This is a sentence-transformers model finetuned from deepvk/USER-bge-m3. It maps sentences & paragraphs to a 1024-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: deepvk/USER-bge-m3
Maximum Sequence Length: 512 tokens
Output Dimensionality: 1024 tokens
Similarity Function: Cosine Similarity

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}) with Transformer model: XLMRobertaModel 
  (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, '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("Data-Lab/USER-bge-m3-embedder-td")
# Run inference
sentences = [
    'детская каша',
    'Каша овсяная детская "Мишка" Сладкая овсяная каша с голубикой и бананами. Можно приготовить на кокосовом молоке',
    'Десерт "Тирамису", 300 г Изысканный итальянский десерт в нестандартном исполнении. В нашем Тирамису много &lpar;очень много!&rpar; сливочного крема и Маскарпоне, поэтому лакомство невероятно нежное!',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Triplet

Dataset: dev
Evaluated with TripletEvaluator

Metric	Value
cosine_accuracy	0.9188
dot_accuracy	0.0803
manhattan_accuracy	0.917
euclidean_accuracy	0.9188
max_accuracy	0.9188

Training Details

Training Dataset

Unnamed Dataset

Size: 10,189 training samples
Columns: sentence_0, sentence_1, and sentence_2
Approximate statistics based on the first 1000 samples:
sentence_0 sentence_1 sentence_2
type string string string
details
min: 3 tokens
mean: 7.85 tokens
max: 30 tokens
min: 6 tokens
mean: 61.74 tokens
max: 377 tokens
min: 5 tokens
mean: 64.71 tokens
max: 393 tokens

Samples:

sentence_0	sentence_1	sentence_2
`хурма`	`Чипсы из хурмы, 25 г Натуральные чипсы из хурмы, без сахара. Мягкие, медово-фруктовые`	`Салат мимоза, 300 г Классический салат мимоза с горбушей, отварными овощами и куриными желтками.`
`жареное мясо`	`КК_котлета куриная жареная, вес`	Баклажаны "Пармиджано" Мама миа, это же настоящая итальянская пармиджана! Нежные ломтики баклажанов, много томатов и ещё больше тягучего сыра. Очень насыщенно, сочно и аппетитно пряно. Баклажаны для этого рецепта не обжариваются, а запекаются в духовке, что делает блюдо более полезным и изысканным.
`бедро цыпленка бройлера`	`Бедро цыплят-бройлеров Халяль 1 кг Сочное бедро цыпленка, подходит для маринования, тушения и запекания`	`Мясо бедра (Филе бедра) индейки в маринаде "Чесночный" 1 кг Диетическое, нежирное филе бедра индейки с деликатным вкусом и ароматом. В меру подсолено и приправлено острым чесночком и травами.`

Loss: TripletLoss with these parameters:
json
1{ 2 "distance_metric": "TripletDistanceMetric.COSINE", 3 "triplet_margin": 0.5 4}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 4
per_device_eval_batch_size: 4
fp16: True
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: steps
prediction_loss_only: True
per_device_train_batch_size: 4
per_device_eval_batch_size: 4
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: 3
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
use_ipex: 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: True
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}
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: False
hub_always_push: False
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
include_inputs_for_metrics: False
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
eval_use_gather_object: False
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	Training Loss	dev_max_accuracy
0.3928	500	0.2477	-
0.7855	1000	0.182	0.9064
1.0	1273	-	0.9073
1.1783	1500	0.157	-
1.5711	2000	0.1234	0.9029
1.9639	2500	0.0993	-
2.0	2546	-	0.9179
2.3566	3000	0.0864	0.9170
2.7494	3500	0.0691	-
3.0	3819	-	0.9188

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.2.0
Transformers: 4.44.0
PyTorch: 2.3.1+cu121
Accelerate: 0.31.0
Datasets: 2.20.0
Tokenizers: 0.19.1

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",
}

TripletLoss

bibtex

@misc{hermans2017defense,
    title={In Defense of the Triplet Loss for Person Re-Identification},
    author={Alexander Hermans and Lucas Beyer and Bastian Leibe},
    year={2017},
    eprint={1703.07737},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

	sentence_0	sentence_1	sentence_2
type	string	string	string
details	min: 3 tokens mean: 7.85 tokens max: 30 tokens	min: 6 tokens mean: 61.74 tokens max: 377 tokens	min: 5 tokens mean: 64.71 tokens max: 393 tokens