bansalaman18/reranker-msmarco-v1.1-ettin-encoder-400m-listnet

CrossEncoder based on jhu-clsp/ettin-encoder-400m

This is a Cross Encoder model finetuned from jhu-clsp/ettin-encoder-400m on the ms_marco dataset using the sentence-transformers library. It computes scores for pairs of texts, which can be used for text reranking and semantic search.

Model Details

Model Description

• Model Type: Cross Encoder
• Base model: jhu-clsp/ettin-encoder-400m
• Maximum Sequence Length: 7999 tokens
• Number of Output Labels: 1 label
• Training Dataset:
  • ms_marco
• Language: en

Model Sources

• Documentation: Sentence Transformers Documentation
• Documentation: Cross Encoder Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Cross Encoders on Hugging Face

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 CrossEncoder

# Download from the 🤗 Hub
model = CrossEncoder("bansalaman18/reranker-msmarco-v1.1-ettin-encoder-400m-listnet")
# Get scores for pairs of texts
pairs = [
    ['equal rights amendment definition quizlet', 'Equal Rights Amendment. A constitutional amendment originally introduced in Congress in 1923 and passed by Congress in 1972, stating that equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex.. '],
    ['equal rights amendment definition quizlet', 'The Equal Rights Amendment (ERA) was a proposed amendment to the United States Constitution designed to guarantee equal rights for women. The ERA was originally written by Alice Paul and Crystal Eastman. In 1923, it was introduced in the Congress for the first time. Oregon-Equality of rights under the law shall not be denied or abridged by the state of Oregon or by any political subdivision in this state on account of sex.'],
    ['equal rights amendment definition quizlet', 'The legal analysis for this strategy is outlined in The Equal Rights Amendment: Why the ERA Remains Legally Viable and Properly Before the States, an article by Allison Held, Sheryl Herndon, and Danielle Stager in the Spring 1997 issue of William & Mary Journal of Women and the Law. Alice Paul rewrote the ERA in 1943 to what is now called the Alice Paul Amendment, reflecting the 15th and the 19th Amendments: Equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex..'],
    ['equal rights amendment definition quizlet', "The Equal Rights Amendment is not yet in the U.S. Constitution. The ERA, affirming the equal application of the Constitution to all persons regardless of their sex, was written in 1923 by Alice Paul, suffragist leader and founder of the National Woman's Party. Section 1. Equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex. Section 2. The Congress shall have the power to enforce, by appropriate legislation, the provisions of this article."],
    ['equal rights amendment definition quizlet', '1 Section 1. Equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex. 2  Section 2. 3  The Congress shall have the power to enforce, by appropriate legislation, the provisions of this article. 4  Section 3. Alice Paul rewrote the ERA in 1943 to what is now called the Alice Paul Amendment, reflecting the 15th and the 19th Amendments: Equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex..'],
]
scores = model.predict(pairs)
print(scores.shape)
# (5,)

# Or rank different texts based on similarity to a single text
ranks = model.rank(
    'equal rights amendment definition quizlet',
    [
        'Equal Rights Amendment. A constitutional amendment originally introduced in Congress in 1923 and passed by Congress in 1972, stating that equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex.. ',
        'The Equal Rights Amendment (ERA) was a proposed amendment to the United States Constitution designed to guarantee equal rights for women. The ERA was originally written by Alice Paul and Crystal Eastman. In 1923, it was introduced in the Congress for the first time. Oregon-Equality of rights under the law shall not be denied or abridged by the state of Oregon or by any political subdivision in this state on account of sex.',
        'The legal analysis for this strategy is outlined in The Equal Rights Amendment: Why the ERA Remains Legally Viable and Properly Before the States, an article by Allison Held, Sheryl Herndon, and Danielle Stager in the Spring 1997 issue of William & Mary Journal of Women and the Law. Alice Paul rewrote the ERA in 1943 to what is now called the Alice Paul Amendment, reflecting the 15th and the 19th Amendments: Equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex..',
        "The Equal Rights Amendment is not yet in the U.S. Constitution. The ERA, affirming the equal application of the Constitution to all persons regardless of their sex, was written in 1923 by Alice Paul, suffragist leader and founder of the National Woman's Party. Section 1. Equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex. Section 2. The Congress shall have the power to enforce, by appropriate legislation, the provisions of this article.",
        '1 Section 1. Equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex. 2  Section 2. 3  The Congress shall have the power to enforce, by appropriate legislation, the provisions of this article. 4  Section 3. Alice Paul rewrote the ERA in 1943 to what is now called the Alice Paul Amendment, reflecting the 15th and the 19th Amendments: Equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex..',
    ]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]

Training Details

Training Dataset

ms_marco

• Dataset: ms_marco at a47ee7a
• Size: 78,704 training samples
• Columns: query, docs, and labels
• Approximate statistics based on the first 1000 samples:

  	query	docs	labels
  type	string	list	list
  details	min: 11 characters mean: 34.01 characters max: 137 characters	min: 3 elements mean: 6.50 elements max: 10 elements	min: 3 elements mean: 6.50 elements max: 10 elements

• Samples:

  query	docs	labels
  what is url means	['Photodisc/Photodisc/Getty Images. Definition: URL stands for Uniform Resource Locator. A URL is a formatted text string used by Web browsers, email clients and other software to identify a network resource on the Internet. Network resources are files that can be plain Web pages, other text documents, graphics, or programs. ', 'Noun. 1. URL-the address of a web page on the world wide web. uniform resource locator, universal resource locator. address, computer address, reference - (computer science) the code that identifies where a piece of information is stored. Translations. An Internet address (for example, http://hmhbooks.com/eref/), usually consisting of the access protocol (http), the domain name (hmhbooks.com), and optionally the path to a file or resource residing on the server where the domain name resides (eref).', 'This is the full, unique location of a resource on a network, such as the Internet. A full URL also defines the method by which the resource is to be retrieved, t...	[1, 0, 0, 0, 0, ...]
  what is biopic	["The definition of a biopic is a dramatic movie about a famous person's life. An example of a biopic is the movie What's Love Got To Do With It, about Tina Turner's life.", 'Chapaev. A biographical film, or biopic (/ˈbaɪoʊpɪk/ ; abbreviation for biographical motion picture), is a film that dramatizes the life of a non-fictional or historically-based person or people. Such films show the life of a historical person and the central character’s real name is used. Roger Ebert defended the The Hurricane and distortions in biographical films in general, stating those who seek the truth about a man from the film of his life might as well seek it from his loving grandmother. ... The Hurricane is not a documentary but a parable .. Some biopics purposely stretch the truth', "BIOPICS FILMS Part 1. Biopic Films (or biographical pictures) are a sub-genre of the larger drama and epic film genres, and although they reached a hey-day of popularity in the 1930s, they are still prominent to this day. '...	[1, 1, 0, 0, 0, ...]
  what is an oystercatcher	['The oystercatchers are a group of waders forming the family Haematopodidae, which has a single genus, Haematopus. They are found on coasts worldwide apart from the polar regions and some tropical regions of Africa and South East Asia. The different species of oystercatcher show little variation in shape or appearance. They range from 39–50 cm (15–20 in) in length and 72–91 cm (28–36 in) in wingspan.', 'The Oystercatcher is a species of bird that lives around Europe and Asia. They usually eat shellfishes that are found on beaches and mud. The Eurasian Oystercatcher uses its sharp bill to open the shells of oysters and mussels. It cuts the muscle between the two halves of the shell together, or it crashes the shell against a rock and eats the oyster inside. With bills like that, oystercatchers are dangerous opponents for other birds. They can fight off predators, and often raid other birds to steal their catches. They attack other birds at an average of five-minute intervals during low...	[1, 1, 0, 0, 0, ...]

• Loss: ListNetLoss with these parameters:

  {
      "activation_fn": "torch.nn.modules.linear.Identity",
      "mini_batch_size": 16
  }
  

Evaluation Dataset

ms_marco

• Dataset: ms_marco at a47ee7a
• Size: 1,000 evaluation samples
• Columns: query, docs, and labels
• Approximate statistics based on the first 1000 samples:

  	query	docs	labels
  type	string	list	list
  details	min: 9 characters mean: 34.2 characters max: 86 characters	min: 2 elements mean: 6.00 elements max: 10 elements	min: 2 elements mean: 6.00 elements max: 10 elements

• Samples:

  query	docs	labels
  equal rights amendment definition quizlet	['Equal Rights Amendment. A constitutional amendment originally introduced in Congress in 1923 and passed by Congress in 1972, stating that equality of rights under the law shall not be denied or abridged by the United States or by any state on account of sex.. ', 'The Equal Rights Amendment (ERA) was a proposed amendment to the United States Constitution designed to guarantee equal rights for women. The ERA was originally written by Alice Paul and Crystal Eastman. In 1923, it was introduced in the Congress for the first time. Oregon-Equality of rights under the law shall not be denied or abridged by the state of Oregon or by any political subdivision in this state on account of sex.', 'The legal analysis for this strategy is outlined in The Equal Rights Amendment: Why the ERA Remains Legally Viable and Properly Before the States, an article by Allison Held, Sheryl Herndon, and Danielle Stager in the Spring 1997 issue of William & Mary Journal of Women and the Law. Alice Paul rewrote t...	[1, 0, 0, 0, 0, ...]
  average cost of bathtub refinishing	['1 Professional Bathtub Refinishing: The average cost range for professional bathtub refinishing is $40 - $50 for materials and 200-$275 for labor for one tub. 2 These ratesw can vary, depending on where you live and project complexity.', 'Typical Bathtub Refinishing Costs. 1 Professional Bathtub Refinishing: The average cost range for professional bathtub refinishing is $40 - $50 for materials and 200-$275 for labor for one tub. 2 These ratesw can vary, depending on where you live and project complexity.', 'Typical Bathtub Refinishing Costs. 1 Professional Bathtub Refinishing: The average cost range for professional bathtub refinishing is $40 - $50 for materials and 200-$275 for labor for one tub. 2 Find recommended, local bathtub refinishing pros on Kudzu.', 'Learn about costs and processes associated with refinishing a bathtub. 1 Professional Bathtub Refinishing: The average cost range for professional bathtub refinishing is $40 - $50 for materials and 200-$275 for labor for ...	[1, 0, 0, 0, 0, ...]
  how long is flu infectious?	['The flu is contagious a day before and five days to a week after its onset. Young children, people over 65, those with compromised immune systems, and people with a terminal illness are at risk of death from the flu. Pregnant women are also at greater risk. The flu can also cause premature birth.', 'Sept. 15, 2009 (San Francisco) -- Some swine flu patients are still infected with H1N1 virus that they can transmit to other people eight to 10 days after their symptoms strike, researchers say.', 'The Flu Is Contagious. Most healthy adults may be able to infect other people beginning 1 day before symptoms develop and up to 5 to 7 days after becoming sick. Children may pass the virus for longer than 7 days. Symptoms start 1 to 4 days after the virus enters the body.', 'Although the typical incubation period for influenza is about one to four days, some adults can be contagious from about one day before onset of symptoms for up to two weeks. Other people who develop complications, such as ...	[1, 0, 0, 0, 0, ...]

• Loss: ListNetLoss with these parameters:

  {
      "activation_fn": "torch.nn.modules.linear.Identity",
      "mini_batch_size": 16
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• learning_rate: 2e-05
• num_train_epochs: 5
• seed: 12
• bf16: True
• load_best_model_at_end: True

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 2e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 5
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.0
• warmup_steps: 0
• log_level: passive
• log_level_replica: warning
• log_on_each_node: True
• logging_nan_inf_filter: True
• save_safetensors: True
• save_on_each_node: False
• save_only_model: False
• restore_callback_states_from_checkpoint: False
• no_cuda: False
• use_cpu: False
• use_mps_device: False
• seed: 12
• data_seed: None
• jit_mode_eval: False
• use_ipex: False
• bf16: True
• fp16: False
• fp16_opt_level: O1
• half_precision_backend: auto
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• local_rank: 0
• ddp_backend: None
• tpu_num_cores: None
• tpu_metrics_debug: False
• debug: []
• dataloader_drop_last: False
• dataloader_num_workers: 0
• dataloader_prefetch_factor: None
• past_index: -1
• disable_tqdm: False
• remove_unused_columns: True
• label_names: None
• load_best_model_at_end: True
• ignore_data_skip: False
• fsdp: []
• fsdp_min_num_params: 0
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• tp_size: 0
• fsdp_transformer_layer_cls_to_wrap: None
• accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• 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
• use_legacy_prediction_loop: False
• push_to_hub: False
• resume_from_checkpoint: None
• hub_model_id: None
• hub_strategy: every_save
• hub_private_repo: None
• hub_always_push: False
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• eval_do_concat_batches: True
• fp16_backend: auto
• push_to_hub_model_id: None
• push_to_hub_organization: None
• mp_parameters:
• auto_find_batch_size: False
• full_determinism: False
• torchdynamo: None
• ray_scope: last
• ddp_timeout: 1800
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• include_tokens_per_second: False
• include_num_input_tokens_seen: False
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Click to expand

Epoch	Step	Training Loss	Validation Loss
0.0002	1	2.068	-
0.0203	100	2.0815	2.0816
0.0407	200	2.073	2.0750
0.0610	300	2.0658	2.0712
0.0813	400	2.0714	2.0691
0.1016	500	2.0779	2.0691
0.1220	600	2.0774	2.0675
0.1423	700	2.0643	2.0676
0.1626	800	2.0695	2.0678
0.1830	900	2.07	2.0662
0.2033	1000	2.0693	2.0667
0.2236	1100	2.0696	2.0650
0.2440	1200	2.0758	2.0656
0.2643	1300	2.0669	2.0649
0.2846	1400	2.0691	2.0648
0.3049	1500	2.0657	2.0652
0.3253	1600	2.0671	2.0651
0.3456	1700	2.0666	2.0649
0.3659	1800	2.0707	2.0649
0.3863	1900	2.0789	2.0651
0.4066	2000	2.0813	2.0655
0.4269	2100	2.0714	2.0663
0.4472	2200	2.0672	2.0654
0.4676	2300	2.069	2.0647
0.4879	2400	2.0717	2.0650
0.5082	2500	2.0657	2.0649
0.5286	2600	2.0709	2.0648
0.5489	2700	2.0662	2.0653
0.5692	2800	2.0683	2.0649
0.5896	2900	2.0757	2.0647
0.6099	3000	2.0752	2.0645
0.6302	3100	2.0752	2.0643
0.6505	3200	2.0722	2.0644
0.6709	3300	2.065	2.0654
0.6912	3400	2.0691	2.0647
0.7115	3500	2.0694	2.0645
0.7319	3600	2.0764	2.0651
0.7522	3700	2.072	2.0644
0.7725	3800	2.0759	2.0641
0.7928	3900	2.0598	2.0647
0.8132	4000	2.0725	2.0644
0.8335	4100	2.0708	2.0650
0.8538	4200	2.08	2.0652
0.8742	4300	2.0642	2.0645
0.8945	4400	2.069	2.0644
0.9148	4500	2.0687	2.0645
0.9351	4600	2.0665	2.0646
0.9555	4700	2.0761	2.0647
0.9758	4800	2.0683	2.0648
0.9961	4900	2.0653	2.0654
1.0165	5000	2.066	2.0645
1.0368	5100	2.0563	2.0651
1.0571	5200	2.0656	2.0658
1.0775	5300	2.0676	2.0647
1.0978	5400	2.0605	2.0655
1.1181	5500	2.0596	2.0653
1.1384	5600	2.0693	2.0646
1.1588	5700	2.0576	2.0646
1.1791	5800	2.0565	2.0656
1.1994	5900	2.0666	2.0649
1.2198	6000	2.0623	2.0647
1.2401	6100	2.0538	2.0649
1.2604	6200	2.0654	2.0659
1.2807	6300	2.0637	2.0664
1.3011	6400	2.0617	2.0655
1.3214	6500	2.0662	2.0657
1.3417	6600	2.0685	2.0650
1.3621	6700	2.0562	2.0652
1.3824	6800	2.065	2.0649
1.4027	6900	2.0638	2.0651
1.4231	7000	2.067	2.0653
1.4434	7100	2.0609	2.0654
1.4637	7200	2.0693	2.0647
1.4840	7300	2.0598	2.0652
1.5044	7400	2.0624	2.0656
1.5247	7500	2.0616	2.0662
1.5450	7600	2.0624	2.0653
1.5654	7700	2.0695	2.0651
1.5857	7800	2.0624	2.0646
1.6060	7900	2.0751	2.0642
1.6263	8000	2.062	2.0643
1.6467	8100	2.0611	2.0645
1.6670	8200	2.0538	2.0645
1.6873	8300	2.0608	2.0642
1.7077	8400	2.0578	2.0644
1.7280	8500	2.0597	2.0646
1.7483	8600	2.0592	2.0656
1.7687	8700	2.0542	2.0648
1.7890	8800	2.0532	2.0648
1.8093	8900	2.0586	2.0646
1.8296	9000	2.0622	2.0643
1.8500	9100	2.068	2.0647
1.8703	9200	2.0697	2.0645
1.8906	9300	2.058	2.0637
1.9110	9400	2.057	2.0647
1.9313	9500	2.0684	2.0641
1.9516	9600	2.0585	2.0644
1.9719	9700	2.0583	2.0646
1.9923	9800	2.0634	2.0643
2.0126	9900	2.0536	2.0675
2.0329	10000	2.0499	2.0696
2.0533	10100	2.0433	2.0685
2.0736	10200	2.0445	2.0699
2.0939	10300	2.0432	2.0729
2.1143	10400	2.0435	2.0696
2.1346	10500	2.0454	2.0708
2.1549	10600	2.052	2.0700
2.1752	10700	2.0351	2.0696
2.1956	10800	2.0388	2.0721
2.2159	10900	2.0394	2.0724
2.2362	11000	2.0497	2.0688
2.2566	11100	2.0529	2.0707
2.2769	11200	2.045	2.0701
2.2972	11300	2.053	2.0694
2.3175	11400	2.0553	2.0710
2.3379	11500	2.0373	2.0696
2.3582	11600	2.0324	2.0698
2.3785	11700	2.052	2.0685
2.3989	11800	2.0411	2.0694
2.4192	11900	2.0523	2.0686
2.4395	12000	2.0439	2.0693
2.4598	12100	2.0484	2.0683
2.4802	12200	2.0468	2.0697
2.5005	12300	2.039	2.0699
2.5208	12400	2.0407	2.0698
2.5412	12500	2.0512	2.0690
2.5615	12600	2.042	2.0687
2.5818	12700	2.041	2.0700
2.6022	12800	2.0462	2.0698
2.6225	12900	2.0452	2.0699
2.6428	13000	2.0442	2.0703
2.6631	13100	2.036	2.0699
2.6835	13200	2.0499	2.0716
2.7038	13300	2.0386	2.0709
2.7241	13400	2.0425	2.0713
2.7445	13500	2.0419	2.0695
2.7648	13600	2.0397	2.0717
2.7851	13700	2.0438	2.0734
2.8054	13800	2.0375	2.0719
2.8258	13900	2.0374	2.0701
2.8461	14000	2.0407	2.0725
2.8664	14100	2.0377	2.0694
2.8868	14200	2.0447	2.0703
2.9071	14300	2.0425	2.0696
2.9274	14400	2.0386	2.0717
2.9478	14500	2.0565	2.0715
2.9681	14600	2.0448	2.0715
2.9884	14700	2.0367	2.0715
3.0087	14800	2.0279	2.0717
3.0291	14900	2.0247	2.0743
3.0494	15000	2.0207	2.0764
3.0697	15100	2.0195	2.0752
3.0901	15200	2.0211	2.0750
3.1104	15300	2.0189	2.0756
3.1307	15400	2.0208	2.0767
3.1510	15500	2.0244	2.0753
3.1714	15600	2.019	2.0758
3.1917	15700	2.0255	2.0766
3.2120	15800	2.0251	2.0762
3.2324	15900	2.0159	2.0745
3.2527	16000	2.0288	2.0753
3.2730	16100	2.0248	2.0762
3.2934	16200	2.0215	2.0777
3.3137	16300	2.0224	2.0759
3.3340	16400	2.0245	2.0796
3.3543	16500	2.0189	2.0783
3.3747	16600	2.0183	2.0759
3.3950	16700	2.0235	2.0766
3.4153	16800	2.0194	2.0761
3.4357	16900	2.0123	2.0788
3.4560	17000	2.0123	2.0768
3.4763	17100	2.0272	2.0777
3.4966	17200	2.0201	2.0777
3.5170	17300	2.0152	2.0766
3.5373	17400	2.0153	2.0753
3.5576	17500	2.0288	2.0782
3.5780	17600	2.0132	2.0777
3.5983	17700	2.0235	2.0773
3.6186	17800	2.0178	2.0751
3.6390	17900	2.0163	2.0767
3.6593	18000	2.0225	2.0755
3.6796	18100	2.0273	2.0752
3.6999	18200	2.0255	2.0774
3.7203	18300	2.0259	2.0757
3.7406	18400	2.0214	2.0767
3.7609	18500	2.0181	2.0763
3.7813	18600	2.0175	2.0752
3.8016	18700	2.0157	2.0761
3.8219	18800	2.0231	2.0769
3.8422	18900	2.0149	2.0760
3.8626	19000	2.0263	2.0759
3.8829	19100	2.02	2.0764
3.9032	19200	2.0161	2.0750
3.9236	19300	2.0269	2.0763
3.9439	19400	2.0255	2.0759
3.9642	19500	2.0215	2.0764
3.9845	19600	2.0124	2.0760
4.0049	19700	2.0265	2.0756
4.0252	19800	2.0055	2.0784
4.0455	19900	2.009	2.0798
4.0659	20000	2.0121	2.0795
4.0862	20100	2.0096	2.0781
4.1065	20200	2.0154	2.0798
4.1269	20300	2.0198	2.0799
4.1472	20400	2.0094	2.0799
4.1675	20500	2.0057	2.0793
4.1878	20600	2.0143	2.0796
4.2082	20700	2.008	2.0821
4.2285	20800	2.0105	2.0801
4.2488	20900	2.0047	2.0814
4.2692	21000	1.9969	2.0807
4.2895	21100	2.0119	2.0799
4.3098	21200	2.0115	2.0803
4.3301	21300	2.0139	2.0800
4.3505	21400	2.0063	2.0793
4.3708	21500	2.0014	2.0811
4.3911	21600	2.0011	2.0798
4.4115	21700	2.0187	2.0816
4.4318	21800	2.0164	2.0801
4.4521	21900	2.014	2.0798
4.4725	22000	2.0102	2.0803
4.4928	22100	2.0042	2.0801
4.5131	22200	2.0083	2.0810
4.5334	22300	2.0105	2.0807
4.5538	22400	2.012	2.0813
4.5741	22500	2.0156	2.0808
4.5944	22600	2.004	2.0811
4.6148	22700	2.0173	2.0805
4.6351	22800	2.0106	2.0804
4.6554	22900	2.0099	2.0808
4.6757	23000	2.0062	2.0808
4.6961	23100	2.0173	2.0808
4.7164	23200	2.0081	2.0804
4.7367	23300	2.0172	2.0802
4.7571	23400	2.0084	2.0808
4.7774	23500	2.0066	2.0809
4.7977	23600	2.0122	2.0803
4.8181	23700	2.0133	2.0810
4.8384	23800	2.0142	2.0804
4.8587	23900	2.0046	2.0807
4.8790	24000	2.0163	2.0806
4.8994	24100	2.0046	2.0808
4.9197	24200	2.009	2.0808
4.9400	24300	2.0092	2.0808
4.9604	24400	2.0084	2.0808
4.9807	24500	2.0028	2.0807

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.11.13
• Sentence Transformers: 5.0.0
• Transformers: 4.51.0
• PyTorch: 2.9.1+cu126
• Accelerate: 1.8.1
• Datasets: 3.6.0
• Tokenizers: 0.21.4-dev.0

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",
}

ListNetLoss

@inproceedings{cao2007learning,
    title={Learning to Rank: From Pairwise Approach to Listwise Approach},
    author={Cao, Zhe and Qin, Tao and Liu, Tie-Yan and Tsai, Ming-Feng and Li, Hang},
    booktitle={Proceedings of the 24th international conference on Machine learning},
    pages={129--136},
    year={2007}
}