metadata

base_model: BAAI/bge-m3
datasets: []
language: []
library_name: sentence-transformers
metrics:
  - pearson_cosine
  - spearman_cosine
  - pearson_manhattan
  - spearman_manhattan
  - pearson_euclidean
  - spearman_euclidean
  - pearson_dot
  - spearman_dot
  - pearson_max
  - spearman_max
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:4532
  - loss:CoSENTLoss
widget:
  - source_sentence: гантели грифы штанги гири
    sentences:
      - гири
      - коммутатор poe web настраиваемый utp3526ts-psb
      - игровой монитор lg xg2705
  - source_sentence: vt vt9602
    sentences:
      - подгрифок для скрипки 1 4 wittner ultra 918141
      - электросамокат white siberia nerpa pro 3600w 2023 elka зеленый
      - компьютер pc itmultra 2 v 2
  - source_sentence: фен dyson supersonic hd08 replika
    sentences:
      - стабилизатор smooth-x combo белый
      - dyson supersonic hd08 replika
      - ip-dal30ir0280p ver2
  - source_sentence: aresa ar-4205
    sentences:
      - холодильник olto rf-140 c  черный
      - aresa ar-3905
      - champion g200vk-1
  - source_sentence: букеты шаров сеты для детей
    sentences:
      - букеты шаров сеты для него
      - дрипка geekvape loop rda
      - труба гладкая жесткая 16 мм 3 м
model-index:
  - name: SentenceTransformer based on BAAI/bge-m3
    results:
      - task:
          type: semantic-similarity
          name: Semantic Similarity
        dataset:
          name: sts dev
          type: sts-dev
        metrics:
          - type: pearson_cosine
            value: 0.9092748477762634
            name: Pearson Cosine
          - type: spearman_cosine
            value: 0.8959000349666695
            name: Spearman Cosine
          - type: pearson_manhattan
            value: 0.9103703525656046
            name: Pearson Manhattan
          - type: spearman_manhattan
            value: 0.8944672696951159
            name: Spearman Manhattan
          - type: pearson_euclidean
            value: 0.9102936678180418
            name: Pearson Euclidean
          - type: spearman_euclidean
            value: 0.8945285994969848
            name: Spearman Euclidean
          - type: pearson_dot
            value: 0.8951660474126123
            name: Pearson Dot
          - type: spearman_dot
            value: 0.8872903553527511
            name: Spearman Dot
          - type: pearson_max
            value: 0.9103703525656046
            name: Pearson Max
          - type: spearman_max
            value: 0.8959000349666695
            name: Spearman Max

SentenceTransformer based on BAAI/bge-m3

This is a sentence-transformers model finetuned from BAAI/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: BAAI/bge-m3
Maximum Sequence Length: 8192 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': 8192, 'do_lower_case': False}) with Transformer model: XLMRobertaModel 
  (1): Pooling({'word_embedding_dimension': 1024, '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})
)

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 SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("seregadgl101/test_bge_10ep")
# Run inference
sentences = [
    'букеты шаров сеты для детей',
    'букеты шаров сеты для него',
    'дрипка geekvape loop rda',
]
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

Semantic Similarity

Dataset: sts-dev
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.9093
spearman_cosine	0.8959
pearson_manhattan	0.9104
spearman_manhattan	0.8945
pearson_euclidean	0.9103
spearman_euclidean	0.8945
pearson_dot	0.8952
spearman_dot	0.8873
pearson_max	0.9104
spearman_max	0.8959

Training Details

Training Dataset

Unnamed Dataset

Size: 4,532 training samples
Columns: sentence1, sentence2, and score
Approximate statistics based on the first 1000 samples:
sentence1 sentence2 score
type string string float
details
min: 4 tokens
mean: 14.45 tokens
max: 48 tokens

min: 3 tokens
mean: 13.09 tokens
max: 51 tokens

min: 0.0
mean: 0.6
max: 1.0

	sentence1	sentence2	score
type	string	string	float
details	min: 4 tokens mean: 14.45 tokens max: 48 tokens	min: 3 tokens mean: 13.09 tokens max: 51 tokens	min: 0.0 mean: 0.6 max: 1.0

Samples:

sentence1	sentence2	score
`батут evo jump internal 12ft`	`батут evo jump internal 12ft`	`1.0`
`наручные часы orient casual`	`наручные часы orient`	`1.0`
`электрический духовой шкаф weissgauff eov 19 mw`	`электрический духовой шкаф weissgauff eov 19 mx`	`0.4`

Loss: CoSENTLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "pairwise_cos_sim"
}

Evaluation Dataset

Unnamed Dataset

Size: 504 evaluation samples
Columns: sentence1, sentence2, and score
Approximate statistics based on the first 1000 samples:
sentence1 sentence2 score
type string string float
details
min: 4 tokens
mean: 14.93 tokens
max: 48 tokens

min: 4 tokens
mean: 13.1 tokens
max: 40 tokens

min: 0.0
mean: 0.59
max: 1.0

	sentence1	sentence2	score
type	string	string	float
details	min: 4 tokens mean: 14.93 tokens max: 48 tokens	min: 4 tokens mean: 13.1 tokens max: 40 tokens	min: 0.0 mean: 0.59 max: 1.0

Samples:

sentence1	sentence2	score
`потолочный светильник yeelight smart led ceiling light c2001s500`	`yeelight smart led ceiling light c2001s500`	`1.0`
`канцелярские принадлежности`	`канцелярские принадлежности разные`	`0.4`
`usb-магнитола acv avs-1718g`	`автомагнитола acv avs-1718g`	`1.0`

Loss: CoSENTLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "pairwise_cos_sim"
}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
gradient_accumulation_steps: 16
learning_rate: 2e-05
num_train_epochs: 10
lr_scheduler_type: cosine
warmup_ratio: 0.1
save_only_model: True
fp16: 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: steps
prediction_loss_only: True
per_device_train_batch_size: 8
per_device_eval_batch_size: 8
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 16
eval_accumulation_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: 10
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: True
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: 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}
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
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	loss	sts-dev_spearman_cosine
1.4109	50	2.1693	0.7897
2.8219	100	2.3041	0.8553
4.2328	150	2.4628	0.8737
5.6437	200	2.5485	0.8877
7.0547	250	2.4879	0.8945
8.4656	300	2.5508	0.8955
9.8765	350	2.5626	0.8959

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.0.1
Transformers: 4.41.2
PyTorch: 2.1.2+cu121
Accelerate: 0.31.0
Datasets: 2.20.0
Tokenizers: 0.19.1

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",
}

CoSENTLoss

@online{kexuefm-8847,
    title={CoSENT: A more efficient sentence vector scheme than Sentence-BERT},
    author={Su Jianlin},
    year={2022},
    month={Jan},
    url={https://kexue.fm/archives/8847},
}