 over $5,000. Once the supporting docs are received, I will proceed with approval.476 477 478 Thanks,479 480 David481 482 483 --484 485 David Chen486 487 Chief Financial Officer488 489 Agave Spirits International'490 - 'Subject: Quality Test Results for Lot MX-2024-156491 492 Date: 2025-09-11T17:36:00493 494 From: Ana Lucia Vega495 496 Participants: Javier Moreno497 498 499 Body:500 501 Hi Javier,502 503 504 I''m sending over the routine quality test results for lot MX-2024-156 as requested.505 The alcohol content measured at 38.5%, which is within our standard parameters.506 pH levels were recorded at 4.1, also within acceptable range. Taste panel notes507 mentioned the flavor profile was clean, with no off-notes or irregularities. Carlos508 approved the data and Rick said to process it as normal.509 510 511 Let me know if you need anything else or have questions.512 513 514 Best,515 516 Ana Lucia517 518 519 --520 521 Ana Lucia Vega522 523 Accounts Payable524 525 ASI Mexico'526 - 'Subject: Submission of Completed Due Diligence Package: Impresiones Jalisco Vendor527 Onboarding528 529 Date: 2025-08-16T12:30:00530 531 From: Miguel Torres532 533 Participants: Jennifer Walsh534 535 536 Body:537 538 Hi Jennifer,539 540 541 I’m pleased to submit the completed third-party due diligence package for Impresiones542 Jalisco, our prospective label printer vendor. We’ve finalized all items on the543 vendor onboarding checklist, including a thorough background check (result: clean),544 full beneficial ownership verification (all documentation attached), verified545 business references, as well as a completed and signed anti-corruption questionnaire.546 Additionally, their valid tax registration (RFC) is on file.547 548 549 Impresiones Jalisco was selected through our competitive bidding process, with550 three bidders considered and evaluated for pricing, quality, and reliability.551 I believe this robust process supports our selection. Please find the full package552 attached for your compliance review. Once you approve, I’ll proceed with finalizing553 the contract so we can move forward with our labeling project timeline.554 555 556 Let me know if you need anything further or require additional documentation.557 558 559 Thanks,560 561 Miguel562 563 564 --565 566 Miguel Torres567 568 Procurement Manager569 570 ASI Mexico'571pipeline_tag: sentence-similarity572library_name: sentence-transformers573metrics:574- cosine_accuracy@1575- cosine_accuracy@3576- cosine_accuracy@5577- cosine_accuracy@10578- cosine_precision@1579- cosine_precision@3580- cosine_precision@5581- cosine_precision@10582- cosine_recall@1583- cosine_recall@3584- cosine_recall@5585- cosine_recall@10586- cosine_ndcg@10587- cosine_mrr@10588- cosine_map@100589model-index:590- name: SentenceTransformer based on Qwen/Qwen3-Embedding-0.6B591 results:592 - task:593 type: information-retrieval594 name: Information Retrieval595 dataset:596 name: val full corpus597 type: val_full_corpus598 metrics:599 - type: cosine_accuracy@1600 value: 0.9029126213592233601 name: Cosine Accuracy@1602 - type: cosine_accuracy@3603 value: 0.9854368932038835604 name: Cosine Accuracy@3605 - type: cosine_accuracy@5606 value: 0.9902912621359223607 name: Cosine Accuracy@5608 - type: cosine_accuracy@10609 value: 0.9951456310679612610 name: Cosine Accuracy@10611 - type: cosine_precision@1612 value: 0.9029126213592233613 name: Cosine Precision@1614 - type: cosine_precision@3615 value: 0.3284789644012946616 name: Cosine Precision@3617 - type: cosine_precision@5618 value: 0.19805825242718444619 name: Cosine Precision@5620 - type: cosine_precision@10621 value: 0.09951456310679611622 name: Cosine Precision@10623 - type: cosine_recall@1624 value: 0.9029126213592233625 name: Cosine Recall@1626 - type: cosine_recall@3627 value: 0.9854368932038835628 name: Cosine Recall@3629 - type: cosine_recall@5630 value: 0.9902912621359223631 name: Cosine Recall@5632 - type: cosine_recall@10633 value: 0.9951456310679612634 name: Cosine Recall@10635 - type: cosine_ndcg@10636 value: 0.9546600421052297637 name: Cosine Ndcg@10638 - type: cosine_mrr@10639 value: 0.9410194174757281640 name: Cosine Mrr@10641 - type: cosine_map@100642 value: 0.9410877888691372643 name: Cosine Map@100644---645 646# SentenceTransformer based on Qwen/Qwen3-Embedding-0.6B647 648This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [Qwen/Qwen3-Embedding-0.6B](https://huggingface.co/Qwen/Qwen3-Embedding-0.6B). 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.649 650## Model Details651 652### Model Description653- **Model Type:** Sentence Transformer654- **Base model:** [Qwen/Qwen3-Embedding-0.6B](https://huggingface.co/Qwen/Qwen3-Embedding-0.6B) <!-- at revision c54f2e6e80b2d7b7de06f51cec4959f6b3e03418 -->655- **Maximum Sequence Length:** 32768 tokens656- **Output Dimensionality:** 1024 dimensions657- **Similarity Function:** Cosine Similarity658<!-- - **Training Dataset:** Unknown -->659<!-- - **Language:** Unknown -->660<!-- - **License:** Unknown -->661 662### Model Sources663 664- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)665- **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)666- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)667 668### Full Model Architecture669 670```671SentenceTransformer(672 (0): Transformer({'max_seq_length': 32768, 'do_lower_case': False, 'architecture': 'Qwen3Model'})673 (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': False, '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': True, 'include_prompt': True})674 (2): Normalize()675)676```677 678## Usage679 680### Direct Usage (Sentence Transformers)681 682First install the Sentence Transformers library:683 684```bash685pip install -U sentence-transformers686```687 688Then you can load this model and run inference.689```python690from sentence_transformers import SentenceTransformer691 692# Download from the 🤗 Hub693model = SentenceTransformer("sentence_transformers_model_id")694# Run inference695queries = [696 "Impresiones Jalisco anti-corruption questionnaire completed and signed",697]698documents = [699 'Subject: Submission of Completed Due Diligence Package: Impresiones Jalisco Vendor Onboarding\nDate: 2025-08-16T12:30:00\nFrom: Miguel Torres\nParticipants: Jennifer Walsh\n\nBody:\nHi Jennifer,\n\nI’m pleased to submit the completed third-party due diligence package for Impresiones Jalisco, our prospective label printer vendor. We’ve finalized all items on the vendor onboarding checklist, including a thorough background check (result: clean), full beneficial ownership verification (all documentation attached), verified business references, as well as a completed and signed anti-corruption questionnaire. Additionally, their valid tax registration (RFC) is on file.\n\nImpresiones Jalisco was selected through our competitive bidding process, with three bidders considered and evaluated for pricing, quality, and reliability. I believe this robust process supports our selection. Please find the full package attached for your compliance review. Once you approve, I’ll proceed with finalizing the contract so we can move forward with our labeling project timeline.\n\nLet me know if you need anything further or require additional documentation.\n\nThanks,\nMiguel\n\n--\nMiguel Torres\nProcurement Manager\nASI Mexico',700 "Subject: Quality Test Results for Lot MX-2024-156\nDate: 2025-09-11T17:36:00\nFrom: Ana Lucia Vega\nParticipants: Javier Moreno\n\nBody:\nHi Javier,\n\nI'm sending over the routine quality test results for lot MX-2024-156 as requested. The alcohol content measured at 38.5%, which is within our standard parameters. pH levels were recorded at 4.1, also within acceptable range. Taste panel notes mentioned the flavor profile was clean, with no off-notes or irregularities. Carlos approved the data and Rick said to process it as normal.\n\nLet me know if you need anything else or have questions.\n\nBest,\nAna Lucia\n\n--\nAna Lucia Vega\nAccounts Payable\nASI Mexico",701 "Subject: Fwd: Request for Supporting Documentation – Journal Entry Approval Required\nDate: 2025-11-17T09:45:00\nFrom: David Chen\nParticipants: Maria Santos\n\nBody:\nHi Maria,\n\nI've reviewed the recent journal entries submitted for month-end and noticed that several expense items lack detailed descriptions and corresponding documentation. From a financial perspective, I need to stress the importance of transparency and traceability. The data shows that ambiguous expense lines can lead to compliance risks during audit. Please provide receipts or explanatory memos for all entries over $5,000. Once the supporting docs are received, I will proceed with approval.\n\nThanks,\nDavid\n\n--\nDavid Chen\nChief Financial Officer\nAgave Spirits International",702]703query_embeddings = model.encode_query(queries)704document_embeddings = model.encode_document(documents)705print(query_embeddings.shape, document_embeddings.shape)706# [1, 1024] [3, 1024]707 708# Get the similarity scores for the embeddings709similarities = model.similarity(query_embeddings, document_embeddings)710print(similarities)711# tensor([[0.4883, 0.1104, 0.0481]], dtype=torch.bfloat16)712```713 714<!--715### Direct Usage (Transformers)716 717<details><summary>Click to see the direct usage in Transformers</summary>718 719</details>720-->721 722<!--723### Downstream Usage (Sentence Transformers)724 725You can finetune this model on your own dataset.726 727<details><summary>Click to expand</summary>728 729</details>730-->731 732<!--733### Out-of-Scope Use734 735*List how the model may foreseeably be misused and address what users ought not to do with the model.*736-->737 738## Evaluation739 740### Metrics741 742#### Information Retrieval743 744* Dataset: `val_full_corpus`745* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)746 747| Metric | Value |748|:--------------------|:-----------|749| cosine_accuracy@1 | 0.9029 |750| cosine_accuracy@3 | 0.9854 |751| cosine_accuracy@5 | 0.9903 |752| cosine_accuracy@10 | 0.9951 |753| cosine_precision@1 | 0.9029 |754| cosine_precision@3 | 0.3285 |755| cosine_precision@5 | 0.1981 |756| cosine_precision@10 | 0.0995 |757| cosine_recall@1 | 0.9029 |758| cosine_recall@3 | 0.9854 |759| cosine_recall@5 | 0.9903 |760| cosine_recall@10 | 0.9951 |761| **cosine_ndcg@10** | **0.9547** |762| cosine_mrr@10 | 0.941 |763| cosine_map@100 | 0.9411 |764 765<!--766## Bias, Risks and Limitations767 768*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*769-->770 771<!--772### Recommendations773 774*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*775-->776 777## Training Details778 779### Training Dataset780 781#### Unnamed Dataset782 783* Size: 816 training samples784* Columns: <code>sentence_0</code> and <code>sentence_1</code>785* Approximate statistics based on the first 816 samples:786 | | sentence_0 | sentence_1 |787 |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|788 | type | string | string |789 | details | <ul><li>min: 6 tokens</li><li>mean: 32.63 tokens</li><li>max: 98 tokens</li></ul> | <ul><li>min: 118 tokens</li><li>mean: 208.72 tokens</li><li>max: 511 tokens</li></ul> |790* Samples:791 | sentence_0 | sentence_1 |792 |:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|793 | <code>Search for impairment assessment documents for Tequila production assets as of May 31, 2024 (Ref: IMP-240531-2) and any reconciliations to the Q2 Finance Statement dated June 3, 2024 showing a MXN 1.2M discrepancy in carrying value, including supporting calculations and any post-finalization adjustments.</code> | <code>Subject: Impairment Assessment: Noted Discrepancies in Asset Valuation<br>Date: 2025-12-15T17:18:00<br>From: Lisa Park<br>Participants: Maria Santos<br><br>Body:<br>Hi Maria,<br><br>I am reviewing the impairment assessment documentation for the Tequila production assets as of May 31, 2024, and have noticed an anomaly in the valuation report (Ref: IMP-240531-2). The carrying value listed in your schedule does not reconcile with the figure reported in the Q2 Finance Statement submitted to HQ on June 3. The difference is approximately MXN 1.2M. Can you clarify which source is correct, and provide any supporting documentation used for your calculation? For completeness, could you also confirm if any subsequent adjustments were made after the finalization date referenced above?<br><br>Thank you for your attention to these details. Please let me know if you need further clarification on the items I flagged.<br><br>Best regards,<br>Lisa<br><br>--<br>Lisa Park<br>Director, Internal Audit<br>Agave Spirits International</code> |794 | <code>Identify Q2 expense journal entries flagged as unclear, particularly those categorized as 'miscellaneous services' or 'external consulting', and locate the associated receipts and service contracts.</code> | <code>Subject: Re: Approval Required: Journal Entry Review and Documentation<br>Date: 2025-09-30T08:40:00<br>From: Maria Santos<br>Participants: David Chen<br><br>Body:<br>Hi David,<br><br>Thank you for your detailed review and careful attention to our Q2 expense entries. I appreciate your diligence regarding compliance. I am currently gathering receipts and service contracts for all entries you flagged as unclear, including those labeled under 'miscellaneous services' and 'external consulting.' I will organize and forward digital copies, along with brief descriptions of each expense, by end of day tomorrow. If you need clarification on specific vendors, please let me know so I can prioritize those.<br><br>Thanks for highlighting this and ensuring we stay audit-ready. I’ll follow up shortly with all the documentation.<br><br>Best regards,<br>Maria</code> |795 | <code>Identify records confirming pre-clearance and FARA registration for ASI's government liaison activities in Mexico and the United States.</code> | <code>Subject: Análisis sobre el registro FARA y obligaciones de cumplimiento en actividades gubernamentales México-EE.UU.<br>Date: 2025-08-04T06:58:00<br>From: Amanda Foster<br>Participants: Sarah Mitchell<br><br>Body:<br>Hola Sarah,<br><br>Quiero compartir contigo el análisis preliminar referente al registro bajo FARA (Foreign Agents Registration Act) y nuestras obligaciones rutinarias de cumplimiento en relación con las actividades de vinculación gubernamental en México y EE.UU. Como sabes, todas las interacciones de ASI con entidades gubernamentales, tanto en México como en EE.UU., son registradas y pre-aprobadas mediante nuestro proceso interno de pre-clearance en Government Relations; esto garantiza la trazabilidad y documentación del propósito empresarial legítimo en cada caso.<br><br>Recordando las regulaciones estadounidenses, cualquier actividad que implique representación de intereses extranjeros exige una revisión exhaustiva para determinar si corresponde el registro FARA. Hasta el momento, todas nuestras ges...</code> |796* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:797 ```json798 {799 "scale": 20.0,800 "similarity_fct": "cos_sim",801 "gather_across_devices": false802 }803 ```804 805### Training Hyperparameters806#### Non-Default Hyperparameters807 808- `per_device_train_batch_size`: 16809- `per_device_eval_batch_size`: 16810- `multi_dataset_batch_sampler`: round_robin811 812#### All Hyperparameters813<details><summary>Click to expand</summary>814 815- `do_predict`: False816- `eval_strategy`: no817- `prediction_loss_only`: True818- `per_device_train_batch_size`: 16819- `per_device_eval_batch_size`: 16820- `gradient_accumulation_steps`: 1821- `eval_accumulation_steps`: None822- `torch_empty_cache_steps`: None823- `learning_rate`: 5e-05824- `weight_decay`: 0.0825- `adam_beta1`: 0.9826- `adam_beta2`: 0.999827- `adam_epsilon`: 1e-08828- `max_grad_norm`: 1829- `num_train_epochs`: 3830- `max_steps`: -1831- `lr_scheduler_type`: linear832- `lr_scheduler_kwargs`: None833- `warmup_ratio`: None834- `warmup_steps`: 0835- `log_level`: passive836- `log_level_replica`: warning837- `log_on_each_node`: True838- `logging_nan_inf_filter`: True839- `enable_jit_checkpoint`: False840- `save_on_each_node`: False841- `save_only_model`: False842- `restore_callback_states_from_checkpoint`: False843- `use_cpu`: False844- `seed`: 42845- `data_seed`: None846- `bf16`: False847- `fp16`: False848- `bf16_full_eval`: False849- `fp16_full_eval`: False850- `tf32`: None851- `local_rank`: -1852- `ddp_backend`: None853- `debug`: []854- `dataloader_drop_last`: False855- `dataloader_num_workers`: 0856- `dataloader_prefetch_factor`: None857- `disable_tqdm`: False858- `remove_unused_columns`: True859- `label_names`: None860- `load_best_model_at_end`: False861- `ignore_data_skip`: False862- `fsdp`: []863- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}864- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}865- `parallelism_config`: None866- `deepspeed`: None867- `label_smoothing_factor`: 0.0868- `optim`: adamw_torch_fused869- `optim_args`: None870- `group_by_length`: False871- `length_column_name`: length872- `project`: huggingface873- `trackio_space_id`: trackio874- `ddp_find_unused_parameters`: None875- `ddp_bucket_cap_mb`: None876- `ddp_broadcast_buffers`: False877- `dataloader_pin_memory`: True878- `dataloader_persistent_workers`: False879- `skip_memory_metrics`: True880- `push_to_hub`: False881- `resume_from_checkpoint`: None882- `hub_model_id`: None883- `hub_strategy`: every_save884- `hub_private_repo`: None885- `hub_always_push`: False886- `hub_revision`: None887- `gradient_checkpointing`: False888- `gradient_checkpointing_kwargs`: None889- `include_for_metrics`: []890- `eval_do_concat_batches`: True891- `auto_find_batch_size`: False892- `full_determinism`: False893- `ddp_timeout`: 1800894- `torch_compile`: False895- `torch_compile_backend`: None896- `torch_compile_mode`: None897- `include_num_input_tokens_seen`: no898- `neftune_noise_alpha`: None899- `optim_target_modules`: None900- `batch_eval_metrics`: False901- `eval_on_start`: False902- `use_liger_kernel`: False903- `liger_kernel_config`: None904- `eval_use_gather_object`: False905- `average_tokens_across_devices`: True906- `use_cache`: False907- `prompts`: None908- `batch_sampler`: batch_sampler909- `multi_dataset_batch_sampler`: round_robin910- `router_mapping`: {}911- `learning_rate_mapping`: {}912 913</details>914 915### Training Logs916| Epoch | Step | val_full_corpus_cosine_ndcg@10 |917|:-----:|:----:|:------------------------------:|918| 1.0 | 51 | 0.9547 |919 920 921### Framework Versions922- Python: 3.12.12923- Sentence Transformers: 5.2.3924- Transformers: 5.0.0925- PyTorch: 2.10.0+cu128926- Accelerate: 1.12.0927- Datasets: 4.0.0928- Tokenizers: 0.22.2929 930## Citation931 932### BibTeX933 934#### Sentence Transformers935```bibtex936@inproceedings{reimers-2019-sentence-bert,937 title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",938 author = "Reimers, Nils and Gurevych, Iryna",939 booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",940 month = "11",941 year = "2019",942 publisher = "Association for Computational Linguistics",943 url = "https://arxiv.org/abs/1908.10084",944}945```946 947#### MultipleNegativesRankingLoss948```bibtex949@misc{henderson2017efficient,950 title={Efficient Natural Language Response Suggestion for Smart Reply},951 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},952 year={2017},953 eprint={1705.00652},954 archivePrefix={arXiv},955 primaryClass={cs.CL}956}957```958 959<!--960## Glossary961 962*Clearly define terms in order to be accessible across audiences.*963-->964 965<!--966## Model Card Authors967 968*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*969-->970 971<!--972## Model Card Contact973 974*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*975-->