This is a Cross Encoder model finetuned from cross-encoder/ms-marco-MiniLM-L6-v2 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: cross-encoder/ms-marco-MiniLM-L6-v2
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.

python

from sentence_transformers import CrossEncoder

# Download from the 🤗 Hub
model = CrossEncoder("cross_encoder_model_id")
# Get scores for pairs of inputs
pairs = [
    ['skill: manage health care staff', 'skill: manage healthcare staff'],
    ['skill: manage health care staff', 'skill: manage staff'],
    ['skill: manage health care staff', 'skill: manage physiotherapy staff'],
    ['skill: manage health care staff', 'skill: work with nursing staff'],
    ['skill: manage health care staff', 'skill: manage chiropractic staff'],
]
scores = model.predict(pairs)
print(scores)
# [ 7.403   3.9133  0.9482 -2.4038 -3.8219]

# Or rank different texts based on similarity to a single text
ranks = model.rank(
    'skill: manage health care staff',
    [
        'skill: manage healthcare staff',
        'skill: manage staff',
        'skill: manage physiotherapy staff',
        'skill: work with nursing staff',
        'skill: manage chiropractic staff',
    ]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]

Training Details

Training Dataset

Unnamed Dataset

Size: 62,318 training samples
Columns: sentence_A, sentence_B, and label
Approximate statistics based on the first 100 samples:
sentence_A sentence_B label
type string string float
modality text text
details
min: 7 tokens
mean: 8.14 tokens
max: 11 tokens
min: 6 tokens
mean: 7.94 tokens
max: 15 tokens
min: 0.0
mean: 0.22
max: 1.0

Samples:

sentence_A	sentence_B	label
`skill: manage security systems`	`skill: manage technical security systems`	`1.0`
`skill: manage security systems`	`skill: Physical access management software`	`0.0`
`skill: manage security systems`	`skill: manage theft prevention`	`0.0`

Loss: BinaryCrossEntropyLoss with these parameters:
json
1{ 2 "activation_fn": "torch.nn.modules.linear.Identity", 3 "pos_weight": null 4}

Evaluation Dataset

Unnamed Dataset

Size: 7,949 evaluation samples
Columns: sentence_A, sentence_B, and label
Approximate statistics based on the first 100 samples:
sentence_A sentence_B label
type string string float
modality text text
details
min: 6 tokens
mean: 9.23 tokens
max: 14 tokens
min: 6 tokens
mean: 8.06 tokens
max: 16 tokens
min: 0.0
mean: 0.21
max: 1.0

Samples:

sentence_A	sentence_B	label
`skill: manage health care staff`	`skill: manage healthcare staff`	`1.0`
`skill: manage health care staff`	`skill: manage staff`	`0.0`
`skill: manage health care staff`	`skill: manage physiotherapy staff`	`0.0`

Loss: BinaryCrossEntropyLoss with these parameters:
json
1{ 2 "activation_fn": "torch.nn.modules.linear.Identity", 3 "pos_weight": null 4}

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 32
num_train_epochs: 2
learning_rate: 2e-05
warmup_steps: 0.1
per_device_eval_batch_size: 64
load_best_model_at_end: True

All Hyperparameters

Click to expand

per_device_train_batch_size: 32
num_train_epochs: 2
max_steps: -1
learning_rate: 2e-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: 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: 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: True
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

Click to expand

Epoch	Step	Training Loss	Validation Loss
0.0103	20	0.4811	-
0.0205	40	0.4186	-
0.0308	60	0.3713	-
0.0411	80	0.4399	-
0.0513	100	0.3115	-
0.0616	120	0.3429	-
0.0719	140	0.3541	-
0.0821	160	0.3054	-
0.0924	180	0.2590	-
0.1027	200	0.2288	-
0.1129	220	0.2604	-
0.1232	240	0.2387	-
0.1335	260	0.2248	-
0.1437	280	0.2176	-
0.1540	300	0.2260	-
0.1643	320	0.2212	-
0.1745	340	0.1946	-
0.1848	360	0.1926	-
0.1951	380	0.2705	-
0.2053	400	0.1964	-
0.2156	420	0.2341	-
0.2259	440	0.3130	-
0.2361	460	0.2246	-
0.2464	480	0.2541	-
0.2567	500	0.2328	-
0.2669	520	0.1759	-
0.2772	540	0.2553	-
0.2875	560	0.2113	-
0.2977	580	0.1925	-
0.3080	600	0.2061	-
0.3183	620	0.2053	-
0.3285	640	0.1886	-
0.3388	660	0.2209	-
0.3491	680	0.1731	-
0.3593	700	0.2043	-
0.3696	720	0.1732	-
0.3799	740	0.1905	-
0.3901	760	0.1987	-
0.4004	780	0.2041	-
0.4107	800	0.1861	-
0.4209	820	0.1963	-
0.4312	840	0.2504	-
0.4415	860	0.1384	-
0.4517	880	0.1731	-
0.4620	900	0.2093	-
0.4723	920	0.1958	-
0.4825	940	0.1908	-
0.4928	960	0.1373	-
0.5031	980	0.1874	-
0.5133	1000	0.1905	-
0.5236	1020	0.1678	-
0.5339	1040	0.1669	-
0.5441	1060	0.1743	-
0.5544	1080	0.2382	-
0.5647	1100	0.1753	-
0.5749	1120	0.1357	-
0.5852	1140	0.1531	-
0.5955	1160	0.1768	-
0.6057	1180	0.1541	-
0.6160	1200	0.1734	-
0.6263	1220	0.1747	-
0.6366	1240	0.1619	-
0.6468	1260	0.1658	-
0.6571	1280	0.1491	-
0.6674	1300	0.1838	-
0.6776	1320	0.1480	-
0.6879	1340	0.2149	-
0.6982	1360	0.1989	-
0.7084	1380	0.1488	-
0.7187	1400	0.1604	-
0.7290	1420	0.1450	-
0.7392	1440	0.1516	-
0.7495	1460	0.1387	-
0.7598	1480	0.1702	-
0.7700	1500	0.1389	-
0.7803	1520	0.1544	-
0.7906	1540	0.1408	-
0.8008	1560	0.1472	-
0.8111	1580	0.1578	-
0.8214	1600	0.1756	-
0.8316	1620	0.1655	-
0.8419	1640	0.1523	-
0.8522	1660	0.1431	-
0.8624	1680	0.1922	-
0.8727	1700	0.1619	-
0.8830	1720	0.1778	-
0.8932	1740	0.1891	-
0.9035	1760	0.1858	-
0.9138	1780	0.1597	-
0.9240	1800	0.1609	-
0.9343	1820	0.1330	-
0.9446	1840	0.1735	-
0.9548	1860	0.1859	-
0.9651	1880	0.1486	-
0.9754	1900	0.1113	-
0.9856	1920	0.1812	-
0.9959	1940	0.1076	-
1.0	1948	-	0.2882
1.0062	1960	0.1339	-
1.0164	1980	0.1221	-
1.0267	2000	0.1282	-
1.0370	2020	0.1004	-
1.0472	2040	0.1151	-
1.0575	2060	0.1477	-
1.0678	2080	0.1306	-
1.0780	2100	0.1328	-
1.0883	2120	0.1676	-
1.0986	2140	0.1171	-
1.1088	2160	0.1300	-
1.1191	2180	0.1291	-
1.1294	2200	0.1282	-
1.1396	2220	0.1393	-
1.1499	2240	0.1019	-
1.1602	2260	0.1536	-
1.1704	2280	0.1487	-
1.1807	2300	0.0985	-
1.1910	2320	0.1407	-
1.2012	2340	0.1455	-
1.2115	2360	0.1396	-
1.2218	2380	0.1301	-
1.2320	2400	0.1315	-
1.2423	2420	0.1156	-
1.2526	2440	0.1522	-
1.2628	2460	0.1360	-
1.2731	2480	0.1219	-
1.2834	2500	0.1306	-
1.2936	2520	0.1129	-
1.3039	2540	0.1561	-
1.3142	2560	0.1061	-
1.3244	2580	0.0967	-
1.3347	2600	0.0994	-
1.3450	2620	0.1115	-
1.3552	2640	0.1141	-
1.3655	2660	0.1325	-
1.3758	2680	0.1174	-
1.3860	2700	0.0973	-
1.3963	2720	0.1415	-
1.4066	2740	0.1691	-
1.4168	2760	0.1166	-
1.4271	2780	0.1349	-
1.4374	2800	0.1228	-
1.4476	2820	0.0988	-
1.4579	2840	0.1239	-
1.4682	2860	0.1255	-
1.4784	2880	0.1151	-
1.4887	2900	0.0715	-
1.4990	2920	0.1206	-
1.5092	2940	0.1042	-
1.5195	2960	0.1194	-
1.5298	2980	0.0994	-
1.5400	3000	0.1479	-
1.5503	3020	0.1188	-
1.5606	3040	0.1463	-
1.5708	3060	0.1245	-
1.5811	3080	0.1013	-
1.5914	3100	0.0991	-
1.6016	3120	0.0989	-
1.6119	3140	0.1648	-
1.6222	3160	0.1098	-
1.6324	3180	0.0953	-
1.6427	3200	0.1601	-
1.6530	3220	0.1237	-
1.6632	3240	0.1820	-
1.6735	3260	0.1152	-
1.6838	3280	0.1195	-
1.6940	3300	0.0973	-
1.7043	3320	0.0914	-
1.7146	3340	0.1002	-
1.7248	3360	0.1709	-
1.7351	3380	0.1032	-
1.7454	3400	0.1132	-
1.7556	3420	0.1182	-
1.7659	3440	0.1193	-
1.7762	3460	0.1011	-
1.7864	3480	0.1343	-
1.7967	3500	0.1555	-
1.8070	3520	0.1558	-
1.8172	3540	0.0966	-
1.8275	3560	0.1053	-
1.8378	3580	0.1604	-
1.8480	3600	0.1547	-
1.8583	3620	0.1147	-
1.8686	3640	0.1026	-
1.8789	3660	0.1250	-
1.8891	3680	0.1431	-
1.8994	3700	0.1058	-
1.9097	3720	0.1016	-
1.9199	3740	0.1298	-
1.9302	3760	0.1082	-
1.9405	3780	0.1271	-
1.9507	3800	0.0844	-
1.9610	3820	0.1266	-
1.9713	3840	0.1311	-
1.9815	3860	0.0937	-
1.9918	3880	0.0833	-
2.0	3896	-	0.3104

The bold row denotes the saved checkpoint.

Training Time

Training: 4.6 minutes
Evaluation: 4.6 seconds
Total: 4.7 minutes

Framework Versions

Python: 3.13.13
Sentence Transformers: 5.5.0
Transformers: 5.8.1
PyTorch: 2.12.0
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

bibtex

@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",
}

	sentence_A	sentence_B	label
type	string	string	float
modality	text	text
details	min: 7 tokens mean: 8.14 tokens max: 11 tokens	min: 6 tokens mean: 7.94 tokens max: 15 tokens	min: 0.0 mean: 0.22 max: 1.0

rohit-vt/tink-reranker-phase4_minilm_real

CrossEncoder based on cross-encoder/ms-marco-MiniLM-L6-v2

Model Details

Model Description

Model Sources

Full Model Architecture

Usage

Direct Usage (Sentence Transformers)

Training Details

Training Dataset

Unnamed Dataset

Evaluation Dataset

Unnamed Dataset

Training Hyperparameters

Non-Default Hyperparameters

All Hyperparameters

Training Logs

Training Time

Framework Versions

Additional Resources

Citation

BibTeX

Sentence Transformers

rohit-vt/
tink-reranker-phase4_minilm_real