metadata

base_model: sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
datasets: []
language: []
library_name: sentence-transformers
metrics:
  - cosine_accuracy
  - cosine_accuracy_threshold
  - cosine_f1
  - cosine_f1_threshold
  - cosine_precision
  - cosine_recall
  - cosine_ap
  - dot_accuracy
  - dot_accuracy_threshold
  - dot_f1
  - dot_f1_threshold
  - dot_precision
  - dot_recall
  - dot_ap
  - manhattan_accuracy
  - manhattan_accuracy_threshold
  - manhattan_f1
  - manhattan_f1_threshold
  - manhattan_precision
  - manhattan_recall
  - manhattan_ap
  - euclidean_accuracy
  - euclidean_accuracy_threshold
  - euclidean_f1
  - euclidean_f1_threshold
  - euclidean_precision
  - euclidean_recall
  - euclidean_ap
  - max_accuracy
  - max_accuracy_threshold
  - max_f1
  - max_f1_threshold
  - max_precision
  - max_recall
  - max_ap
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:410745
  - loss:ContrastiveLoss
widget:
  - source_sentence: وینچ
    sentences:
      - >-
        ترقه شکلاتی ( هفت ترقه ) ناریه پارس درجه 1 بسته 15 عددی ترقه شکلاتی (
        هفت ترقه ) ناریه پارس درجه 1 بسته 15 عددی 10عدد ناریه ترقه شکلاتی هفت
        ترقه بار تازه بدون رطوبت وخرابی مارک معتبر نورافشانی
      - پارچه میکرو کجراه
      - >-
        Car winch-1500LBS-KARA وینچ خودرو آفرود ۶۸۰ کیلوگرم کارا ۱۵۰۰lbs وینچ
        خودرویی (جلو ماشینی) 1500LBS کارا (KARA)
  - source_sentence: ' وسپا '
    sentences:
      - پولوشرت زرد وسپا
      - دوچرخه بند سقفی  لیفان X70 ایکس 70 آلومینیومی طرح منابو
      - >-
        دوچرخه ویوا Oxygen سایز 26 دوچرخه 26 ويوا OXYGEN دوچرخه کوهستان ویوا مدل
        OXYGEN سایز 26
  - source_sentence: دوچرخه المپیا سایز 27 5
    sentences:
      - "دوچرخه شهری المپیا کد 16220 سایز 16 دوچرخه شهری المپیا\_کد 16220 سایز 16 دوچرخه المپیا کد 16220 سایز 16 - OLYMPIA"
      - لامپ اس ام دی خودرو مدل 8B بسته 2 عددی
      - قیمت کمپرس سنج موتور
  - source_sentence: دچرخه ی
    sentences:
      - هیدروفیشیال ۷ کاره نیوفیس پلاس متور سنگین ۲۰۲۲
      - جامدادی کیوت
      - جعبه ی کادو ی رنگی
  - source_sentence: هایومکس
    sentences:
      - انگشتر حدید صینی کد2439
      - ژل هایومکس ولومایزر 2 سی سی
      - >-
        دزدگیر پاناتک مدل P-CA501 دزدگیر پاناتک P-CA501-2 دزدگیر پاناتک مدل
        P-CA501-2
model-index:
  - name: >-
      SentenceTransformer based on
      sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
    results:
      - task:
          type: binary-classification
          name: Binary Classification
        dataset:
          name: Unknown
          type: unknown
        metrics:
          - type: cosine_accuracy
            value: 0.8396327702184535
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.7623803019523621
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.8951804502771806
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.7234876751899719
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.8454428891975638
            name: Cosine Precision
          - type: cosine_recall
            value: 0.9511359538406059
            name: Cosine Recall
          - type: cosine_ap
            value: 0.9296495014804667
            name: Cosine Ap
          - type: dot_accuracy
            value: 0.8127916913166371
            name: Dot Accuracy
          - type: dot_accuracy_threshold
            value: 18.16492462158203
            name: Dot Accuracy Threshold
          - type: dot_f1
            value: 0.8798154233377613
            name: Dot F1
          - type: dot_f1_threshold
            value: 17.596263885498047
            name: Dot F1 Threshold
          - type: dot_precision
            value: 0.82272025942101
            name: Dot Precision
          - type: dot_recall
            value: 0.9454261329486717
            name: Dot Recall
          - type: dot_ap
            value: 0.9138496334192171
            name: Dot Ap
          - type: manhattan_accuracy
            value: 0.8362584631565109
            name: Manhattan Accuracy
          - type: manhattan_accuracy_threshold
            value: 56.61064910888672
            name: Manhattan Accuracy Threshold
          - type: manhattan_f1
            value: 0.892930089729684
            name: Manhattan F1
          - type: manhattan_f1_threshold
            value: 60.147003173828125
            name: Manhattan F1 Threshold
          - type: manhattan_precision
            value: 0.8403818109505502
            name: Manhattan Precision
          - type: manhattan_recall
            value: 0.9524882798413271
            name: Manhattan Recall
          - type: manhattan_ap
            value: 0.9274603777518026
            name: Manhattan Ap
          - type: euclidean_accuracy
            value: 0.8366528626832315
            name: Euclidean Accuracy
          - type: euclidean_accuracy_threshold
            value: 3.691666603088379
            name: Euclidean Accuracy Threshold
          - type: euclidean_f1
            value: 0.8933491652479936
            name: Euclidean F1
          - type: euclidean_f1_threshold
            value: 3.691666603088379
            name: Euclidean F1 Threshold
          - type: euclidean_precision
            value: 0.8525051194539249
            name: Euclidean Precision
          - type: euclidean_recall
            value: 0.9383038826782065
            name: Euclidean Recall
          - type: euclidean_ap
            value: 0.9275301813554955
            name: Euclidean Ap
          - type: max_accuracy
            value: 0.8396327702184535
            name: Max Accuracy
          - type: max_accuracy_threshold
            value: 56.61064910888672
            name: Max Accuracy Threshold
          - type: max_f1
            value: 0.8951804502771806
            name: Max F1
          - type: max_f1_threshold
            value: 60.147003173828125
            name: Max F1 Threshold
          - type: max_precision
            value: 0.8525051194539249
            name: Max Precision
          - type: max_recall
            value: 0.9524882798413271
            name: Max Recall
          - type: max_ap
            value: 0.9296495014804667
            name: Max Ap
          - type: cosine_accuracy
            value: 0.831416113411775
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.7449432611465454
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.8897548675482456
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.7427525520324707
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.8502039810530351
            name: Cosine Precision
          - type: cosine_recall
            value: 0.9331650438754658
            name: Cosine Recall
          - type: cosine_ap
            value: 0.9252554285491397
            name: Cosine Ap
          - type: dot_accuracy
            value: 0.8083437410986218
            name: Dot Accuracy
          - type: dot_accuracy_threshold
            value: 18.16763687133789
            name: Dot Accuracy Threshold
          - type: dot_f1
            value: 0.8761684843089249
            name: Dot F1
          - type: dot_f1_threshold
            value: 17.106109619140625
            name: Dot F1 Threshold
          - type: dot_precision
            value: 0.8156272661348803
            name: Dot Precision
          - type: dot_recall
            value: 0.9464178386825339
            name: Dot Recall
          - type: dot_ap
            value: 0.9078782883891188
            name: Dot Ap
          - type: manhattan_accuracy
            value: 0.827735051162383
            name: Manhattan Accuracy
          - type: manhattan_accuracy_threshold
            value: 53.94535446166992
            name: Manhattan Accuracy Threshold
          - type: manhattan_f1
            value: 0.887467671202069
            name: Manhattan F1
          - type: manhattan_f1_threshold
            value: 59.66460418701172
            name: Manhattan F1 Threshold
          - type: manhattan_precision
            value: 0.8336590260906306
            name: Manhattan Precision
          - type: manhattan_recall
            value: 0.9487017670393076
            name: Manhattan Recall
          - type: manhattan_ap
            value: 0.9230969972500983
            name: Manhattan Ap
          - type: euclidean_accuracy
            value: 0.8274282959749337
            name: Euclidean Accuracy
          - type: euclidean_accuracy_threshold
            value: 3.4869043827056885
            name: Euclidean Accuracy Threshold
          - type: euclidean_f1
            value: 0.8874656133173449
            name: Euclidean F1
          - type: euclidean_f1_threshold
            value: 3.7965426445007324
            name: Euclidean F1 Threshold
          - type: euclidean_precision
            value: 0.8363423648594751
            name: Euclidean Precision
          - type: euclidean_recall
            value: 0.9452458228152422
            name: Euclidean Recall
          - type: euclidean_ap
            value: 0.9231713715918721
            name: Euclidean Ap
          - type: max_accuracy
            value: 0.831416113411775
            name: Max Accuracy
          - type: max_accuracy_threshold
            value: 53.94535446166992
            name: Max Accuracy Threshold
          - type: max_f1
            value: 0.8897548675482456
            name: Max F1
          - type: max_f1_threshold
            value: 59.66460418701172
            name: Max F1 Threshold
          - type: max_precision
            value: 0.8502039810530351
            name: Max Precision
          - type: max_recall
            value: 0.9487017670393076
            name: Max Recall
          - type: max_ap
            value: 0.9252554285491397
            name: Max Ap

SentenceTransformer based on sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2

This is a sentence-transformers model finetuned from sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2. It maps sentences & paragraphs to a 384-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: sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
Maximum Sequence Length: 128 tokens
Output Dimensionality: 384 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': 128, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 384, '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("DashReza7/sentence-transformers_paraphrase-multilingual-MiniLM-L12-v2_FINETUNED_on_torob_data_v4")
# Run inference
sentences = [
    'هایومکس',
    'ژل هایومکس ولومایزر 2 سی سی',
    'دزدگیر پاناتک مدل P-CA501 دزدگیر پاناتک P-CA501-2 دزدگیر پاناتک مدل P-CA501-2',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

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

Evaluation

Metrics

Binary Classification

Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.8396
cosine_accuracy_threshold	0.7624
cosine_f1	0.8952
cosine_f1_threshold	0.7235
cosine_precision	0.8454
cosine_recall	0.9511
cosine_ap	0.9296
dot_accuracy	0.8128
dot_accuracy_threshold	18.1649
dot_f1	0.8798
dot_f1_threshold	17.5963
dot_precision	0.8227
dot_recall	0.9454
dot_ap	0.9138
manhattan_accuracy	0.8363
manhattan_accuracy_threshold	56.6106
manhattan_f1	0.8929
manhattan_f1_threshold	60.147
manhattan_precision	0.8404
manhattan_recall	0.9525
manhattan_ap	0.9275
euclidean_accuracy	0.8367
euclidean_accuracy_threshold	3.6917
euclidean_f1	0.8933
euclidean_f1_threshold	3.6917
euclidean_precision	0.8525
euclidean_recall	0.9383
euclidean_ap	0.9275
max_accuracy	0.8396
max_accuracy_threshold	56.6106
max_f1	0.8952
max_f1_threshold	60.147
max_precision	0.8525
max_recall	0.9525
max_ap	0.9296

Binary Classification

Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.8314
cosine_accuracy_threshold	0.7449
cosine_f1	0.8898
cosine_f1_threshold	0.7428
cosine_precision	0.8502
cosine_recall	0.9332
cosine_ap	0.9253
dot_accuracy	0.8083
dot_accuracy_threshold	18.1676
dot_f1	0.8762
dot_f1_threshold	17.1061
dot_precision	0.8156
dot_recall	0.9464
dot_ap	0.9079
manhattan_accuracy	0.8277
manhattan_accuracy_threshold	53.9454
manhattan_f1	0.8875
manhattan_f1_threshold	59.6646
manhattan_precision	0.8337
manhattan_recall	0.9487
manhattan_ap	0.9231
euclidean_accuracy	0.8274
euclidean_accuracy_threshold	3.4869
euclidean_f1	0.8875
euclidean_f1_threshold	3.7965
euclidean_precision	0.8363
euclidean_recall	0.9452
euclidean_ap	0.9232
max_accuracy	0.8314
max_accuracy_threshold	53.9454
max_f1	0.8898
max_f1_threshold	59.6646
max_precision	0.8502
max_recall	0.9487
max_ap	0.9253

Training Details

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 128
per_device_eval_batch_size: 128
learning_rate: 2e-05
num_train_epochs: 1
warmup_ratio: 0.1
fp16: True

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: steps
prediction_loss_only: True
per_device_train_batch_size: 128
per_device_eval_batch_size: 128
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 1
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: 1
max_steps: -1
lr_scheduler_type: linear
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
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: 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
batch_sampler: batch_sampler
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	loss	max_ap
None	0	-	-	0.8131
0.1558	500	0.0262	-	-
0.3116	1000	0.0184	-	-
0.4674	1500	0.0173	-	-
0.6232	2000	0.0164	0.0155	0.9253
0.7791	2500	0.016	-	-
0.9349	3000	0.0155	-	-
1.0	3209	-	-	0.9296

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.0.1
Transformers: 4.42.4
PyTorch: 2.4.0+cu121
Accelerate: 0.32.1
Datasets: 2.21.0
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",
}

ContrastiveLoss

@inproceedings{hadsell2006dimensionality,
    author={Hadsell, R. and Chopra, S. and LeCun, Y.},
    booktitle={2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06)}, 
    title={Dimensionality Reduction by Learning an Invariant Mapping}, 
    year={2006},
    volume={2},
    number={},
    pages={1735-1742},
    doi={10.1109/CVPR.2006.100}
}