TakoData/chart-reranker

CrossEncoder based on Alibaba-NLP/gte-multilingual-reranker-base

This is a Cross Encoder model finetuned from Alibaba-NLP/gte-multilingual-reranker-base 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: Alibaba-NLP/gte-multilingual-reranker-base
• Maximum Sequence Length: 512 tokens
• Number of Output Labels: 1 label

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("cross_encoder_model_id")
# Get scores for pairs of texts
pairs = [
    ['include the popular publications as well', 'Title: "Americans\' Library use - past 3 months (United States)"\nCollections: YouGov Trackers\nDatasets: YouGovTrackerValueV2\nChart Type: survey:timeseries\nSources: YouGov'],
    ['Give it a good research topic', 'Title: "The most important issues facing the country (United Kingdom)"\nCollections: YouGov Trackers\nDatasets: YouGovTrackerValueV2\nChart Type: survey:timeseries\nSources: YouGov'],
    ['When and where are the Denver Broncos playing the Kansas City Chiefs?', 'Title: "Denver Broncos at Kansas City Chiefs"\nCollections: Football\nChart Type: game_score:football'],
    ['49ers vs Seahawks', 'Title: "Seahawk Deep Ocean Technology, Inc. Overview"\nCollections: Companies\nChart Type: company:finance\nCanonical forms: "Overview"="Stock Overview"\nSources: S&P Global'],
    ['Comparative review of JBL vs Marshall 2025 Bluetooth speakers', 'Title: "B&C Speakers Overview"\nCollections: Companies\nChart Type: company:finance\nCanonical forms: "B&C Speakers"="B&C Speakers S.p.A.", "Overview"="Stock Overview"\nSources: S&P Global'],
]
scores = model.predict(pairs)
print(scores.shape)
# (5,)

# Or rank different texts based on similarity to a single text
ranks = model.rank(
    'include the popular publications as well',
    [
        'Title: "Americans\' Library use - past 3 months (United States)"\nCollections: YouGov Trackers\nDatasets: YouGovTrackerValueV2\nChart Type: survey:timeseries\nSources: YouGov',
        'Title: "The most important issues facing the country (United Kingdom)"\nCollections: YouGov Trackers\nDatasets: YouGovTrackerValueV2\nChart Type: survey:timeseries\nSources: YouGov',
        'Title: "Denver Broncos at Kansas City Chiefs"\nCollections: Football\nChart Type: game_score:football',
        'Title: "Seahawk Deep Ocean Technology, Inc. Overview"\nCollections: Companies\nChart Type: company:finance\nCanonical forms: "Overview"="Stock Overview"\nSources: S&P Global',
        'Title: "B&C Speakers Overview"\nCollections: Companies\nChart Type: company:finance\nCanonical forms: "B&C Speakers"="B&C Speakers S.p.A.", "Overview"="Stock Overview"\nSources: S&P Global',
    ]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]

Evaluation

Metrics

Cross Encoder Correlation

• Dataset: validation
• Evaluated with CrossEncoderCorrelationEvaluator

Metric	Value
pearson	0.8721
spearman	0.8685

Training Details

Training Dataset

Unnamed Dataset

• Size: 24,504 training samples
• Columns: sentence_0, sentence_1, and label
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1	label
  type	string	string	float
  details	min: 2 characters mean: 86.83 characters max: 993 characters	min: 77 characters mean: 169.16 characters max: 360 characters	min: 0.0 mean: 0.42 max: 1.0

• Samples:

  sentence_0	sentence_1	label
  include the popular publications as well	Title: "Americans' Library use - past 3 months (United States)" Collections: YouGov Trackers Datasets: YouGovTrackerValueV2 Chart Type: survey:timeseries Sources: YouGov	0.5
  Give it a good research topic	Title: "The most important issues facing the country (United Kingdom)" Collections: YouGov Trackers Datasets: YouGovTrackerValueV2 Chart Type: survey:timeseries Sources: YouGov	1.0
  When and where are the Denver Broncos playing the Kansas City Chiefs?	Title: "Denver Broncos at Kansas City Chiefs" Collections: Football Chart Type: game_score:football	1.0

• Loss: BinaryCrossEntropyLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• num_train_epochs: 5
• fp16: 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: 32
• per_device_eval_batch_size: 32
• 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: 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: 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: 42
• data_seed: None
• jit_mode_eval: False
• bf16: False
• fp16: True
• 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: False
• 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}
• 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}
• parallelism_config: None
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch_fused
• optim_args: None
• adafactor: False
• 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
• 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
• hub_revision: None
• 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: 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
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	validation_spearman
0.1305	100	-	0.7594
0.2611	200	-	0.7951
0.3916	300	-	0.8050
0.5222	400	-	0.8200
0.6527	500	0.468	0.8290
0.7833	600	-	0.8331
0.9138	700	-	0.8347
1.0	766	-	0.8434
1.0444	800	-	0.8432
1.1749	900	-	0.8467
1.3055	1000	0.4135	0.8473
1.4360	1100	-	0.8475
1.5666	1200	-	0.8535
1.6971	1300	-	0.8518
1.8277	1400	-	0.8571
1.9582	1500	0.3747	0.8577
2.0	1532	-	0.8556
2.0888	1600	-	0.8587
2.2193	1700	-	0.8609
2.3499	1800	-	0.8612
2.4804	1900	-	0.8619
2.6110	2000	0.3515	0.8626
2.7415	2100	-	0.8622
2.8721	2200	-	0.8653
3.0	2298	-	0.8656
3.0026	2300	-	0.8656
3.1332	2400	-	0.8643
3.2637	2500	0.3421	0.8646
3.3943	2600	-	0.8654
3.5248	2700	-	0.8666
3.6554	2800	-	0.8640
3.7859	2900	-	0.8685

Framework Versions

• Python: 3.12.11
• Sentence Transformers: 5.1.2
• Transformers: 4.57.1
• PyTorch: 2.8.0+cu128
• Accelerate: 1.11.0
• Datasets: 4.2.0
• Tokenizers: 0.22.1

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",
}