--- tags: - sentence-transformers - sentence-similarity - feature-extraction - generated_from_trainer - dataset_size:1441905 - loss:CachedMultipleNegativesRankingLoss widget: - source_sentence: Treponema caused disease or disorder sentences: - bejel - tumor of ureter - debrisoquine, ultrarapid metabolism of - source_sentence: B cell (antibody) deficiencies sentences: - distal phalanx of digit IV - well-differentiated fetal adenocarcinoma of the lung - deficiency of humoral immunity - source_sentence: Elevated AdoHcy concentration sentences: - gepulste Abgabe - Elevated circulating S-adenosyl-L-homocysteine concentration - Frequently cries for no reason - source_sentence: Isoelectric focusing of serum transferrin consistent with CDG type II sentences: - Amblyomma aureolatum - squamous cell carcinoma of the bile duct - Abnormal isoelectric focusing of serum transferrin, type 2 pattern - source_sentence: Light-chain amyloidosis sentences: - partial deletion of the long arm of chromosome X - Teneria teneriensis - amyloidosis primary systemic pipeline_tag: sentence-similarity library_name: sentence-transformers metrics: - cosine_accuracy@1 - cosine_accuracy@3 - cosine_accuracy@5 - cosine_accuracy@10 - cosine_precision@1 - cosine_precision@3 - cosine_precision@5 - cosine_precision@10 - cosine_recall@1 - cosine_recall@3 - cosine_recall@5 - cosine_recall@10 - cosine_ndcg@10 - cosine_mrr@10 - cosine_map@100 model-index: - name: SentenceTransformer results: - task: type: information-retrieval name: Information Retrieval dataset: name: owl ontology eval type: owl_ontology_eval metrics: - type: cosine_accuracy@1 value: 0.6302799165287473 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.8147801683816651 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.8775275239260272 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.9268187378570915 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.6302799165287473 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.27634261591230724 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.17979420018709072 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.09566812981218968 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.6216929313281044 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.8081120625554675 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.8723585426111152 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.9241442997289582 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.7796907170635903 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.7342337217921898 name: Cosine Mrr@10 - type: cosine_map@100 value: 0.734065731352359 name: Cosine Map@100 --- # SentenceTransformer This is a [sentence-transformers](https://www.SBERT.net) model trained on the json dataset. 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 - **Maximum Sequence Length:** 1024 tokens - **Output Dimensionality:** 384 dimensions - **Similarity Function:** Cosine Similarity - **Training Dataset:** - json ### 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': 1024, '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: ```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("pankajrajdeo/bond-embed-v1-fp16") # Run inference sentences = [ 'Light-chain amyloidosis', 'amyloidosis primary systemic', 'partial deletion of the long arm of chromosome X', ] 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 #### Information Retrieval * Dataset: `owl_ontology_eval` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.6303 | | cosine_accuracy@3 | 0.8148 | | cosine_accuracy@5 | 0.8775 | | cosine_accuracy@10 | 0.9268 | | cosine_precision@1 | 0.6303 | | cosine_precision@3 | 0.2763 | | cosine_precision@5 | 0.1798 | | cosine_precision@10 | 0.0957 | | cosine_recall@1 | 0.6217 | | cosine_recall@3 | 0.8081 | | cosine_recall@5 | 0.8724 | | cosine_recall@10 | 0.9241 | | **cosine_ndcg@10** | **0.7797** | | cosine_mrr@10 | 0.7342 | | cosine_map@100 | 0.7341 | ## Training Details ### Training Dataset #### json * Dataset: json * Size: 1,441,905 training samples * Columns: anchor and positive * Approximate statistics based on the first 1000 samples: | | anchor | positive | |:--------|:---------------------------------------------------------------------------------|:---------------------------------------------------------------------------------| | type | string | string | | details | | | * Samples: | anchor | positive | |:-------------------------------------|:-------------------------------------| | Mangshan horned toad | Mangshan spadefoot toad | | Leuconotopicos borealis | Picoides borealis | | Cylindrella teneriensis | Teneria teneriensis | * Loss: [CachedMultipleNegativesRankingLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cachedmultiplenegativesrankingloss) with these parameters: ```json { "scale": 20.0, "similarity_fct": "cos_sim" } ``` ### Training Hyperparameters #### Non-Default Hyperparameters - `eval_strategy`: steps - `per_device_train_batch_size`: 1024 - `learning_rate`: 1.5e-05 - `num_train_epochs`: 5 - `lr_scheduler_type`: cosine - `warmup_ratio`: 0.05 - `bf16`: True - `dataloader_num_workers`: 32 - `load_best_model_at_end`: True - `gradient_checkpointing`: 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`: 1024 - `per_device_eval_batch_size`: 8 - `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`: 1.5e-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`: 5 - `max_steps`: -1 - `lr_scheduler_type`: cosine - `lr_scheduler_kwargs`: {} - `warmup_ratio`: 0.05 - `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`: 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`: 32 - `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`: None - `hub_always_push`: False - `hub_revision`: None - `gradient_checkpointing`: True - `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`: False - `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`: False - `prompts`: None - `batch_sampler`: batch_sampler - `multi_dataset_batch_sampler`: proportional
### Training Logs | Epoch | Step | Training Loss | owl_ontology_eval_cosine_ndcg@10 | |:------:|:----:|:-------------:|:--------------------------------:| | 0.0717 | 100 | 1.3232 | - | | 0.1434 | 200 | 1.021 | - | | 0.2151 | 300 | 0.9633 | - | | 0.2867 | 400 | 0.9068 | - | | 0.3297 | 460 | - | 0.7207 | | 0.3584 | 500 | 0.8723 | - | | 0.4301 | 600 | 0.852 | - | | 0.5018 | 700 | 0.8161 | - | | 0.5735 | 800 | 0.7939 | - | | 0.6452 | 900 | 0.7935 | - | | 0.6595 | 920 | - | 0.7364 | | 0.7168 | 1000 | 0.7646 | - | | 0.7885 | 1100 | 0.7464 | - | | 0.8602 | 1200 | 0.7376 | - | | 0.9319 | 1300 | 0.7313 | - | | 0.9892 | 1380 | - | 0.7468 | | 1.0036 | 1400 | 0.7099 | - | | 1.0753 | 1500 | 0.6884 | - | | 1.1470 | 1600 | 0.6776 | - | | 1.2186 | 1700 | 0.6694 | - | | 1.2903 | 1800 | 0.6641 | - | | 1.3190 | 1840 | - | 0.7561 | | 1.3620 | 1900 | 0.6526 | - | | 1.4337 | 2000 | 0.6524 | - | | 1.5054 | 2100 | 0.6364 | - | | 1.5771 | 2200 | 0.6339 | - | | 1.6487 | 2300 | 0.626 | 0.7614 | | 1.7204 | 2400 | 0.6197 | - | | 1.7921 | 2500 | 0.6193 | - | | 1.8638 | 2600 | 0.6155 | - | | 1.9355 | 2700 | 0.6142 | - | | 1.9785 | 2760 | - | 0.7662 | | 2.0072 | 2800 | 0.5853 | - | | 2.0789 | 2900 | 0.5824 | - | | 2.1505 | 3000 | 0.5769 | - | | 2.2222 | 3100 | 0.5765 | - | | 2.2939 | 3200 | 0.5608 | - | | 2.3082 | 3220 | - | 0.7698 | | 2.3656 | 3300 | 0.5695 | - | | 2.4373 | 3400 | 0.5641 | - | | 2.5090 | 3500 | 0.5638 | - | | 2.5806 | 3600 | 0.554 | - | | 2.6380 | 3680 | - | 0.7735 | | 2.6523 | 3700 | 0.5539 | - | | 2.7240 | 3800 | 0.5495 | - | | 2.7957 | 3900 | 0.5556 | - | | 2.8674 | 4000 | 0.5397 | - | | 2.9391 | 4100 | 0.5447 | - | | 2.9677 | 4140 | - | 0.7757 | | 3.0108 | 4200 | 0.5331 | - | | 3.0824 | 4300 | 0.5336 | - | | 3.1541 | 4400 | 0.5346 | - | | 3.2258 | 4500 | 0.5247 | - | | 3.2975 | 4600 | 0.5241 | 0.7775 | | 3.3692 | 4700 | 0.5257 | - | | 3.4409 | 4800 | 0.5241 | - | | 3.5125 | 4900 | 0.5171 | - | | 3.5842 | 5000 | 0.5215 | - | | 3.6272 | 5060 | - | 0.7787 | | 3.6559 | 5100 | 0.5203 | - | | 3.7276 | 5200 | 0.5214 | - | | 3.7993 | 5300 | 0.5266 | - | | 3.8710 | 5400 | 0.5127 | - | | 3.9427 | 5500 | 0.5062 | - | | 3.9570 | 5520 | - | 0.7790 | | 4.0143 | 5600 | 0.5104 | - | | 4.0860 | 5700 | 0.5155 | - | | 4.1577 | 5800 | 0.5042 | - | | 4.2294 | 5900 | 0.5174 | - | | 4.2867 | 5980 | - | 0.7797 | | 4.3011 | 6000 | 0.509 | - | | 4.3728 | 6100 | 0.5106 | - | | 4.4444 | 6200 | 0.5076 | - | | 4.5161 | 6300 | 0.5046 | - | | 4.5878 | 6400 | 0.5077 | - | | 4.6165 | 6440 | - | 0.7795 | | 4.6595 | 6500 | 0.5114 | - | | 4.7312 | 6600 | 0.5103 | - | | 4.8029 | 6700 | 0.5106 | - | | 4.8746 | 6800 | 0.5102 | - | | 4.9462 | 6900 | 0.5076 | 0.7797 | ### Framework Versions - Python: 3.11.11 - Sentence Transformers: 3.4.1 - Transformers: 4.53.2 - PyTorch: 2.6.0+cu124 - Accelerate: 1.5.2 - Datasets: 3.2.0 - Tokenizers: 0.21.0 ## 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", } ``` #### CachedMultipleNegativesRankingLoss ```bibtex @misc{gao2021scaling, title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup}, author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan}, year={2021}, eprint={2101.06983}, archivePrefix={arXiv}, primaryClass={cs.LG} } ```