You can deploy this model and start earning money today!

Views

No views yet

SentenceTransformer based on intfloat/multilingual-e5-base

Name: sentence-transformer
Rating: 2.5 (70 reviews)

This is a sentence-transformers model finetuned from intfloat/multilingual-e5-base on the train dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

Model Type: Sentence Transformer
Base model: intfloat/multilingual-e5-base
Maximum Sequence Length: 512 tokens
Output Dimensionality: 768 dimensions
Similarity Function: Cosine Similarity
Training Dataset:
- train

Model Sources

Documentation: Sentence Transformers Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False, 'architecture': 'XLMRobertaModel'})
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
  (2): Normalize()
)

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 SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
sentences = [
    'Прием осуществляется только по результатам дополнительных вступительных испытаний (ДВИ профильной, творческой и/или профессиональной направленности), проводимых МГУ в 2022 году.',
    'Как осуществляется прием по специальной квоте для детей военнослужащих и сотрудников, погибших или получивших увечье, в МГУ в 2022 году?',
    'На каком основании осуществляется отнесение поступающих к числу детей военнослужащих и сотрудников в пределах специальной квоты?',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.8577, 0.7622],
#         [0.8577, 1.0000, 0.8551],
#         [0.7622, 0.8551, 1.0000]])

Evaluation

Metrics

Information Retrieval

Dataset: dim_768
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 768 3}

Metric	Value
cosine_accuracy@1	0.4091
cosine_accuracy@3	0.5455
cosine_accuracy@5	0.6818
cosine_accuracy@10	0.9091
cosine_precision@1	0.4091
cosine_precision@3	0.1818
cosine_precision@5	0.1364
cosine_precision@10	0.0909
cosine_recall@1	0.4091
cosine_recall@3	0.5455
cosine_recall@5	0.6818
cosine_recall@10	0.9091
cosine_ndcg@10	0.6219
cosine_mrr@10	0.5343
cosine_map@100	0.54

Information Retrieval

Dataset: dim_512
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 512 3}

Metric	Value
cosine_accuracy@1	0.4091
cosine_accuracy@3	0.5455
cosine_accuracy@5	0.6818
cosine_accuracy@10	0.8636
cosine_precision@1	0.4091
cosine_precision@3	0.1818
cosine_precision@5	0.1364
cosine_precision@10	0.0864
cosine_recall@1	0.4091
cosine_recall@3	0.5455
cosine_recall@5	0.6818
cosine_recall@10	0.8636
cosine_ndcg@10	0.6067
cosine_mrr@10	0.5278
cosine_map@100	0.5366

Information Retrieval

Dataset: dim_256
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 256 3}

Metric	Value
cosine_accuracy@1	0.4091
cosine_accuracy@3	0.5455
cosine_accuracy@5	0.7727
cosine_accuracy@10	0.9091
cosine_precision@1	0.4091
cosine_precision@3	0.1818
cosine_precision@5	0.1545
cosine_precision@10	0.0909
cosine_recall@1	0.4091
cosine_recall@3	0.5455
cosine_recall@5	0.7727
cosine_recall@10	0.9091
cosine_ndcg@10	0.6197
cosine_mrr@10	0.5319
cosine_map@100	0.5387

Information Retrieval

Dataset: dim_128
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 128 3}

Metric	Value
cosine_accuracy@1	0.4091
cosine_accuracy@3	0.5455
cosine_accuracy@5	0.6364
cosine_accuracy@10	0.8636
cosine_precision@1	0.4091
cosine_precision@3	0.1818
cosine_precision@5	0.1273
cosine_precision@10	0.0864
cosine_recall@1	0.4091
cosine_recall@3	0.5455
cosine_recall@5	0.6364
cosine_recall@10	0.8636
cosine_ndcg@10	0.5941
cosine_mrr@10	0.5131
cosine_map@100	0.5219

Information Retrieval

Dataset: dim_64
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 64 3}

Metric	Value
cosine_accuracy@1	0.5
cosine_accuracy@3	0.6364
cosine_accuracy@5	0.7727
cosine_accuracy@10	0.9091
cosine_precision@1	0.5
cosine_precision@3	0.2121
cosine_precision@5	0.1545
cosine_precision@10	0.0909
cosine_recall@1	0.5
cosine_recall@3	0.6364
cosine_recall@5	0.7727
cosine_recall@10	0.9091
cosine_ndcg@10	0.6675
cosine_mrr@10	0.594
cosine_map@100	0.5989

Information Retrieval

Dataset: dim_768
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 768 3}

Metric	Value
cosine_accuracy@1	0.5455
cosine_accuracy@3	0.9091
cosine_accuracy@5	0.9091
cosine_accuracy@10	1.0
cosine_precision@1	0.5455
cosine_precision@3	0.303
cosine_precision@5	0.1818
cosine_precision@10	0.1
cosine_recall@1	0.5455
cosine_recall@3	0.9091
cosine_recall@5	0.9091
cosine_recall@10	1.0
cosine_ndcg@10	0.7865
cosine_mrr@10	0.7167
cosine_map@100	0.7167

Information Retrieval

Dataset: dim_512
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 512 3}

Metric	Value
cosine_accuracy@1	0.6364
cosine_accuracy@3	0.8636
cosine_accuracy@5	0.9091
cosine_accuracy@10	1.0
cosine_precision@1	0.6364
cosine_precision@3	0.2879
cosine_precision@5	0.1818
cosine_precision@10	0.1
cosine_recall@1	0.6364
cosine_recall@3	0.8636
cosine_recall@5	0.9091
cosine_recall@10	1.0
cosine_ndcg@10	0.8169
cosine_mrr@10	0.7584
cosine_map@100	0.7584

Information Retrieval

Dataset: dim_256
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 256 3}

Metric	Value
cosine_accuracy@1	0.6364
cosine_accuracy@3	0.9545
cosine_accuracy@5	0.9545
cosine_accuracy@10	1.0
cosine_precision@1	0.6364
cosine_precision@3	0.3182
cosine_precision@5	0.1909
cosine_precision@10	0.1
cosine_recall@1	0.6364
cosine_recall@3	0.9545
cosine_recall@5	0.9545
cosine_recall@10	1.0
cosine_ndcg@10	0.8414
cosine_mrr@10	0.7879
cosine_map@100	0.7879

Information Retrieval

Dataset: dim_128
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 128 3}

Metric	Value
cosine_accuracy@1	0.5909
cosine_accuracy@3	1.0
cosine_accuracy@5	1.0
cosine_accuracy@10	1.0
cosine_precision@1	0.5909
cosine_precision@3	0.3333
cosine_precision@5	0.2
cosine_precision@10	0.1
cosine_recall@1	0.5909
cosine_recall@3	1.0
cosine_recall@5	1.0
cosine_recall@10	1.0
cosine_ndcg@10	0.8133
cosine_mrr@10	0.75
cosine_map@100	0.75

Information Retrieval

Dataset: dim_64
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 64 3}

Metric	Value
cosine_accuracy@1	0.8182
cosine_accuracy@3	0.9545
cosine_accuracy@5	1.0
cosine_accuracy@10	1.0
cosine_precision@1	0.8182
cosine_precision@3	0.3182
cosine_precision@5	0.2
cosine_precision@10	0.1
cosine_recall@1	0.8182
cosine_recall@3	0.9545
cosine_recall@5	1.0
cosine_recall@10	1.0
cosine_ndcg@10	0.9159
cosine_mrr@10	0.8879
cosine_map@100	0.8879

Training Details

Training Dataset

train

Dataset: train
Size: 22 training samples
Columns: positive and anchor
Approximate statistics based on the first 22 samples:
positive anchor
type string string
details
min: 8 tokens
mean: 38.55 tokens
max: 90 tokens
min: 12 tokens
mean: 25.95 tokens
max: 39 tokens

Samples:

positive	anchor
`The provided text is empty, so there is no main topic to discuss.`	`What is the main topic of the provided text?`
На прием в пределах специальной квоты имеют право дети военнослужащих и сотрудников федеральных органов исполнительной власти и государственных органов, где предусмотрена военная служба, а также сотрудников органов внутренних дел РФ, которые участвовали или участвуют в специальной военной операции на территориях Донецкой Народной Республики, Луганской Народной Республики и Украины, в том числе погибших (умерших) при исполнении обязанностей.	`Какие категории лиц имеют право на прием в пределах специальной квоты в МГУ в 2022 году согласно представленным Особенностям?`
`Указ № 268 с учетом указанного приказа распространяется на прием на обучение по программам высшего образования – бакалавриата, магистратуры и специалитета.`	`На какие образовательные программы распространяется Указ № 268 с учетом приказа Минобрнауки от 21 августа 2020 г.?`

Loss: MatryoshkaLoss with these parameters:
json
1{ 2 "loss": "MultipleNegativesRankingLoss", 3 "matryoshka_dims": [ 4 768, 5 512, 6 256, 7 128, 8 64 9 ], 10 "matryoshka_weights": [ 11 1, 12 1, 13 1, 14 1, 15 1 16 ], 17 "n_dims_per_step": -1 18}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: epoch
per_device_train_batch_size: 4
per_device_eval_batch_size: 16
gradient_accumulation_steps: 16
learning_rate: 2e-05
num_train_epochs: 5
lr_scheduler_type: cosine
warmup_ratio: 0.1
bf16: True
load_best_model_at_end: True
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: epoch
prediction_loss_only: True
per_device_train_batch_size: 4
per_device_eval_batch_size: 16
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 16
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: 5
max_steps: -1
lr_scheduler_type: cosine
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: 42
data_seed: None
jit_mode_eval: 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: 0
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}
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: no_duplicates
multi_dataset_batch_sampler: proportional
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	dim_768_cosine_ndcg@10	dim_512_cosine_ndcg@10	dim_256_cosine_ndcg@10	dim_128_cosine_ndcg@10	dim_64_cosine_ndcg@10
-1	-1	0.6232	0.6067	0.6197	0.5875	0.6656
1.0	1	0.6219	0.6067	0.6197	0.5941	0.6675
-1	-1	0.6219	0.6067	0.6197	0.5941	0.6675
1.0	1	0.6219	0.6067	0.6197	0.5941	0.6675
2.0	2	0.6763	0.6801	0.7014	0.6618	0.7527
3.0	3	0.7702	0.7747	0.7452	0.7744	0.8252
4.0	4	0.7834	0.7995	0.8285	0.7979	0.8900
5.0	5	0.7865	0.8169	0.8414	0.8133	0.9159

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.12.12
Sentence Transformers: 5.1.2
Transformers: 4.57.1
PyTorch: 2.8.0+cu126
Accelerate: 1.11.0
Datasets: 4.0.0
Tokenizers: 0.22.1

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",
}

MatryoshkaLoss

bibtex

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

MultipleNegativesRankingLoss

bibtex

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

	positive	anchor
type	string	string
details	min: 8 tokens mean: 38.55 tokens max: 90 tokens	min: 12 tokens mean: 25.95 tokens max: 39 tokens

Renan1997/sentence-transformer

SentenceTransformer based on intfloat/multilingual-e5-base

Model Details

Model Description

Model Sources

Full Model Architecture

Usage

Direct Usage (Sentence Transformers)

Evaluation

Metrics

Information Retrieval

Information Retrieval

Information Retrieval

Information Retrieval

Information Retrieval

Information Retrieval

Information Retrieval

Information Retrieval

Information Retrieval

Information Retrieval

Training Details

Training Dataset

train

Training Hyperparameters

Non-Default Hyperparameters

All Hyperparameters

Training Logs

Framework Versions

Citation

BibTeX

Sentence Transformers

MatryoshkaLoss

MultipleNegativesRankingLoss

Renan1997/
sentence-transformer