Add new CrossEncoder model

bbc4e48 verified 2 days ago

17.9 kB

language:
  - en
tags:
  - sentence-transformers
  - cross-encoder
  - reranker
  - generated_from_trainer
  - dataset_size:5749
  - loss:BinaryCrossEntropyLoss
base_model: distilbert/distilroberta-base
datasets:
  - sentence-transformers/stsb
pipeline_tag: text-ranking
library_name: sentence-transformers
metrics:
  - pearson
  - spearman
model-index:
  - name: CrossEncoder based on distilbert/distilroberta-base
    results:
      - task:
          type: cross-encoder-correlation
          name: Cross Encoder Correlation
        dataset:
          name: stsb validation
          type: stsb-validation
        metrics:
          - type: pearson
            value: 0.8864227817727027
            name: Pearson
          - type: spearman
            value: 0.8837678149208236
            name: Spearman
      - task:
          type: cross-encoder-correlation
          name: Cross Encoder Correlation
        dataset:
          name: stsb test
          type: stsb-test
        metrics:
          - type: pearson
            value: 0.8503521391700528
            name: Pearson
          - type: spearman
            value: 0.8403655772346184
            name: Spearman

CrossEncoder based on distilbert/distilroberta-base

This is a Cross Encoder model finetuned from distilbert/distilroberta-base on the stsb 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: distilbert/distilroberta-base
Maximum Sequence Length: 512 tokens
Number of Output Labels: 1 label
Supported Modality: Text
Training Dataset:
- stsb
Language: en

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': 'RobertaForSequenceClassification'})
)

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("omkar334/reranker-distilroberta-base-stsb")
# Get scores for pairs of inputs
pairs = [
    ['A man with a hard hat is dancing.', 'A man wearing a hard hat is dancing.'],
    ['A young child is riding a horse.', 'A child is riding a horse.'],
    ['A man is feeding a mouse to a snake.', 'The man is feeding a mouse to the snake.'],
    ['A woman is playing the guitar.', 'A man is playing guitar.'],
    ['A woman is playing the flute.', 'A man is playing a flute.'],
]
scores = model.predict(pairs)
print(scores)
# [0.9598 0.9533 0.9566 0.3766 0.4535]

# Or rank different texts based on similarity to a single text
ranks = model.rank(
    'A man with a hard hat is dancing.',
    [
        'A man wearing a hard hat is dancing.',
        'A child is riding a horse.',
        'The man is feeding a mouse to the snake.',
        'A man is playing guitar.',
        'A man is playing a flute.',
    ]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]

Evaluation

Metrics

Cross Encoder Correlation

Datasets: stsb-validation and stsb-test
Evaluated with CrossEncoderCorrelationEvaluator

Metric	stsb-validation	stsb-test
pearson	0.8864	0.8504
spearman	0.8838	0.8404

Training Details

Training Dataset

stsb

Dataset: stsb at ab7a5ac
Size: 5,749 training samples
Columns: sentence1, sentence2, and score

Approximate statistics based on the first 100 samples:

	sentence1	sentence2	score
type	string	string	float
modality	text	text
details	min: 7 tokens mean: 9.49 tokens max: 14 tokens	min: 7 tokens mean: 9.61 tokens max: 17 tokens	min: 0.1 mean: 0.66 max: 1.0

Samples:

sentence1	sentence2	score
`A plane is taking off.`	`An air plane is taking off.`	`1.0`
`A man is playing a large flute.`	`A man is playing a flute.`	`0.76`
`A man is spreading shreded cheese on a pizza.`	`A man is spreading shredded cheese on an uncooked pizza.`	`0.76`

Loss: BinaryCrossEntropyLoss with these parameters:

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

Evaluation Dataset

stsb

Dataset: stsb at ab7a5ac
Size: 1,500 evaluation samples
Columns: sentence1, sentence2, and score

Approximate statistics based on the first 100 samples:

	sentence1	sentence2	score
type	string	string	float
modality	text	text
details	min: 7 tokens mean: 10.04 tokens max: 19 tokens	min: 7 tokens mean: 9.98 tokens max: 18 tokens	min: 0.0 mean: 0.53 max: 1.0

Samples:

sentence1	sentence2	score
`A man with a hard hat is dancing.`	`A man wearing a hard hat is dancing.`	`1.0`
`A young child is riding a horse.`	`A child is riding a horse.`	`0.95`
`A man is feeding a mouse to a snake.`	`The man is feeding a mouse to the snake.`	`1.0`

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: 64
num_train_epochs: 4
warmup_steps: 0.1
bf16: True
per_device_eval_batch_size: 64

All Hyperparameters

Click to expand

per_device_train_batch_size: 64
num_train_epochs: 4
max_steps: -1
learning_rate: 5e-05
lr_scheduler_type: linear
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: 64
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: []
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: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: proportional
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	Validation Loss	stsb-validation_spearman	stsb-test_spearman
-1	-1	-	-	-0.0362	-
0.2222	20	0.6909	-	-	-
0.4444	40	0.6506	-	-	-
0.6667	60	0.5969	-	-	-
0.8889	80	0.5680	0.5461	0.8552	-
1.1111	100	0.5551	-	-	-
1.3333	120	0.5379	-	-	-
1.5556	140	0.5449	-	-	-
1.7778	160	0.5443	0.5342	0.8777	-
2.0	180	0.5373	-	-	-
2.2222	200	0.5287	-	-	-
2.4444	220	0.5248	-	-	-
2.6667	240	0.5283	0.5383	0.8785	-
2.8889	260	0.5251	-	-	-
3.1111	280	0.5156	-	-	-
3.3333	300	0.5093	-	-	-
3.5556	320	0.5164	0.5369	0.8824	-
3.7778	340	0.5152	-	-	-
4.0	360	0.5208	0.5331	0.8838	-
-1	-1	-	-	-	0.8404

Training Time

Training: 3.2 minutes
Evaluation: 15.8 seconds
Total: 3.5 minutes

Framework Versions

Python: 3.11.14
Sentence Transformers: 5.6.0.dev0
Transformers: 5.9.0
PyTorch: 2.12.0
Accelerate: 1.13.0
Datasets: 4.8.5
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",
}