TeraflopAI/teraflopai-jina-nano-caselaw

SentenceTransformer

This is a sentence-transformers model trained on the parquet dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for retrieval.

Model Details

Model Description

• Model Type: Sentence Transformer
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity
• Supported Modality: Text
• Training Dataset:
  • parquet

Model Sources

• Documentation: Sentence Transformers Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'transformer_task': 'feature-extraction', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'last_hidden_state'}}, 'module_output_name': 'token_embeddings', 'architecture': 'EuroBertModel'})
  (1): Pooling({'embedding_dimension': 768, 'pooling_mode': 'lasttoken', 'include_prompt': True})
  (2): Normalize({})
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("TeraflopAI/teraflopai-jina-nano-legal")
# Run inference
queries = [
    "Under what circumstances can a plaintiff recover damages for lost profits when a defendant's breach of contract involves the failure to provide telephone service?",
]
documents = [
    'UNITED STATES COURT OF APPEALS\n                            UNITED STATES COURT OF APPEALS\n                    FOR THE FIRST CIRCUIT\n                                FOR THE FIRST CIRCUIT\n                                         \nNo. 94-1711\n\n                  SAS OF PUERTO RICO, INC.,\n\n                    Plaintiff, Appellant,\n\n                              v.\n\n                PUERTO RICO TELEPHONE COMPANY,\n\n                     Defendant, Appellee. \n\n                                         \n\n         APPEAL FROM THE UNITED STATES DISTRICT COURT\n\n               FOR THE DISTRICT OF PUERTO RICO\n\n        [Hon. Jose Antonio Fuste, U.S. District Judge]\n                                                                 \n\n                                         \n\n                            Before\n\n                    Torruella, Chief Judge,\n                                                      \n\n                    Boudin, Circuit Judge,\n                                                     \n\n              and Boyle,* Senior District Judge.\n                                                           \n\n                                         \n\nLaurence  Z.  Shiekman  with  whom  M.  Duncan  Grant,  Frank   M.\n                                                                              \nRapoport,  Michael A. Ceramella and Pepper, Hamilton & Scheetz were on\n                                                                      \nbrief for appellant.\nPhilip J. Mause with whom Joaquin A. Marquez and Drinker Biddle  &\n                                                                              \nReath were on brief for appellee.\n             \n\n                                         \n\n                      February 21, 1995\n                                         \n\n                \n\n*Of the District of Rhode Island, sitting by designation.',
    'Nor did the BIA abuse its discretion by denying Chen\'s motion to reopen, which alleged that he suffered from the ineffective assistance of counsel. To prevail on such a claim, the alien must first comply with certain procedures set forth in Matter of Lozada, 19 I. & N. Dec. 637 (BIA 1988). Here, the BIA properly noted that besides filing a supporting affidavit, Chen made no effort to comply with the requirements enumerated in Lozada. Chen not only failed to notify his former counsel of the allegations of ineffective assistance and to allow him an opportunity to respond, he also failed to file a complaint with a disciplinary authority or provide an explanation for not doing so. See Twum v. INS, 411 F.3d 54, 59 (2d Cir.2005) (citing Lozada, 19 I. & N. Dec. at 639). \n\nBy failing to substantially comply with Lozada, Chen "forfeit[ed][his] ineffective assistance of counsel claim." Jian Yun Zheng v. U.S. Dep\'t of Justice, 409 F.3d 43, 47 (2d Cir.2005). While it is true that "slavish adherence" to Lozada\'s requirements is not necessary in certain circumstances, and while the BIA acknowledged in its June 2006 decision that the brief written by Chen\'s former counsel was deficient, this is not a case in which the facts supporting a "claim of ineffective assistance are clear on the face of the record," which may excuse the failure to comply with Lozada. Yi Long Yang, 478 F.3d at 142-43. The facts here are distinct from the circumstances presented in Yi Long Yang, in that Chen\'s former counsel was not disbarred, nor was there evidence that the agency explicitly assumed his competence. See id. at 142.',
    'We believe the same prerequisite should operate in this\ncase. The requirement that parties seeking Rule 60(b) relief\nshow some prospect of succeeding on the merits flows from\nthe basic principle that courts should revive previously-\ndismissed claims only if they have some reason to believe that\ndoing so will not ultimately waste judicial resources. See\nMurray,\n52 F.3d at 355\n. This principle holds true here:\nreviving Thomas\'s appeal will constitute an "empty exercise\nor futile gesture,"\nid.,\n unless Thomas has some possibility of\nprevailing. \n\n       Indeed, we see two especially good reasons to condition\nthe grant of Thomas\'s motion for reconsideration on his\ndemonstrating a chance of succeeding on the merits. First,\nThomas claims that his appeal should be reinstated because\nthe PLRA\'s three-strikes provision is unconstitutional as\napplied to him. For this court to reach out and decide this\ndifficult and important question simply to reinstate a pointless\nappeal would violate the norm of constitutional avoidance to\nwhich we generally adhere. See Kalka v. Hawk,\n215 F.3d 90, 97\n (D.C. Cir. 2000) ("Federal courts should not decide\nconstitutional questions unless it is necessary to do so.").\nSecond, the PLRA provides that a court "shall dismiss" an\nIFP litigant\'s case if the "appeal . . . is frivolous or malicious\n. . . [or] fails to state a claim on which relief may be granted."\n28 U.S.C. § 1915\n(e)(2). Thus, even were we to grant Thomas\nIFP status and reinstate his appeal, we would then have to\n\x0c                                7\npromptly dismiss the case if his claims lack merit. What could\nbe a more "futile gesture" than reinstating an appeal only to\nthen immediately dismiss it?',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 768] [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[ 0.6921, -0.0759,  0.0219]])

Evaluation

Metrics

Information Retrieval

• Dataset: jina-nano_lr5e-5_warmup0.1_bs8k-caselaw
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.8919
cosine_accuracy@3	0.9569
cosine_accuracy@5	0.9727
cosine_accuracy@10	0.9853
cosine_precision@1	0.8919
cosine_precision@3	0.319
cosine_precision@5	0.1945
cosine_precision@10	0.0985
cosine_recall@1	0.8919
cosine_recall@3	0.9569
cosine_recall@5	0.9727
cosine_recall@10	0.9853
cosine_ndcg@10	0.9413
cosine_mrr@10	0.9269
cosine_map@100	0.9276

Training Details

Training Dataset

parquet

• Dataset: parquet

• Size: 36,118,859 training samples

• Columns: question and answer

• Approximate statistics based on the first 1000 samples:

  	question	answer
  type	string	string
  details	min: 15 tokens mean: 28.74 tokens max: 50 tokens	min: 40 tokens mean: 308.21 tokens max: 512 tokens

• Samples:

  question	answer
  What is the legal standard and procedure for granting a defendant's request for leave to withdraw as counsel when the attorney certifies that no nonfrivolous issues exist for appeal?	Appeal by the defendant from two judgments of the Supreme Court, Queens County (Rosengarten, J.), both rendered May 5, 2003, convicting him of burglary in the first degree, robbery in the first degree, and burglary in the second degree under indictment No. 3417/01, and burglary in the first degree, robbery in the first degree, and burglary in the second degree under indictment No. 1182/02, upon his pleas of guilty, and imposing sentences. Ordered that the judgments are affirmed. We have reviewed the record and agree with the defendant's assigned counsel that there are no nonfrivolous issues which could be raised on appeal. Counsel's application for leave to withdraw as counsel is granted (see Anders v California, 386 US 738 [1967]; People v Paige, 54 AD2d 631 [1976]; cf. People v *606Gonzalez, 47 NY2d 606 [1979]). Adams, J.P., Cozier, Ritter and Skelos, JJ., concur.
  Are state-law tort claims alleging defective labeling of generic drugs preempted by federal law?	ORDER

  JOSEPH N. LAPLANTE, District Judge.

  This case presents a question currently pending before three different federal courts of appeal: whether state-law tort claims alleging the defective labeling of generic drugs are preempted by federal law. See Morris v. Wyeth, Inc., No. 09-5509 (6th Cir. Apr. 27, 2009); Demahy v. Wyeth, Inc., No. 08-31204 (5th Cir. Dec. 16, 2008); Mensing v. Wyeth, Inc., No. 08-3850 (8th Cir. Dec. 10, 2008). The defendants, Mutual Pharmaceutical Company, Inc. and United Research Laboratories, Inc., move for judgment on the pleadings, see Fed.R.Civ.P. 12(c), on claims by the plaintiffs, Karen L. and Gregory S. Bartlett, alleging that Karen suffered serious injuries from Sulindac, a generic drug manufactured by the defendants. The defendants argue that all of the plaintiffs' state-law causes of action are pre-empted by Title I of the Drug Price Competition and Patent Term Restoration Act of 1984, 1 part of the Hatch-Waxman Amendments to the Federal Food, Drug,... | | Under what circumstances can a trial court's finding of competency to stand trial be challenged when multiple expert evaluations conclude the defendant is competent but exhibit bizarre conduct? | Prior to trial, Card was examined by two court-appointed psychologists for the purpose of determining whether he was competent to stand trial. Following examinations, both psychologists concluded that Card was competent to stand trial pursuant to the criteria set forth in Rule 3.211, Florida Rule of Criminal Procedure. After the initial reports of the two court-appointed *1175 experts were filed, the defense filed a motion for the appointment of a forensic psychiatrist to examine Card. The court acquiesced to this request. Although the forensic psychiatrist did not file his report with the court until a few months after the court issued its order finding Card competent to stand trial, the forensic psychiatrist also concluded that Card was competent. Further, although the various reports filed by the experts indicate bizarre conduct and behavioral problems, the trial court was never presented with evidence providing reasonable grounds to believe that Card was not competent to stand tria... |

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "CachedMultipleNegativesRankingLoss",
      "matryoshka_dims": [
          768,
          512,
          256,
          128
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Evaluation Dataset

parquet

• Dataset: parquet
• Size: 10,000 evaluation samples
• Columns: question and answer
• Approximate statistics based on the first 1000 samples:

  	question	answer
  type	string	string
  details	min: 14 tokens mean: 28.88 tokens max: 52 tokens	min: 67 tokens mean: 312.84 tokens max: 512 tokens

• Samples:

  question	answer
  What specific factors do Texas courts consider when determining if terminating a parent's rights serves the child's best interest?	In determining whether termination is in the child's best interest, we apply the following factors laid out in Holley v. Adams, 544 S.W.2d 367, 371–72 (Tex. 1976). Those factors include, but are not limited to:           1.       The child's desires;             2.       The child's physical and emotional needs, now and in the future;             3.       The emotional and physical danger to the child, now and in the future;             4.       The parental ability of the individuals seeking custody;             5.       The programs available to assist these individuals in promoting the child's best interest;             6.       The plans for the child by the individual or agency seeking custody;             7.       The stability of the home or proposed placement;             8.       The parent's act or omissions that may indicate the existing parent-child relationship is not the proper one; and             9.       Any excuse for the parent's acts or omissions.
  Under what circumstances do separate criminal acts fail to constitute a single continuous transaction for the purpose of admitting evidence of one act to prove another?	¶37 The case before us is far more analogous to Hildreth than to the others. Gallegos stabbed Victim in a park and was later apprehended. Then, while at the police station, Gallegos acted violently, resulting in additional charges. Gallegos's violent behavior at the police station did not "facilitate[ ] flight" from the earlier attack, nor could the later crimes be characterized as "a single [violent] spree," as we would characterize a string of robberies, for example. See Benson , 2014 UT App 92 , ¶¶ 13-14, 325 P.3d 855 . Neither do Gallegos's crimes demonstrate "a distinct behavioral arc of increasingly aggressive and opportunistic transgressions." Burke , 2011 UT App 168 , ¶ 24, 256 P.3d 1102 . Instead, this case is more like Hildreth , where the defendant committed a sequence of offenses, but those offenses were not otherwise related to each other. See 2010 UT App 209 , ¶ 32, 238 P.3d 444 . Here, the stabbing at the park and the violent behavior at the police station are so indepen...
  What level of culpability, such as actual knowledge or reckless disregard, must a plaintiff prove to establish an Eighth Amendment violation for deliberate indifference?	Wilson v. Seiter, ___ U.S. at -, ___, 111 S.Ct. at 2324-25, 2327.

  The Seventh Circuit recently observed that "[i]n order to show `deliberate indifference,' a plaintiff is required to prove that the prison official's action was deliberate or reckless in the criminal sense." Santiago v. Lane, 894 F.2d 218 (7th Cir.1990) (emphasis added) (footnote omitted). The United States Supreme Court has cited the Seventh Circuit's criminal recklessness standard with approval. Whitley v. Albers, 475 U.S. 312, 321, 106 S.Ct. 1078, 1085, 89 L.Ed.2d 251 (1986), citing Duckworth v. Franzen, 780 F.2d 645, 653 (7th Cir.1985), cert. denied, 479 U.S. 816, 107 S.Ct. 71, 93 L.Ed.2d 28 (1986). In Franzen, the Seventh Circuit noted that punishment under the Eighth Amendment "implies at a minimum actual knowledge of impending harm easily preventable, so that a conscious, culpable refusal to prevent the harm can be inferred from the defendant's failure to prevent it." 780 F.2d at 653. See also Wilks v. You... |
• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "CachedMultipleNegativesRankingLoss",
      "matryoshka_dims": [
          768,
          512,
          256,
          128
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 2048
• num_train_epochs: 1
• lr_scheduler_type: cosine
• warmup_steps: 0.1
• bf16: True
• per_device_eval_batch_size: 512
• prompts: {'question': 'Query: ', 'answer': 'Document: '}
• batch_sampler: no_duplicates

All Hyperparameters

Click to expand

• per_device_train_batch_size: 2048
• num_train_epochs: 1
• max_steps: -1
• learning_rate: 5e-05
• lr_scheduler_type: cosine
• lr_scheduler_kwargs: None
• warmup_steps: 0.1
• optim: adamw_torch_fused
• optim_args: None
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• optim_target_modules: None
• gradient_accumulation_steps: 1
• average_tokens_across_devices: True
• max_grad_norm: 1.0
• label_smoothing_factor: 0.0
• bf16: True
• fp16: False
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• use_liger_kernel: False
• liger_kernel_config: None
• use_cache: False
• neftune_noise_alpha: None
• torch_empty_cache_steps: None
• auto_find_batch_size: False
• log_on_each_node: True
• logging_nan_inf_filter: True
• include_num_input_tokens_seen: no
• log_level: passive
• log_level_replica: warning
• disable_tqdm: False
• project: huggingface
• trackio_space_id: None
• trackio_bucket_id: None
• trackio_static_space_id: None
• per_device_eval_batch_size: 512
• prediction_loss_only: True
• eval_on_start: False
• eval_do_concat_batches: True
• eval_use_gather_object: False
• eval_accumulation_steps: None
• include_for_metrics: []
• batch_eval_metrics: False
• save_only_model: False
• save_on_each_node: False
• enable_jit_checkpoint: False
• push_to_hub: False
• hub_private_repo: None
• hub_model_id: None
• hub_strategy: every_save
• hub_always_push: False
• hub_revision: None
• load_best_model_at_end: False
• ignore_data_skip: False
• restore_callback_states_from_checkpoint: False
• full_determinism: False
• seed: 42
• data_seed: None
• use_cpu: 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
• dataloader_drop_last: True
• dataloader_num_workers: 0
• dataloader_pin_memory: True
• dataloader_persistent_workers: False
• dataloader_prefetch_factor: None
• remove_unused_columns: True
• label_names: None
• train_sampling_strategy: random
• length_column_name: length
• ddp_find_unused_parameters: None
• ddp_bucket_cap_mb: None
• ddp_broadcast_buffers: False
• ddp_static_graph: None
• ddp_backend: None
• ddp_timeout: 1800
• fsdp: []
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• deepspeed: None
• debug: []
• skip_memory_metrics: True
• do_predict: False
• resume_from_checkpoint: None
• warmup_ratio: None
• local_rank: -1
• prompts: {'question': 'Query: ', 'answer': 'Document: '}
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Click to expand

Epoch	Step	Training Loss	Validation Loss	jina-nano_lr5e-5_warmup0.1_bs8k-caselaw_cosine_ndcg@10
0.0023	10	5.4249	-	-
0.0045	20	5.3218	-	-
0.0068	30	4.9505	-	-
0.0091	40	4.4023	-	-
0.0113	50	3.3952	-	-
0.0136	60	2.6740	-	-
0.0159	70	2.1146	-	-
0.0181	80	1.8614	-	-
0.0204	90	1.6447	-	-
0.0227	100	1.5237	-	-
0.0249	110	1.4115	-	-
0.0272	120	1.3367	-	-
0.0295	130	1.2759	-	-
0.0318	140	1.2270	-	-
0.0340	150	1.1613	-	-
0.0363	160	1.1364	-	-
0.0386	170	1.0903	-	-
0.0408	180	1.0284	-	-
0.0431	190	1.0257	-	-
0.0454	200	0.9858	-	-
0.0476	210	0.9542	-	-
0.0499	220	0.9389	-	-
0.0522	230	0.9077	-	-
0.0544	240	0.8852	-	-
0.0567	250	0.8622	-	-
0.0590	260	0.8439	-	-
0.0612	270	0.8217	-	-
0.0635	280	0.8145	-	-
0.0658	290	0.8084	-	-
0.0680	300	0.7711	-	-
0.0703	310	0.7579	-	-
0.0726	320	0.7618	-	-
0.0748	330	0.7377	-	-
0.0771	340	0.7256	-	-
0.0794	350	0.7078	-	-
0.0817	360	0.7000	-	-
0.0839	370	0.6873	-	-
0.0862	380	0.6894	-	-
0.0885	390	0.6586	-	-
0.0907	400	0.6632	-	-
0.0930	410	0.6472	-	-
0.0953	420	0.6528	-	-
0.0975	430	0.6439	-	-
0.0998	440	0.6319	-	-
0.1021	450	0.6030	-	-
0.1043	460	0.6269	-	-
0.1066	470	0.6085	-	-
0.1089	480	0.5979	-	-
0.1111	490	0.5931	-	-
0.1134	500	0.5888	-	-
0.1157	510	0.5724	-	-
0.1179	520	0.5808	-	-
0.1202	530	0.5661	-	-
0.1225	540	0.5597	-	-
0.1247	550	0.5463	-	-
0.1270	560	0.5348	-	-
0.1293	570	0.5364	-	-
0.1315	580	0.5389	-	-
0.1338	590	0.5351	-	-
0.1361	600	0.5199	-	-
0.1384	610	0.5188	-	-
0.1406	620	0.5135	-	-
0.1429	630	0.5083	-	-
0.1452	640	0.4999	-	-
0.1474	650	0.4986	-	-
0.1497	660	0.4980	-	-
0.1520	670	0.4951	-	-
0.1542	680	0.4930	-	-
0.1565	690	0.4842	-	-
0.1588	700	0.4848	-	-
0.1610	710	0.4835	-	-
0.1633	720	0.4846	-	-
0.1656	730	0.4760	-	-
0.1678	740	0.4719	-	-
0.1701	750	0.4746	-	-
0.1724	760	0.4576	-	-
0.1746	770	0.4709	-	-
0.1769	780	0.4430	-	-
0.1792	790	0.4496	-	-
0.1814	800	0.4570	-	-
0.1837	810	0.4452	-	-
0.1860	820	0.4495	-	-
0.1883	830	0.4468	-	-
0.1905	840	0.4468	-	-
0.1928	850	0.4446	-	-
0.1951	860	0.4372	-	-
0.1973	870	0.4376	-	-
0.1996	880	0.4287	-	-
0.2019	890	0.4268	-	-
0.2041	900	0.4239	-	-
0.2064	910	0.4274	-	-
0.2087	920	0.4234	-	-
0.2109	930	0.4190	-	-
0.2132	940	0.4260	-	-
0.2155	950	0.4247	-	-
0.2177	960	0.4137	-	-
0.2200	970	0.4141	-	-
0.2223	980	0.4114	-	-
0.2245	990	0.4137	-	-
0.2268	1000	0.4088	-	-
0.2291	1010	0.3989	-	-
0.2313	1020	0.4162	-	-
0.2336	1030	0.4048	-	-
0.2359	1040	0.4026	-	-
0.2381	1050	0.4044	-	-
0.2404	1060	0.4017	-	-
0.2427	1070	0.4064	-	-
0.2450	1080	0.3970	-	-
0.2472	1090	0.3985	-	-
0.2495	1100	0.3884	-	-
0.2518	1110	0.3841	-	-
0.2540	1120	0.3960	-	-
0.2563	1130	0.3964	-	-
0.2586	1140	0.3910	-	-
0.2608	1150	0.3892	-	-
0.2631	1160	0.3839	-	-
0.2654	1170	0.3786	-	-
0.2676	1180	0.3800	-	-
0.2699	1190	0.3865	-	-
0.2722	1200	0.3812	-	-
0.2744	1210	0.3748	-	-
0.2767	1220	0.3823	-	-
0.2790	1230	0.3771	-	-
0.2812	1240	0.3732	-	-
0.2835	1250	0.3707	-	-
0.2858	1260	0.3709	-	-
0.2880	1270	0.3689	-	-
0.2903	1280	0.3793	-	-
0.2926	1290	0.3704	-	-
0.2949	1300	0.3697	-	-
0.2971	1310	0.3647	-	-
0.2994	1320	0.3687	-	-
0.3001	1323	-	0.1011	0.9300
0.3017	1330	0.3662	-	-
0.3039	1340	0.3656	-	-
0.3062	1350	0.3626	-	-
0.3085	1360	0.3625	-	-
0.3107	1370	0.3663	-	-
0.3130	1380	0.3525	-	-
0.3153	1390	0.3496	-	-
0.3175	1400	0.3588	-	-
0.3198	1410	0.3579	-	-
0.3221	1420	0.3487	-	-
0.3243	1430	0.3537	-	-
0.3266	1440	0.3521	-	-
0.3289	1450	0.3512	-	-
0.3311	1460	0.3576	-	-
0.3334	1470	0.3510	-	-
0.3357	1480	0.3467	-	-
0.3379	1490	0.3488	-	-
0.3402	1500	0.3410	-	-
0.3425	1510	0.3425	-	-
0.3447	1520	0.3558	-	-
0.3470	1530	0.3483	-	-
0.3493	1540	0.3490	-	-
0.3516	1550	0.3435	-	-
0.3538	1560	0.3420	-	-
0.3561	1570	0.3363	-	-
0.3584	1580	0.3470	-	-
0.3606	1590	0.3415	-	-
0.3629	1600	0.3491	-	-
0.3652	1610	0.3386	-	-
0.3674	1620	0.3389	-	-
0.3697	1630	0.3334	-	-
0.3720	1640	0.3403	-	-
0.3742	1650	0.3330	-	-
0.3765	1660	0.3359	-	-
0.3788	1670	0.3319	-	-
0.3810	1680	0.3359	-	-
0.3833	1690	0.3354	-	-
0.3856	1700	0.3308	-	-
0.3878	1710	0.3359	-	-
0.3901	1720	0.3312	-	-
0.3924	1730	0.3374	-	-
0.3946	1740	0.3260	-	-
0.3969	1750	0.3307	-	-
0.3992	1760	0.3269	-	-
0.4015	1770	0.3285	-	-
0.4037	1780	0.3296	-	-
0.4060	1790	0.3332	-	-
0.4083	1800	0.3350	-	-
0.4105	1810	0.3291	-	-
0.4128	1820	0.3253	-	-
0.4151	1830	0.3256	-	-
0.4173	1840	0.3274	-	-
0.4196	1850	0.3200	-	-
0.4219	1860	0.3147	-	-
0.4241	1870	0.3221	-	-
0.4264	1880	0.3194	-	-
0.4287	1890	0.3250	-	-
0.4309	1900	0.3249	-	-
0.4332	1910	0.3159	-	-
0.4355	1920	0.3291	-	-
0.4377	1930	0.3179	-	-
0.4400	1940	0.3256	-	-
0.4423	1950	0.3223	-	-
0.4445	1960	0.3182	-	-
0.4468	1970	0.3120	-	-
0.4491	1980	0.3151	-	-
0.4513	1990	0.3162	-	-
0.4536	2000	0.3090	-	-
0.4559	2010	0.3098	-	-
0.4582	2020	0.3136	-	-
0.4604	2030	0.3171	-	-
0.4627	2040	0.3138	-	-
0.4650	2050	0.3142	-	-
0.4672	2060	0.3118	-	-
0.4695	2070	0.3207	-	-
0.4718	2080	0.3166	-	-
0.4740	2090	0.3205	-	-
0.4763	2100	0.3201	-	-
0.4786	2110	0.3060	-	-
0.4808	2120	0.3158	-	-
0.4831	2130	0.3118	-	-
0.4854	2140	0.3107	-	-
0.4876	2150	0.3037	-	-
0.4899	2160	0.3071	-	-
0.4922	2170	0.3126	-	-
0.4944	2180	0.3108	-	-
0.4967	2190	0.3061	-	-
0.4990	2200	0.3023	-	-
0.5012	2210	0.3134	-	-
0.5035	2220	0.3065	-	-
0.5058	2230	0.3100	-	-
0.5081	2240	0.3067	-	-
0.5103	2250	0.3087	-	-
0.5126	2260	0.3021	-	-
0.5149	2270	0.3022	-	-
0.5171	2280	0.3109	-	-
0.5194	2290	0.3072	-	-
0.5217	2300	0.3043	-	-
0.5239	2310	0.2994	-	-
0.5262	2320	0.3024	-	-
0.5285	2330	0.3041	-	-
0.5307	2340	0.3021	-	-
0.5330	2350	0.2989	-	-
0.5353	2360	0.3041	-	-
0.5375	2370	0.2980	-	-
0.5398	2380	0.2995	-	-
0.5421	2390	0.2961	-	-
0.5443	2400	0.2929	-	-
0.5466	2410	0.2972	-	-
0.5489	2420	0.3008	-	-
0.5511	2430	0.3028	-	-
0.5534	2440	0.2995	-	-
0.5557	2450	0.3066	-	-
0.5579	2460	0.3032	-	-
0.5602	2470	0.2961	-	-
0.5625	2480	0.2957	-	-
0.5648	2490	0.3023	-	-
0.5670	2500	0.3043	-	-
0.5693	2510	0.3009	-	-
0.5716	2520	0.2968	-	-
0.5738	2530	0.2969	-	-
0.5761	2540	0.2882	-	-
0.5784	2550	0.2956	-	-
0.5806	2560	0.2983	-	-
0.5829	2570	0.2951	-	-
0.5852	2580	0.2885	-	-
0.5874	2590	0.2977	-	-
0.5897	2600	0.2970	-	-
0.5920	2610	0.2925	-	-
0.5942	2620	0.2835	-	-
0.5965	2630	0.2973	-	-
0.5988	2640	0.2915	-	-
0.6001	2646	-	0.0793	0.9393
0.6010	2650	0.2957	-	-
0.6033	2660	0.2978	-	-
0.6056	2670	0.2961	-	-
0.6078	2680	0.2990	-	-
0.6101	2690	0.2980	-	-
0.6124	2700	0.2937	-	-
0.6147	2710	0.2954	-	-
0.6169	2720	0.2903	-	-
0.6192	2730	0.2992	-	-
0.6215	2740	0.2894	-	-
0.6237	2750	0.2895	-	-
0.6260	2760	0.2905	-	-
0.6283	2770	0.2907	-	-
0.6305	2780	0.2919	-	-
0.6328	2790	0.2821	-	-
0.6351	2800	0.2964	-	-
0.6373	2810	0.2917	-	-
0.6396	2820	0.2910	-	-
0.6419	2830	0.2914	-	-
0.6441	2840	0.2965	-	-
0.6464	2850	0.2905	-	-
0.6487	2860	0.2951	-	-
0.6509	2870	0.2909	-	-
0.6532	2880	0.2849	-	-
0.6555	2890	0.2879	-	-
0.6577	2900	0.2895	-	-
0.6600	2910	0.2921	-	-
0.6623	2920	0.2875	-	-
0.6645	2930	0.2876	-	-
0.6668	2940	0.2953	-	-
0.6691	2950	0.2885	-	-
0.6714	2960	0.2860	-	-
0.6736	2970	0.2850	-	-
0.6759	2980	0.2943	-	-
0.6782	2990	0.2877	-	-
0.6804	3000	0.2856	-	-
0.6827	3010	0.2951	-	-
0.6850	3020	0.2856	-	-
0.6872	3030	0.2945	-	-
0.6895	3040	0.2860	-	-
0.6918	3050	0.2917	-	-
0.6940	3060	0.2905	-	-
0.6963	3070	0.2978	-	-
0.6986	3080	0.2915	-	-
0.7008	3090	0.2860	-	-
0.7031	3100	0.2864	-	-
0.7054	3110	0.2920	-	-
0.7076	3120	0.2813	-	-
0.7099	3130	0.2853	-	-
0.7122	3140	0.2827	-	-
0.7144	3150	0.2865	-	-
0.7167	3160	0.2829	-	-
0.7190	3170	0.2816	-	-
0.7213	3180	0.2835	-	-
0.7235	3190	0.2866	-	-
0.7258	3200	0.2846	-	-
0.7281	3210	0.2835	-	-
0.7303	3220	0.2874	-	-
0.7326	3230	0.2808	-	-
0.7349	3240	0.2898	-	-
0.7371	3250	0.2850	-	-
0.7394	3260	0.2912	-	-
0.7417	3270	0.2878	-	-
0.7439	3280	0.2880	-	-
0.7462	3290	0.2932	-	-
0.7485	3300	0.2818	-	-
0.7507	3310	0.2840	-	-
0.7530	3320	0.2833	-	-
0.7553	3330	0.2868	-	-
0.7575	3340	0.2871	-	-
0.7598	3350	0.2825	-	-
0.7621	3360	0.2914	-	-
0.7643	3370	0.2899	-	-
0.7666	3380	0.2845	-	-
0.7689	3390	0.2875	-	-
0.7711	3400	0.2921	-	-
0.7734	3410	0.2823	-	-
0.7757	3420	0.2869	-	-
0.7780	3430	0.2909	-	-
0.7802	3440	0.2818	-	-
0.7825	3450	0.2902	-	-
0.7848	3460	0.2845	-	-
0.7870	3470	0.2838	-	-
0.7893	3480	0.2889	-	-
0.7916	3490	0.2831	-	-
0.7938	3500	0.2898	-	-
0.7961	3510	0.2846	-	-
0.7984	3520	0.2904	-	-
0.8006	3530	0.2870	-	-
0.8029	3540	0.2733	-	-
0.8052	3550	0.2823	-	-
0.8074	3560	0.2805	-	-
0.8097	3570	0.2847	-	-
0.8120	3580	0.2785	-	-
0.8142	3590	0.2895	-	-
0.8165	3600	0.2793	-	-
0.8188	3610	0.2875	-	-
0.8210	3620	0.2883	-	-
0.8233	3630	0.2839	-	-
0.8256	3640	0.2833	-	-
0.8279	3650	0.2841	-	-
0.8301	3660	0.2867	-	-
0.8324	3670	0.2882	-	-
0.8347	3680	0.2772	-	-
0.8369	3690	0.2921	-	-
0.8392	3700	0.2866	-	-
0.8415	3710	0.2773	-	-
0.8437	3720	0.2826	-	-
0.8460	3730	0.2802	-	-
0.8483	3740	0.2847	-	-
0.8505	3750	0.2828	-	-
0.8528	3760	0.2926	-	-
0.8551	3770	0.2869	-	-
0.8573	3780	0.2807	-	-
0.8596	3790	0.2897	-	-
0.8619	3800	0.2833	-	-
0.8641	3810	0.2856	-	-
0.8664	3820	0.2880	-	-
0.8687	3830	0.2913	-	-
0.8709	3840	0.2857	-	-
0.8732	3850	0.2864	-	-
0.8755	3860	0.2796	-	-
0.8778	3870	0.2796	-	-
0.8800	3880	0.2850	-	-
0.8823	3890	0.2863	-	-
0.8846	3900	0.2912	-	-
0.8868	3910	0.2837	-	-
0.8891	3920	0.2907	-	-
0.8914	3930	0.2849	-	-
0.8936	3940	0.2822	-	-
0.8959	3950	0.2839	-	-
0.8982	3960	0.2810	-	-
0.9002	3969	-	0.0754	0.9412
0.9004	3970	0.2905	-	-
0.9027	3980	0.2904	-	-
0.9050	3990	0.2768	-	-
0.9072	4000	0.2836	-	-
0.9095	4010	0.2885	-	-
0.9118	4020	0.2873	-	-
0.9140	4030	0.2857	-	-
0.9163	4040	0.2850	-	-
0.9186	4050	0.2828	-	-
0.9208	4060	0.2848	-	-
0.9231	4070	0.2858	-	-
0.9254	4080	0.2841	-	-
0.9276	4090	0.2792	-	-
0.9299	4100	0.2784	-	-
0.9322	4110	0.2845	-	-
0.9345	4120	0.2869	-	-
0.9367	4130	0.2810	-	-
0.9390	4140	0.2841	-	-
0.9413	4150	0.2788	-	-
0.9435	4160	0.2938	-	-
0.9458	4170	0.2870	-	-
0.9481	4180	0.2785	-	-
0.9503	4190	0.2795	-	-
0.9526	4200	0.2863	-	-
0.9549	4210	0.2869	-	-
0.9571	4220	0.2815	-	-
0.9594	4230	0.2842	-	-
0.9617	4240	0.2828	-	-
0.9639	4250	0.2822	-	-
0.9662	4260	0.2806	-	-
0.9685	4270	0.2820	-	-
0.9707	4280	0.2858	-	-
0.9730	4290	0.2810	-	-
0.9753	4300	0.2864	-	-
0.9775	4310	0.2790	-	-
0.9798	4320	0.2803	-	-
0.9821	4330	0.2815	-	-
0.9844	4340	0.2767	-	-
0.9866	4350	0.2899	-	-
0.9889	4360	0.2839	-	-
0.9912	4370	0.2831	-	-
0.9934	4380	0.2911	-	-
0.9957	4390	0.2913	-	-
0.9980	4400	0.2834	-	-
1.0	4409	-	0.0753	0.9413
-1	-1	-	-	0.9413

Training Time

• Training: 11.6 hours
• Evaluation: 21.5 minutes
• Total: 11.9 hours

Framework Versions

• Python: 3.12.13
• Sentence Transformers: 5.4.1
• Transformers: 5.8.0
• PyTorch: 2.11.0+cu130
• Accelerate: 1.13.0
• Datasets: 4.8.5
• Tokenizers: 0.22.2

Citation

BibTeX

Sentence Transformers

@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",
}

MatryoshkaLoss

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    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},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

CachedMultipleNegativesRankingLoss

@misc{gao2021scaling,
    title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
    author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan},
    year={2021},
    eprint={2101.06983},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}