ayushexel/reranker-MiniLM-L6-H384-uncased-gooaq-bce-495000

ModernBERT-base trained on GooAQ

This is a Cross Encoder model finetuned from nreimers/MiniLM-L6-H384-uncased 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: nreimers/MiniLM-L6-H384-uncased
• Maximum Sequence Length: 512 tokens
• Number of Output Labels: 1 label
• Language: en
• License: apache-2.0

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("ayushexel/reranker-MiniLM-L6-H384-uncased-gooaq-bce-495000")
# Get scores for pairs of texts
pairs = [
    ["in grey's anatomy how does izzie die?", 'After speculation that Izzie would be killed off in the fifth season, the character was diagnosed with Stage 4 metastatic melanoma.'],
    ["in grey's anatomy how does izzie die?", "Izzie later admitted to George that she was in love with him, leaving him speechless. George later admitted he loved Izzie too, despite his strange reaction to her when she confessed her love to him. Their relationship was soon discovered by George's wife, Callie and the two got a divorce."],
    ["in grey's anatomy how does izzie die?", "The episode in which Derek Shepherd (Patrick Dempsey) dies is one that most Grey's Anatomy fans will never forget. The fateful incident occurred in season 11, episode 21, and it was titled, “How To Save a Life.” The attending doctor who failed to save McDreamy's life recently appeared in an episode of Grey's Anatomy."],
    ["in grey's anatomy how does izzie die?", "Richard Webber, Grey's Anatomy fans are nervous he'll die, though nothing is set in stone on the show yet. Warning: Spoilers for Season 16, Episode 19 of Grey's Anatomy follow."],
    ["in grey's anatomy how does izzie die?", "Izzie eventually forgives him, and they begin dating again until Denny enters the picture. After Denny's death they begin dating yet again and following her recovery from cancer they get married, but it doesn't last."],
]
scores = model.predict(pairs)
print(scores.shape)
# (5,)

# Or rank different texts based on similarity to a single text
ranks = model.rank(
    "in grey's anatomy how does izzie die?",
    [
        'After speculation that Izzie would be killed off in the fifth season, the character was diagnosed with Stage 4 metastatic melanoma.',
        "Izzie later admitted to George that she was in love with him, leaving him speechless. George later admitted he loved Izzie too, despite his strange reaction to her when she confessed her love to him. Their relationship was soon discovered by George's wife, Callie and the two got a divorce.",
        "The episode in which Derek Shepherd (Patrick Dempsey) dies is one that most Grey's Anatomy fans will never forget. The fateful incident occurred in season 11, episode 21, and it was titled, “How To Save a Life.” The attending doctor who failed to save McDreamy's life recently appeared in an episode of Grey's Anatomy.",
        "Richard Webber, Grey's Anatomy fans are nervous he'll die, though nothing is set in stone on the show yet. Warning: Spoilers for Season 16, Episode 19 of Grey's Anatomy follow.",
        "Izzie eventually forgives him, and they begin dating again until Denny enters the picture. After Denny's death they begin dating yet again and following her recovery from cancer they get married, but it doesn't last.",
    ]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]

Evaluation

Metrics

Cross Encoder Reranking

• Dataset: gooaq-dev
• Evaluated with CrossEncoderRerankingEvaluator with these parameters:

  {
      "at_k": 10,
      "always_rerank_positives": false
  }
  

Metric	Value
map	0.5291 (+0.1486)
mrr@10	0.5258 (+0.1553)
ndcg@10	0.5805 (+0.1477)

Cross Encoder Reranking

• Datasets: NanoMSMARCO_R100, NanoNFCorpus_R100 and NanoNQ_R100
• Evaluated with CrossEncoderRerankingEvaluator with these parameters:

  {
      "at_k": 10,
      "always_rerank_positives": true
  }
  

Metric	NanoMSMARCO_R100	NanoNFCorpus_R100	NanoNQ_R100
map	0.2939 (-0.1956)	0.3242 (+0.0632)	0.2769 (-0.1427)
mrr@10	0.2772 (-0.2003)	0.5253 (+0.0255)	0.2629 (-0.1638)
ndcg@10	0.3678 (-0.1726)	0.3345 (+0.0095)	0.3325 (-0.1682)

Cross Encoder Nano BEIR

• Dataset: NanoBEIR_R100_mean
• Evaluated with CrossEncoderNanoBEIREvaluator with these parameters:

  {
      "dataset_names": [
          "msmarco",
          "nfcorpus",
          "nq"
      ],
      "rerank_k": 100,
      "at_k": 10,
      "always_rerank_positives": true
  }
  

Metric	Value
map	0.2984 (-0.0917)
mrr@10	0.3552 (-0.1128)
ndcg@10	0.3449 (-0.1104)

Training Details

Training Dataset

Unnamed Dataset

• Size: 2,749,365 training samples
• Columns: question, answer, and label
• Approximate statistics based on the first 1000 samples:

  	question	answer	label
  type	string	string	int
  details	min: 19 characters mean: 42.17 characters max: 79 characters	min: 54 characters mean: 246.01 characters max: 399 characters	0: ~81.90% 1: ~18.10%

• Samples:

  question	answer	label
  in grey's anatomy how does izzie die?	After speculation that Izzie would be killed off in the fifth season, the character was diagnosed with Stage 4 metastatic melanoma.	1
  in grey's anatomy how does izzie die?	Izzie later admitted to George that she was in love with him, leaving him speechless. George later admitted he loved Izzie too, despite his strange reaction to her when she confessed her love to him. Their relationship was soon discovered by George's wife, Callie and the two got a divorce.	0
  in grey's anatomy how does izzie die?	The episode in which Derek Shepherd (Patrick Dempsey) dies is one that most Grey's Anatomy fans will never forget. The fateful incident occurred in season 11, episode 21, and it was titled, “How To Save a Life.” The attending doctor who failed to save McDreamy's life recently appeared in an episode of Grey's Anatomy.	0

• Loss: BinaryCrossEntropyLoss with these parameters:

  {
      "activation_fn": "torch.nn.modules.linear.Identity",
      "pos_weight": 5
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 256
• per_device_eval_batch_size: 256
• learning_rate: 2e-05
• num_train_epochs: 1
• warmup_ratio: 0.1
• seed: 12
• bf16: True
• dataloader_num_workers: 12
• 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: 256
• per_device_eval_batch_size: 256
• 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: 1
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.1
• 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: 12
• 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
• dispatch_batches: None
• split_batches: 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

Training Logs

Epoch	Step	Training Loss	gooaq-dev_ndcg@10	NanoMSMARCO_R100_ndcg@10	NanoNFCorpus_R100_ndcg@10	NanoNQ_R100_ndcg@10	NanoBEIR_R100_mean_ndcg@10
-1	-1	-	0.1141 (-0.3187)	0.0667 (-0.4737)	0.2984 (-0.0267)	0.0318 (-0.4689)	0.1323 (-0.3231)
0.0001	1	1.2808	-	-	-	-	-
0.0186	200	1.196	-	-	-	-	-
0.0372	400	1.1939	-	-	-	-	-
0.0559	600	1.1823	-	-	-	-	-
0.0745	800	1.1506	-	-	-	-	-
0.0931	1000	0.9972	-	-	-	-	-
0.1117	1200	0.9336	-	-	-	-	-
0.1304	1400	0.898	-	-	-	-	-
0.1490	1600	0.8582	-	-	-	-	-
0.1676	1800	0.8391	-	-	-	-	-
0.1862	2000	0.8153	-	-	-	-	-
0.2048	2200	0.7999	-	-	-	-	-
0.2235	2400	0.7793	-	-	-	-	-
0.2421	2600	0.7889	-	-	-	-	-
0.2607	2800	0.7576	-	-	-	-	-
0.2793	3000	0.7592	-	-	-	-	-
0.2980	3200	0.7543	-	-	-	-	-
0.3166	3400	0.7437	-	-	-	-	-
0.3352	3600	0.7426	-	-	-	-	-
0.3538	3800	0.7337	-	-	-	-	-
0.3724	4000	0.7312	-	-	-	-	-
0.3911	4200	0.7212	-	-	-	-	-
0.4097	4400	0.7281	-	-	-	-	-
0.4283	4600	0.7166	-	-	-	-	-
0.4469	4800	0.7167	-	-	-	-	-
0.4655	5000	0.7175	-	-	-	-	-
0.4842	5200	0.7176	-	-	-	-	-
0.5028	5400	0.7141	-	-	-	-	-
0.5214	5600	0.6963	-	-	-	-	-
0.5400	5800	0.6888	-	-	-	-	-
0.5587	6000	0.6937	-	-	-	-	-
0.5773	6200	0.7009	-	-	-	-	-
0.5959	6400	0.6887	-	-	-	-	-
0.6145	6600	0.6933	-	-	-	-	-
0.6331	6800	0.692	-	-	-	-	-
0.6518	7000	0.6874	-	-	-	-	-
0.6704	7200	0.6792	-	-	-	-	-
0.6890	7400	0.6772	-	-	-	-	-
0.7076	7600	0.6804	-	-	-	-	-
0.7263	7800	0.6728	-	-	-	-	-
0.7449	8000	0.6703	-	-	-	-	-
0.7635	8200	0.6844	-	-	-	-	-
0.7821	8400	0.6663	-	-	-	-	-
0.8007	8600	0.6775	-	-	-	-	-
0.8194	8800	0.6647	-	-	-	-	-
0.8380	9000	0.6818	-	-	-	-	-
0.8566	9200	0.6724	-	-	-	-	-
0.8752	9400	0.6748	-	-	-	-	-
0.8939	9600	0.6567	-	-	-	-	-
0.9125	9800	0.6682	-	-	-	-	-
0.9311	10000	0.6747	-	-	-	-	-
0.9497	10200	0.6618	-	-	-	-	-
0.9683	10400	0.6625	-	-	-	-	-
0.9870	10600	0.6629	-	-	-	-	-
-1	-1	-	0.5805 (+0.1477)	0.3678 (-0.1726)	0.3345 (+0.0095)	0.3325 (-0.1682)	0.3449 (-0.1104)

Framework Versions

• Python: 3.11.0
• Sentence Transformers: 4.0.1
• Transformers: 4.50.3
• PyTorch: 2.6.0+cu124
• Accelerate: 1.5.2
• Datasets: 3.5.0
• Tokenizers: 0.21.1

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",
}