SentenceTransformer based on sentence-transformers/multi-qa-mpnet-base-dot-v1

This is a sentence-transformers model finetuned from sentence-transformers/multi-qa-mpnet-base-dot-v1. It maps sentences & paragraphs to a 768-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/multi-qa-mpnet-base-dot-v1
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 tokens
• Similarity Function: Dot Product

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: MPNetModel 
  (1): Pooling({'word_embedding_dimension': 768, '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})
)

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("sentence_transformers_model_id")
# Run inference
sentences = [
    'Thyroid-stimulating hormone receptor gene, chromosome 14q31, homozygous mutation',
    'A number sign (#) is used with this entry because of evidence that congenital nongoitrous hypothyroidism-1 (CHNG1) is caused by homozygous or compound heterozygous mutation in the gene encoding the thyroid-stimulating hormone receptor (TSHR; 603372) on chromosome 14q31.\n\nDescription\n\nResistance to thyroid-stimulating hormone (TSH; see 188540), a hallmark of congenital nongoitrous hypothyroidism, causes increased levels of plasma TSH and low levels of thyroid hormone. Only a subset of patients develop frank hypothyroidism; the remainder are euthyroid and asymptomatic (so-called compensated hypothyroidism) and are usually detected by neonatal screening programs (Paschke and Ludgate, 1997).\n\n### Genetic Heterogeneity of Congenital Nongoitrous Hypothyroidism',
    'Visuospatial dysgnosia is a loss of the sense of "whereness" in the relation of oneself to one\'s environment and in the relation of objects to each other.[1] Visuospatial dysgnosia is often linked with topographical disorientation.\n\n## Contents\n\n  * 1 Symptoms\n  * 2 Lesion areas\n  * 3 Case studies\n  * 4 Therapies\n  * 5 References\n\n## Symptoms[edit]\n\nThe syndrome rarely presents itself the same way in every patient. Some symptoms that occur may be:',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

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

Evaluation

Metrics

Information Retrieval

• Evaluated with InformationRetrievalEvaluator

Training Details

Training Dataset

Unnamed Dataset

• Size: 89,218 training samples
• Columns: queries and chunks
• Approximate statistics based on the first 1000 samples:

  	queries	chunks
  type	string	string
  details	min: 7 tokensmean: 18.07 tokensmax: 63 tokens	min: 5 tokensmean: 161.59 tokensmax: 299 tokens

• Samples:

  queries	chunks
  Polyhydramnios, megalencephaly, symptomatic epilepsy	A number sign (#) is used with this entry because of evidence that polyhydramnios, megalencephaly, and symptomatic epilepsy (PMSE) is caused by homozygous mutation in the STRADA gene (608626) on chromosome 17q23.Clinical Features
  Polyhydramnios, megalencephaly, STRADA gene mutation	A number sign (#) is used with this entry because of evidence that polyhydramnios, megalencephaly, and symptomatic epilepsy (PMSE) is caused by homozygous mutation in the STRADA gene (608626) on chromosome 17q23.Clinical Features
  Megalencephaly, symptomatic epilepsy, chromosome 17q23	A number sign (#) is used with this entry because of evidence that polyhydramnios, megalencephaly, and symptomatic epilepsy (PMSE) is caused by homozygous mutation in the STRADA gene (608626) on chromosome 17q23.Clinical Features

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 1,
      "similarity_fct": "dot_score"
  }
  

Evaluation Dataset

Unnamed Dataset

• Size: 18,180 evaluation samples
• Columns: queries and chunks
• Approximate statistics based on the first 1000 samples:

  	queries	chunks
  type	string	string
  details	min: 6 tokensmean: 18.35 tokensmax: 82 tokens	min: 4 tokensmean: 152.55 tokensmax: 312 tokens

• Samples:

  queries	chunks
  Weight loss, anorexia, fatigue, epigastric pain and discomfort	Undifferentiated carcinoma of stomach is a rare epithelial tumour of the stomach that lacks any features of differentiation beyond an epithelial phenotype. The presenting symptoms are usually vague and nonspecific, such as weight loss, anorexia, fatigue, epigastric pain and discomfort, heartburn and nausea, vomiting or hematemesis. Patients may also be asymptomatic. Ascites, jaundice, intestinal obstruction and peripheral lymphadenopathy indicate advanced stages and metastatic spread.
  Heartburn, nausea, vomiting, hematemesis	Undifferentiated carcinoma of stomach is a rare epithelial tumour of the stomach that lacks any features of differentiation beyond an epithelial phenotype. The presenting symptoms are usually vague and nonspecific, such as weight loss, anorexia, fatigue, epigastric pain and discomfort, heartburn and nausea, vomiting or hematemesis. Patients may also be asymptomatic. Ascites, jaundice, intestinal obstruction and peripheral lymphadenopathy indicate advanced stages and metastatic spread.
  Ascites, jaundice, intestinal obstruction, peripheral lymphadenopathy	Undifferentiated carcinoma of stomach is a rare epithelial tumour of the stomach that lacks any features of differentiation beyond an epithelial phenotype. The presenting symptoms are usually vague and nonspecific, such as weight loss, anorexia, fatigue, epigastric pain and discomfort, heartburn and nausea, vomiting or hematemesis. Patients may also be asymptomatic. Ascites, jaundice, intestinal obstruction and peripheral lymphadenopathy indicate advanced stages and metastatic spread.

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 1,
      "similarity_fct": "dot_score"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• learning_rate: 2e-05
• num_train_epochs: 50
• warmup_ratio: 0.1
• fp16: True
• load_best_model_at_end: True
• eval_on_start: True
• batch_sampler: no_duplicates

All Hyperparameters

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• 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: 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: 50
• 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: 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: 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
• eval_on_start: True
• eval_use_gather_object: False
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.11.9
• Sentence Transformers: 3.0.1
• Transformers: 4.43.3
• PyTorch: 2.3.1+cu121
• Accelerate: 0.30.1
• Datasets: 2.19.2
• 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",
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply}, 
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}