--- tags: - sentence-transformers - sentence-similarity - feature-extraction - dense - generated_from_trainer - dataset_size:816 - loss:MultipleNegativesRankingLoss base_model: Qwen/Qwen3-Embedding-0.6B widget: - source_sentence: Search for evidence of regulators requesting clarification on barrel aging logs for NOM-006-SCFI-2012 certification for batches 3124-A and 3125-B. sentences: - 'Subject: Exciting News: New Distribution Agreement Expands Our North American Reach Date: 2025-08-06T16:41:00 From: Thomas Tom Bradford Participants: Sarah Mitchell; Kevin O''Brien; Patricia Reeves Body: Dear Executive Team, I am pleased to announce that Agave Spirits International has finalized a new distribution agreement with Pacific Spirits Group, targeting the U.S. West Coast and Canadian markets. This partnership is expected to increase our premium product presence by at least 35% over the next fiscal year, with flagship brands such as Tequila Oro and Sierra Azul leading the portfolio. This strategic move positions us for accelerated growth and strengthens our competitive advantage in crucial markets. I will share further details during our leadership call next week. In the meantime, please begin preparing your teams to support the anticipated rollout in Q3. Your collaboration is vital to ensuring a successful launch. Thank you for your continued commitment and leadership. Best regards, Tom' - 'Subject: Re: Inventory Discrepancy at Tequila Distillery – Proposal to Resolve Date: 2025-11-21T10:30:00 From: Roberto Garza Participants: Carlos Delgado Body: Hola Carlos, Thank you for flagging this and for approaching it thoughtfully. I completely agree that a recount with Pedro’s team is the right step—showing transparency is key. Mañana at 10am works for me. Before we meet, could you or Pedro please pull the detailed transfer logs from the last quarter? Having those on hand will help clarify if this is a timing or recording issue. As you said, it’s likely just a minor oversight, but it''s best we review all angles. Looking forward to resolving this together with cafecito—appreciate the care you both put into these matters. Saludos, Roberto' - 'Subject: Quality Certification Requirements for Recent Batches – Clarification Needed Date: 2025-12-01T06:07:00 From: Carlos Delgado Participants: Roberto Garza Body: Hola Roberto, Hope you''re having a good afternoon. I wanted to touch base regarding the NOM-006-SCFI-2012 certification process for our most recent lots (especially Batch 3124-A and 3125-B). While the initial chromatographic analyses were within threshold, there were minor variances in the ester content and methanol limits. Nothing fuera de lo común, pero los reguladores have requested clarification on our barrel aging logs for these particular batches. You know, in Mexico, relationships matter, and sometimes a simple phone call or una pequeña cena can resolve these technicalities. This is how business is done here, and I’m confident our local practices meet the spirit and letter of the regulations. Let’s ensure the technical team keeps detailed test result documentation (including los controles de temperatura y humedad en la bodega). If you need my support defending our methodology, I’m happy to take the lead—nuestro equipo es lo más importante. Please advise if you need me to coordinate with external auditors or if a site visit would be beneficial. Un abrazo, Carlos -- Carlos Delgado Country Manager, Mexico Operations Agave Spirits International Tequila, Jalisco' - source_sentence: Find the Team Lunch expense report dated June 14, 2024 with attendee details and business purpose, including the itemized factura and supporting receipts. sentences: - 'Subject: Re: Expense Report Submission: Team Lunch – June 14th, 2024 Date: 2025-12-19T13:33:00 From: Maria Santos Participants: Ana Lucia Vega Body: Hello Ana Lucia, Thank you for your timely submission of the expense report and for providing both the itemized factura and supporting receipts. I have reviewed the documentation, and everything appears to be in order for processing. Please remember that, as per policy, any future expense reports should continue to include full details on attendees and business purpose, which you have done correctly here. I will proceed to send this for approval and notify you if any further information is required. Best regards, Maria' - 'Subject: Re: Cork Vendor Evaluation and Procurement Considerations Date: 2025-08-12T09:26:00 From: Roberto Garza Participants: Pedro Villanueva Body: Hi Pedro, Thanks for putting together the comparison and highlighting the delivery timeline differences. I agree that consistent supply is crucial, especially with permit uncertainty looming. Please move forward with ProCork SA, but ensure we document service guarantees and any penalties for late delivery in the contract. Also, confirm there are no recent supply chain disruptions on their end. Let me know if you encounter any issues during negotiations. Best, Roberto' - 'Subject: Re: Urgent: Water Permit Renewal Delay Threatening Production Levels Date: 2025-08-25T15:02:00 From: Carlos Delgado Participants: Roberto Garza Body: Hola Roberto, Thank you for raising your concerns. I completely understand the pressure this delay puts on the plant, but I want to reassure you that Rick has the situation well in hand. We’ve been investing a lot in our relationship with the local authorities—In Mexico, relationships matter, ya sabes. The mayor has been very receptive to our efforts, and we’ve had several productive meetings (and just a small dinner or two) with municipal leadership. This is how business is done here; these things take relationship building, and I’m confident we’ll have good news soon. Por favor, don’t worry—déjalo en nuestras manos. I’ll keep you posted as soon as I hear anything definitive. Un abrazo, Carlos' - source_sentence: ¿Se utilizará la tarjeta corporativa para la cena con el municipio (seis comensales)? sentences: - 'Subject: Re: Reservación especial en La Tequila para cena con el municipio Date: 2025-09-11T12:29:00 From: Elena Fuentes Participants: Ricardo Mendez Body: Hola Ricardo, La reservación para seis personas está confirmada en el salón privado de Hueso para el viernes solicitado. El costo aproximado es de 3,500 MXN por persona. ¿Prefieres que cubra el pago con la tarjeta corporativa o gestiono la facturación con el proveedor habitual? El restaurante está al tanto de todos los detalles, incluyendo las restricciones alimentarias de la Sra. Sofia Reyes, y nos apoyarán con un menú especial y un brindis de bienvenida. Te mantengo informando cualquier ajuste adicional que propongan y confirmo los detalles finales en cuanto los tenga. Saludos, Elena' - 'Subject: Planning for the Annual Distributor Conference – Maximizing Q4 Impact Date: 2025-12-03T16:29:00 From: Kevin O''Brien Participants: Daniel Wright; Patricia Reeves; Sarah Mitchell Body: Hi Team, As we move into the final stretch of the year, I want to discuss our participation and objectives for the upcoming Distributor Conference. This event is critical for driving Q4 numbers and reinforcing our market share, especially as our competitors ramp up their own outreach. My recommendation is that we prioritize high-value partner meetings and fast-track any compliance reviews that might hold up potential deals. Let''s ensure our competitive positioning is front and center throughout the conference. Please come prepared with actionable strategies and an update on how we can expedite pipeline conversions without unnecessary delay. Looking forward to everyone''s input so we can hit the ground running and close the year strong. Best, Kevin' - 'Subject: Reminder: Performance Review Deadlines and Upcoming Calibration Meetings Date: 2025-08-29T16:46:00 From: Elena Fuentes Participants: Patricia Reeves Body: Hello Managers, Carlos asked me to remind you that the deadline for submitting team performance reviews is next Friday, June 14. Please use the forms circulated last week and ensure they are returned to HR by noon on the deadline. Calibration meetings are scheduled for June 17 and 18. Invitations were sent separately—let me know if you did not receive yours. Contact me if you need replacement forms or have questions about the process. Thank you, Elena -- Elena Fuentes Executive Assistant ASI Mexico' - source_sentence: What is the attendance confirmation deadline by Tuesday for the retirement farewell dinner for José Morales at Casa Agave? sentences: - 'Subject: Regulatory Fee Payment – Strategic Approach Date: 2025-09-15T08:35:00 From: Carlos Delgado Participants: Maria Santos Body: Hola Maria, I hope this message finds you well. Regarding the upcoming regulatory fee payment, I want to remind you that in Mexico, relationships matter as much as paperwork. These cuotas are not just formalities; they reflect our respect for the local authorities, and a small gesture—a coffee, una cenita sencilla—can go a long way to help facilitate smoother approvals. This is how business is done here, and I always defend our approach. Please coordinate with our tax advisors so everything is handled with full transparency, but also with the personal touch that sets us apart. Let me know if you need support with any introductions or if questions arise—protecting our equipo comes first. Saludos, Carlos -- Carlos Delgado Country Manager, Mexico Operations Agave Spirits International Tequila, Jalisco' - 'Subject: Inquiry Regarding Vendor Onboarding: Consultoria Verde de Jalisco Date: 2025-08-13T11:27:00 From: Jennifer Walsh Participants: Miguel Torres Body: Hi Miguel, I am conducting a review of our vendor onboarding process as it pertains to Consultoria Verde de Jalisco. For the record, I need to confirm several specific details to ensure full compliance with ASI’s Third-Party Due Diligence Policy. Please provide documentation or direct responses to the following: 1. Was a background check conducted on Consultoria Verde de Jalisco prior to onboarding? If so, please forward the results or indicate where they are archived. 2. Do we have up-to-date beneficial ownership documentation for this vendor? If yes, please attach or specify the location of these records. 3. Who provided final approval for onboarding this vendor? Please include the name, title, and date of approval. 4. Was the standard third-party due diligence questionnaire completed in full? If so, please provide a copy for our files. As you are aware, all vendors providing government-related services require enhanced screening, and it is critical that we document everything to mitigate potential red flags. Please respond by close of business this Friday so we may ensure our files are complete and in accordance with policy. Thank you for your prompt attention to this matter. Regards, Jennifer -- Jennifer Walsh Vice President, Global Compliance & Ethics Agave Spirits International CONFIDENTIAL - Attorney Work Product' - 'Subject: Announcement: Retirement Farewell Dinner for José Morales Date: 2025-09-08T18:17:00 From: Carlos Delgado Participants: Elena Fuentes; Team Mexico Office Body: Dear team, It is both with pride and a touch of nostalgia that I announce that our esteemed colleague, José Morales, will be retiring at the end of this month after more than 25 years of dedicated service. En México, relationships matter—y por eso, we are organizing just a small dinner, una cenita sencilla, to celebrate his remarkable career and to say ‘gracias’ for his invaluable contributions. Please join us on Thursday, 7:00 PM at Casa Agave. It will be an opportunity to honor José’s legacy, share stories, and reinforce what we do best—support one another. In this business, así se hacen las cosas: juntos, con corazón. Your presence would mean a lot to José and to our team. Por favor, confirm your attendance by Tuesday so we can finalize everything. Any questions, let me know—my door is always open for my equipo. Warm regards, Carlos -- Carlos Delgado Country Manager, Mexico Operations Agave Spirits International Tequila, Jalisco' - source_sentence: Impresiones Jalisco anti-corruption questionnaire completed and signed sentences: - 'Subject: Fwd: Request for Supporting Documentation – Journal Entry Approval Required Date: 2025-11-17T09:45:00 From: David Chen Participants: Maria Santos Body: Hi Maria, I''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. Thanks, David -- David Chen Chief Financial Officer Agave Spirits International' - 'Subject: Quality Test Results for Lot MX-2024-156 Date: 2025-09-11T17:36:00 From: Ana Lucia Vega Participants: Javier Moreno Body: Hi Javier, I''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. Let me know if you need anything else or have questions. Best, Ana Lucia -- Ana Lucia Vega Accounts Payable ASI Mexico' - 'Subject: Submission of Completed Due Diligence Package: Impresiones Jalisco Vendor Onboarding Date: 2025-08-16T12:30:00 From: Miguel Torres Participants: Jennifer Walsh Body: Hi Jennifer, I’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. Impresiones 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. Let me know if you need anything further or require additional documentation. Thanks, Miguel -- Miguel Torres Procurement Manager ASI Mexico' pipeline_tag: sentence-similarity library_name: sentence-transformers metrics: - cosine_accuracy@1 - cosine_accuracy@3 - cosine_accuracy@5 - cosine_accuracy@10 - cosine_precision@1 - cosine_precision@3 - cosine_precision@5 - cosine_precision@10 - cosine_recall@1 - cosine_recall@3 - cosine_recall@5 - cosine_recall@10 - cosine_ndcg@10 - cosine_mrr@10 - cosine_map@100 model-index: - name: SentenceTransformer based on Qwen/Qwen3-Embedding-0.6B results: - task: type: information-retrieval name: Information Retrieval dataset: name: val full corpus type: val_full_corpus metrics: - type: cosine_accuracy@1 value: 0.9029126213592233 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.9854368932038835 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.9902912621359223 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.9951456310679612 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.9029126213592233 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.3284789644012946 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.19805825242718444 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.09951456310679611 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.9029126213592233 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.9854368932038835 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.9902912621359223 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.9951456310679612 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.9546600421052297 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.9410194174757281 name: Cosine Mrr@10 - type: cosine_map@100 value: 0.9410877888691372 name: Cosine Map@100 --- # SentenceTransformer based on Qwen/Qwen3-Embedding-0.6B This 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. ## Model Details ### Model Description - **Model Type:** Sentence Transformer - **Base model:** [Qwen/Qwen3-Embedding-0.6B](https://huggingface.co/Qwen/Qwen3-Embedding-0.6B) - **Maximum Sequence Length:** 32768 tokens - **Output Dimensionality:** 1024 dimensions - **Similarity Function:** Cosine Similarity ### Model Sources - **Documentation:** [Sentence Transformers Documentation](https://sbert.net) - **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers) - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers) ### Full Model Architecture ``` SentenceTransformer( (0): Transformer({'max_seq_length': 32768, 'do_lower_case': False, 'architecture': 'Qwen3Model'}) (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}) (2): Normalize() ) ``` ## Usage ### Direct Usage (Sentence Transformers) First install the Sentence Transformers library: ```bash pip install -U sentence-transformers ``` Then you can load this model and run inference. ```python from sentence_transformers import SentenceTransformer # Download from the 🤗 Hub model = SentenceTransformer("sentence_transformers_model_id") # Run inference queries = [ "Impresiones Jalisco anti-corruption questionnaire completed and signed", ] documents = [ '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', "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", "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", ] query_embeddings = model.encode_query(queries) document_embeddings = model.encode_document(documents) print(query_embeddings.shape, document_embeddings.shape) # [1, 1024] [3, 1024] # Get the similarity scores for the embeddings similarities = model.similarity(query_embeddings, document_embeddings) print(similarities) # tensor([[0.4883, 0.1104, 0.0481]], dtype=torch.bfloat16) ``` ## Evaluation ### Metrics #### Information Retrieval * Dataset: `val_full_corpus` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.9029 | | cosine_accuracy@3 | 0.9854 | | cosine_accuracy@5 | 0.9903 | | cosine_accuracy@10 | 0.9951 | | cosine_precision@1 | 0.9029 | | cosine_precision@3 | 0.3285 | | cosine_precision@5 | 0.1981 | | cosine_precision@10 | 0.0995 | | cosine_recall@1 | 0.9029 | | cosine_recall@3 | 0.9854 | | cosine_recall@5 | 0.9903 | | cosine_recall@10 | 0.9951 | | **cosine_ndcg@10** | **0.9547** | | cosine_mrr@10 | 0.941 | | cosine_map@100 | 0.9411 | ## Training Details ### Training Dataset #### Unnamed Dataset * Size: 816 training samples * Columns: sentence_0 and sentence_1 * Approximate statistics based on the first 816 samples: | | sentence_0 | sentence_1 | |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------| | type | string | string | | details | | | * Samples: | sentence_0 | sentence_1 | |:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | 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. | Subject: Impairment Assessment: Noted Discrepancies in Asset Valuation
Date: 2025-12-15T17:18:00
From: Lisa Park
Participants: Maria Santos

Body:
Hi Maria,

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?

Thank you for your attention to these details. Please let me know if you need further clarification on the items I flagged.

Best regards,
Lisa

--
Lisa Park
Director, Internal Audit
Agave Spirits International
| | 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. | Subject: Re: Approval Required: Journal Entry Review and Documentation
Date: 2025-09-30T08:40:00
From: Maria Santos
Participants: David Chen

Body:
Hi David,

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.

Thanks for highlighting this and ensuring we stay audit-ready. I’ll follow up shortly with all the documentation.

Best regards,
Maria
| | Identify records confirming pre-clearance and FARA registration for ASI's government liaison activities in Mexico and the United States. | Subject: Análisis sobre el registro FARA y obligaciones de cumplimiento en actividades gubernamentales México-EE.UU.
Date: 2025-08-04T06:58:00
From: Amanda Foster
Participants: Sarah Mitchell

Body:
Hola Sarah,

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.

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...
| * Loss: [MultipleNegativesRankingLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters: ```json { "scale": 20.0, "similarity_fct": "cos_sim", "gather_across_devices": false } ``` ### Training Hyperparameters #### Non-Default Hyperparameters - `per_device_train_batch_size`: 16 - `per_device_eval_batch_size`: 16 - `multi_dataset_batch_sampler`: round_robin #### All Hyperparameters
Click to expand - `do_predict`: False - `eval_strategy`: no - `prediction_loss_only`: True - `per_device_train_batch_size`: 16 - `per_device_eval_batch_size`: 16 - `gradient_accumulation_steps`: 1 - `eval_accumulation_steps`: None - `torch_empty_cache_steps`: None - `learning_rate`: 5e-05 - `weight_decay`: 0.0 - `adam_beta1`: 0.9 - `adam_beta2`: 0.999 - `adam_epsilon`: 1e-08 - `max_grad_norm`: 1 - `num_train_epochs`: 3 - `max_steps`: -1 - `lr_scheduler_type`: linear - `lr_scheduler_kwargs`: None - `warmup_ratio`: None - `warmup_steps`: 0 - `log_level`: passive - `log_level_replica`: warning - `log_on_each_node`: True - `logging_nan_inf_filter`: True - `enable_jit_checkpoint`: False - `save_on_each_node`: False - `save_only_model`: False - `restore_callback_states_from_checkpoint`: False - `use_cpu`: False - `seed`: 42 - `data_seed`: None - `bf16`: False - `fp16`: False - `bf16_full_eval`: False - `fp16_full_eval`: False - `tf32`: None - `local_rank`: -1 - `ddp_backend`: None - `debug`: [] - `dataloader_drop_last`: False - `dataloader_num_workers`: 0 - `dataloader_prefetch_factor`: None - `disable_tqdm`: False - `remove_unused_columns`: True - `label_names`: None - `load_best_model_at_end`: False - `ignore_data_skip`: False - `fsdp`: [] - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False} - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None} - `parallelism_config`: None - `deepspeed`: None - `label_smoothing_factor`: 0.0 - `optim`: adamw_torch_fused - `optim_args`: None - `group_by_length`: False - `length_column_name`: length - `project`: huggingface - `trackio_space_id`: trackio - `ddp_find_unused_parameters`: None - `ddp_bucket_cap_mb`: None - `ddp_broadcast_buffers`: False - `dataloader_pin_memory`: True - `dataloader_persistent_workers`: False - `skip_memory_metrics`: True - `push_to_hub`: False - `resume_from_checkpoint`: None - `hub_model_id`: None - `hub_strategy`: every_save - `hub_private_repo`: None - `hub_always_push`: False - `hub_revision`: None - `gradient_checkpointing`: False - `gradient_checkpointing_kwargs`: None - `include_for_metrics`: [] - `eval_do_concat_batches`: True - `auto_find_batch_size`: False - `full_determinism`: False - `ddp_timeout`: 1800 - `torch_compile`: False - `torch_compile_backend`: None - `torch_compile_mode`: None - `include_num_input_tokens_seen`: no - `neftune_noise_alpha`: None - `optim_target_modules`: None - `batch_eval_metrics`: False - `eval_on_start`: False - `use_liger_kernel`: False - `liger_kernel_config`: None - `eval_use_gather_object`: False - `average_tokens_across_devices`: True - `use_cache`: False - `prompts`: None - `batch_sampler`: batch_sampler - `multi_dataset_batch_sampler`: round_robin - `router_mapping`: {} - `learning_rate_mapping`: {}
### Training Logs | Epoch | Step | val_full_corpus_cosine_ndcg@10 | |:-----:|:----:|:------------------------------:| | 1.0 | 51 | 0.9547 | ### Framework Versions - Python: 3.12.12 - Sentence Transformers: 5.2.3 - Transformers: 5.0.0 - PyTorch: 2.10.0+cu128 - Accelerate: 1.12.0 - Datasets: 4.0.0 - Tokenizers: 0.22.2 ## Citation ### BibTeX #### Sentence Transformers ```bibtex @inproceedings{reimers-2019-sentence-bert, title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks", author = "Reimers, Nils and Gurevych, Iryna", booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing", month = "11", year = "2019", publisher = "Association for Computational Linguistics", url = "https://arxiv.org/abs/1908.10084", } ``` #### MultipleNegativesRankingLoss ```bibtex @misc{henderson2017efficient, title={Efficient Natural Language Response Suggestion for Smart Reply}, 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}, year={2017}, eprint={1705.00652}, archivePrefix={arXiv}, primaryClass={cs.CL} } ```