ChengyouXin/cacheverifier-quora

CrossEncoder

This is a Cross Encoder model trained 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
• Maximum Sequence Length: 512 tokens
• Number of Output Labels: 1 label
• Supported Modality: Text

Model Sources

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

Full Model Architecture

CrossEncoder(
  (0): Transformer({'transformer_task': 'sequence-classification', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'logits'}}, 'module_output_name': 'scores', 'architecture': 'BertForSequenceClassification'})
)

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("cross_encoder_model_id")
# Get scores for pairs of inputs
pairs = [
    ['How do you control your horniness?', 'How do I control my horny emotions?'],
    ['What do i do after my MBBS ?', 'What can one do after MBBS?'],
    ['What is the county of Edgware and how does the lifestyle compare to the London Borough of Enfield?', 'What is the district of Edgware and how does the lifestyle compare to the London Borough of Islington?'],
    ['What is a qualified SAP ERP key user?', 'What is the responsibility of SAP ERP key user?'],
    ['Which is the best book for tensor calculus?', 'Which is the best book to study TENSOR for general relativity from basic?'],
]
scores = model.predict(pairs)
print(scores)
# [ 0.0335  0.6294 -2.3788 -0.096  -0.4309]

# Or rank different texts based on similarity to a single text
ranks = model.rank(
    'How do you control your horniness?',
    [
        'How do I control my horny emotions?',
        'What can one do after MBBS?',
        'What is the district of Edgware and how does the lifestyle compare to the London Borough of Islington?',
        'What is the responsibility of SAP ERP key user?',
        'Which is the best book to study TENSOR for general relativity from basic?',
    ]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]

Training Details

Training Dataset

Unnamed Dataset

• Size: 11,928 training samples
• Columns: query, response, and label
• Approximate statistics based on the first 100 samples:

  	query	response	label
  type	string	string	float
  modality	text	text
  details	min: 7 tokens mean: 14.07 tokens max: 24 tokens	min: 8 tokens mean: 14.39 tokens max: 31 tokens	min: 0.0 mean: 0.62 max: 1.0

• Samples:

  query	response	label
  How do you control your horniness?	How do I control my horny emotions?	1.0
  What do i do after my MBBS ?	What can one do after MBBS?	1.0
  What is the county of Edgware and how does the lifestyle compare to the London Borough of Enfield?	What is the district of Edgware and how does the lifestyle compare to the London Borough of Islington?	0.0

• Loss: BinaryCrossEntropyLoss with these parameters:

  {
      "activation_fn": "torch.nn.modules.linear.Identity",
      "pos_weight": null
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 16
• num_train_epochs: 1
• disable_tqdm: True

All Hyperparameters

Click to expand

• per_device_train_batch_size: 16
• num_train_epochs: 1
• max_steps: -1
• learning_rate: 5e-05
• lr_scheduler_type: linear
• lr_scheduler_kwargs: None
• warmup_steps: 0
• 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: False
• 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: True
• project: huggingface
• trackio_space_id: None
• trackio_bucket_id: None
• trackio_static_space_id: None
• per_device_eval_batch_size: 8
• 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: False
• 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: None
• fsdp_config: None
• deepspeed: None
• debug: []
• skip_memory_metrics: True
• do_predict: False
• resume_from_checkpoint: None
• warmup_ratio: None
• local_rank: -1
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss
0.0013	1	2.9614
0.0134	10	0.8912
0.0268	20	0.8821
0.0402	30	0.6854
0.0536	40	0.7558
0.0670	50	0.6960
0.0804	60	0.6753
0.0938	70	0.6979
0.1072	80	0.6919
0.1206	90	0.6373
0.1340	100	0.6750
0.1475	110	0.7235
0.1609	120	0.6508
0.1743	130	0.6698
0.1877	140	0.6603
0.2011	150	0.6601
0.2145	160	0.6269
0.2279	170	0.6568
0.2413	180	0.5662
0.2547	190	0.6341
0.2681	200	0.6649
0.2815	210	0.6582
0.2949	220	0.6966
0.3083	230	0.5850
0.3217	240	0.5919
0.3351	250	0.6952
0.3485	260	0.6682
0.3619	270	0.6402
0.3753	280	0.6923
0.3887	290	0.5896
0.4021	300	0.6448
0.4155	310	0.6208
0.4290	320	0.6557
0.4424	330	0.6780
0.4558	340	0.6057
0.4692	350	0.6660
0.4826	360	0.6834
0.4960	370	0.6351
0.5094	380	0.6442
0.5228	390	0.6002
0.5362	400	0.6454
0.5496	410	0.6431
0.5630	420	0.6146
0.5764	430	0.5826
0.5898	440	0.6906
0.6032	450	0.6260
0.6166	460	0.6390
0.6300	470	0.6107
0.6434	480	0.6381
0.6568	490	0.6296
0.6702	500	0.6163
0.6836	510	0.5750
0.6971	520	0.6387
0.7105	530	0.6353
0.7239	540	0.5639
0.7373	550	0.5501
0.7507	560	0.6608
0.7641	570	0.6868
0.7775	580	0.5937
0.7909	590	0.6198
0.8043	600	0.6683
0.8177	610	0.6228
0.8311	620	0.5776
0.8445	630	0.6115
0.8579	640	0.6536
0.8713	650	0.6366
0.8847	660	0.6278
0.8981	670	0.6331
0.9115	680	0.5928
0.9249	690	0.6246
0.9383	700	0.6273
0.9517	710	0.6254
0.9651	720	0.5991
0.9786	730	0.6309
0.9920	740	0.5972

Training Time

• Training: 19.5 seconds

Framework Versions

• Python: 3.11.6
• Sentence Transformers: 5.6.1
• Transformers: 5.14.1
• PyTorch: 2.13.0+cu130
• Accelerate: 1.14.0
• Datasets: 5.0.0
• Tokenizers: 0.22.2

Additional Resources

• Training and Finetuning Reranker Models with Sentence Transformers: the end-to-end guide for training or finetuning Cross Encoder (reranker) models.
• Multimodal Embedding & Reranker Models with Sentence Transformers: use text, image, audio, and video reranker models through the same API.
• Training and Finetuning Multimodal Embedding & Reranker Models with Sentence Transformers: training multimodal Cross Encoders.

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",
}