--- base_model: cointegrated/rubert-tiny2 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:13690 - loss:ContrastiveLoss widget: - source_sentence: Тренажер на свободных весах DFC HOMEGYM HM019 в Москве sentences: - Независимая бицепс-машина Matrix G7-S40 - Беговая дорожка Stingrey ST-9317 - Мультикомплекс Hasttings Digger HD003-7 - source_sentence: Велотренажер Clear Fit Envy CFB 45 Ego sentences: - Эллиптический тренажер Svensson Body Labs Strideline EXA в Москве - Реабилитационная рама ARMS (стек 60кг) AR082.1х60 - Вертикальная тяга двухпозиционная Vertex OPS 110 - source_sentence: Нижняя тяга Smith SH004 sentences: - Аэробайк BH FITNESS AIRMAG - Велотренажер Freemotion Tour De France Club - Жим ногами под углом 45 градусов Bronze Gym BG-BGR-801 - source_sentence: Эллиптический тренажер BEST FITNESS BFE1 sentences: - Спин-байк DFC OVICX Q200C - Эллиптический тренажер NordicTrack E11.6 (NTEVEL99813) - Беговая дорожка SPIRIT LW650 - source_sentence: Беговая дорожка Hasttings CT100 sentences: - Мини велотренажер с регулируемой высотой Bradex SF 0830 - Беговая дорожка Koenigsmann ML в Москве - Вертикальный велотренажер Sole B94 (2023) model-index: - name: SentenceTransformer based on cointegrated/rubert-tiny2 results: - task: type: binary-classification name: Binary Classification dataset: name: cv type: cv metrics: - type: cosine_accuracy value: 1.0 name: Cosine Accuracy - type: cosine_accuracy_threshold value: 0.7653387784957886 name: Cosine Accuracy Threshold - type: cosine_f1 value: 1.0 name: Cosine F1 - type: cosine_f1_threshold value: 0.7653387784957886 name: Cosine F1 Threshold - type: cosine_precision value: 1.0 name: Cosine Precision - type: cosine_recall value: 1.0 name: Cosine Recall - type: cosine_ap value: 1.0 name: Cosine Ap - type: dot_accuracy value: 1.0 name: Dot Accuracy - type: dot_accuracy_threshold value: 0.7653387784957886 name: Dot Accuracy Threshold - type: dot_f1 value: 1.0 name: Dot F1 - type: dot_f1_threshold value: 0.7653387784957886 name: Dot F1 Threshold - type: dot_precision value: 1.0 name: Dot Precision - type: dot_recall value: 1.0 name: Dot Recall - type: dot_ap value: 1.0 name: Dot Ap - type: manhattan_accuracy value: 1.0 name: Manhattan Accuracy - type: manhattan_accuracy_threshold value: 9.330949783325195 name: Manhattan Accuracy Threshold - type: manhattan_f1 value: 1.0 name: Manhattan F1 - type: manhattan_f1_threshold value: 9.330949783325195 name: Manhattan F1 Threshold - type: manhattan_precision value: 1.0 name: Manhattan Precision - type: manhattan_recall value: 1.0 name: Manhattan Recall - type: manhattan_ap value: 1.0 name: Manhattan Ap - type: euclidean_accuracy value: 1.0 name: Euclidean Accuracy - type: euclidean_accuracy_threshold value: 0.6849288940429688 name: Euclidean Accuracy Threshold - type: euclidean_f1 value: 1.0 name: Euclidean F1 - type: euclidean_f1_threshold value: 0.6849288940429688 name: Euclidean F1 Threshold - type: euclidean_precision value: 1.0 name: Euclidean Precision - type: euclidean_recall value: 1.0 name: Euclidean Recall - type: euclidean_ap value: 1.0 name: Euclidean Ap - type: max_accuracy value: 1.0 name: Max Accuracy - type: max_accuracy_threshold value: 9.330949783325195 name: Max Accuracy Threshold - type: max_f1 value: 1.0 name: Max F1 - type: max_f1_threshold value: 9.330949783325195 name: Max F1 Threshold - type: max_precision value: 1.0 name: Max Precision - type: max_recall value: 1.0 name: Max Recall - type: max_ap value: 1.0 name: Max Ap --- # SentenceTransformer based on cointegrated/rubert-tiny2 This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [cointegrated/rubert-tiny2](https://huggingface.co/cointegrated/rubert-tiny2). It maps sentences & paragraphs to a 312-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:** [cointegrated/rubert-tiny2](https://huggingface.co/cointegrated/rubert-tiny2) - **Maximum Sequence Length:** 2048 tokens - **Output Dimensionality:** 312 tokens - **Similarity Function:** Cosine Similarity ### Model Sources - **Documentation:** [Sentence Transformers Documentation](https://sbert.net) - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers) - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers) ### Full Model Architecture ``` SentenceTransformer( (0): Transformer({'max_seq_length': 2048, 'do_lower_case': False}) with Transformer model: BertModel (1): Pooling({'word_embedding_dimension': 312, '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: ```bash 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 = [ 'Беговая дорожка Hasttings CT100', 'Вертикальный велотренажер Sole B94 (2023)', 'Беговая дорожка Koenigsmann ML в Москве', ] embeddings = model.encode(sentences) print(embeddings.shape) # [3, 312] # Get the similarity scores for the embeddings similarities = model.similarity(embeddings, embeddings) print(similarities.shape) # [3, 3] ``` ## Evaluation ### Metrics #### Binary Classification * Dataset: `cv` * Evaluated with [BinaryClassificationEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.BinaryClassificationEvaluator) | Metric | Value | |:-----------------------------|:--------| | cosine_accuracy | 1.0 | | cosine_accuracy_threshold | 0.7653 | | cosine_f1 | 1.0 | | cosine_f1_threshold | 0.7653 | | cosine_precision | 1.0 | | cosine_recall | 1.0 | | cosine_ap | 1.0 | | dot_accuracy | 1.0 | | dot_accuracy_threshold | 0.7653 | | dot_f1 | 1.0 | | dot_f1_threshold | 0.7653 | | dot_precision | 1.0 | | dot_recall | 1.0 | | dot_ap | 1.0 | | manhattan_accuracy | 1.0 | | manhattan_accuracy_threshold | 9.3309 | | manhattan_f1 | 1.0 | | manhattan_f1_threshold | 9.3309 | | manhattan_precision | 1.0 | | manhattan_recall | 1.0 | | manhattan_ap | 1.0 | | euclidean_accuracy | 1.0 | | euclidean_accuracy_threshold | 0.6849 | | euclidean_f1 | 1.0 | | euclidean_f1_threshold | 0.6849 | | euclidean_precision | 1.0 | | euclidean_recall | 1.0 | | euclidean_ap | 1.0 | | max_accuracy | 1.0 | | max_accuracy_threshold | 9.3309 | | max_f1 | 1.0 | | max_f1_threshold | 9.3309 | | max_precision | 1.0 | | max_recall | 1.0 | | **max_ap** | **1.0** | ## Training Details ### Training Dataset #### Unnamed Dataset * Size: 13,690 training samples * Columns: sentence1, sentence2, and score * Approximate statistics based on the first 1000 samples: | | sentence1 | sentence2 | score | |:--------|:---------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------| | type | string | string | float | | details | | | | * Samples: | sentence1 | sentence2 | score | |:-----------------------------------------------------------------------------------|:-------------------------------------------------------------------|:-----------------| | Велотренажер аэродинамический Spirit Fitness AB900+ Air Bike в Москве | Баттерфляй / Задняя дельта Impulse ExoForm FE9715 | 0.0 | | Эллиптический тренажер Sports Art E835 | Эллиптический тренажер Clear Fit AirElliptical AE 40 | 1.0 | | Мультистанция Nohrd SlimBeam | Сведение бедра UltraGym LF-510 | 0.0 | * Loss: [ContrastiveLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#contrastiveloss) with these parameters: ```json { "distance_metric": "SiameseDistanceMetric.COSINE_DISTANCE", "margin": 0.5, "size_average": true } ``` ### Evaluation Dataset #### Unnamed Dataset * Size: 28 evaluation samples * Columns: sentence1, sentence2, and score * Approximate statistics based on the first 28 samples: | | sentence1 | sentence2 | score | |:--------|:----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|:---------------------------------------------------------------| | type | string | string | float | | details | | | | * Samples: | sentence1 | sentence2 | score | |:----------------------------------------------------------------------------------|:-----------------------------------------------------------------|:-----------------| | Беговая дорожка Carbon Yukon | Кросстренер Octane Fitness Max Trainer MTX в Москве | 0.0 | | Беговая дорожка Беговая дорожка DFC BOSS I T-B1 для реабилитации | Беговая дорожка Protrain N6J | 1.0 | | Грузоблочный тренажер Precor C010ES - жим ногами/икроножные в Москве | Ягодичные мышцы Bronze Gym MNM-016A | 1.0 | * Loss: [ContrastiveLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#contrastiveloss) with these parameters: ```json { "distance_metric": "SiameseDistanceMetric.COSINE_DISTANCE", "margin": 0.5, "size_average": true } ``` ### Training Hyperparameters #### Non-Default Hyperparameters - `eval_strategy`: epoch - `per_device_train_batch_size`: 32 - `per_device_eval_batch_size`: 32 - `num_train_epochs`: 10 - `warmup_ratio`: 0.1 - `fp16`: True #### All Hyperparameters
Click to expand - `overwrite_output_dir`: False - `do_predict`: False - `eval_strategy`: epoch - `prediction_loss_only`: True - `per_device_train_batch_size`: 32 - `per_device_eval_batch_size`: 32 - `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`: 1.0 - `num_train_epochs`: 10 - `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 - `eval_use_gather_object`: False - `batch_sampler`: batch_sampler - `multi_dataset_batch_sampler`: proportional
### Training Logs | Epoch | Step | Training Loss | loss | cv_max_ap | |:------:|:----:|:-------------:|:------:|:---------:| | 0 | 0 | - | - | 0.6247 | | 1.0 | 428 | - | 0.0121 | 0.9407 | | 1.1682 | 500 | 0.0121 | - | - | | 2.0 | 856 | - | 0.0105 | 0.9805 | | 2.3364 | 1000 | 0.0037 | - | - | | 3.0 | 1284 | - | 0.0085 | 0.9821 | | 3.5047 | 1500 | 0.0028 | - | - | | 4.0 | 1712 | - | 0.0073 | 0.9891 | | 4.6729 | 2000 | 0.0025 | - | - | | 5.0 | 2140 | - | 0.0065 | 0.9924 | | 5.8411 | 2500 | 0.0021 | - | - | | 6.0 | 2568 | - | 0.0053 | 0.9963 | | 7.0 | 2996 | - | 0.0055 | 0.9963 | | 7.0093 | 3000 | 0.0018 | - | - | | 8.0 | 3424 | - | 0.0041 | 1.0 | | 8.1776 | 3500 | 0.0015 | - | - | | 9.0 | 3852 | - | 0.0040 | 1.0 | | 9.3458 | 4000 | 0.0014 | - | - | | 10.0 | 4280 | - | 0.0036 | 1.0 | ### Framework Versions - Python: 3.11.8 - Sentence Transformers: 3.1.0 - Transformers: 4.44.2 - PyTorch: 2.4.1+cu118 - Accelerate: 0.34.2 - Datasets: 3.0.0 - 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", } ``` #### ContrastiveLoss ```bibtex @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} } ```