potsu-potsu/medembed-small-biomedical-matryoshka-v2

MedEmbed Biomedical MRL

This is a sentence-transformers 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: 512 tokens
• Output Dimensionality: 384 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • json
• Language: en
• License: apache-2.0

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: BertModel 
  (1): Pooling({'word_embedding_dimension': 384, '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:

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("potsu-potsu/medembed-small-biomedical-matryoshka-v2")
# Run inference
sentences = [
    'What are the effects of the deletion of all three Pcdh clusters (tricluster deletion) in mice?',
    'Multicluster Pcdh diversity is required for mouse olfactory neural circuit assembly. The vertebrate clustered protocadherin (Pcdh) cell surface proteins are encoded by three closely linked gene clusters (Pcdhα, Pcdhβ, and Pcdhγ). Although deletion of individual Pcdh clusters had subtle phenotypic consequences, the loss of all three clusters (tricluster deletion) led to a severe axonal arborization defect and loss of self-avoidance.',
    'Investigators proposed that there have been three extended periods in the evolution of gene regulatory elements. Early vertebrate evolution was characterized by regulatory gains near transcription factors and developmental genes, but this trend was replaced by innovations near extracellular signaling genes, and then innovations near posttranslational protein modifiers.',
]
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: dim_384
• Evaluated with InformationRetrievalEvaluator with these parameters:

  {
      "truncate_dim": 384
  }
  

Metric	Value
cosine_accuracy@10	0.9165
cosine_accuracy@15	0.9264
cosine_accuracy@20	0.9335
cosine_accuracy@25	0.9378
cosine_accuracy@30	0.9406
cosine_precision@10	0.3908
cosine_precision@15	0.3223
cosine_precision@20	0.2713
cosine_precision@25	0.2338
cosine_precision@30	0.205
cosine_recall@10	0.5942
cosine_recall@15	0.6588
cosine_recall@20	0.6988
cosine_recall@25	0.7271
cosine_recall@30	0.7471
cosine_ndcg@10	0.6798
cosine_ndcg@20	0.6889
cosine_ndcg@30	0.7001
cosine_mrr@10	0.7992
cosine_mrr@20	0.8004
cosine_mrr@30	0.8007
cosine_map@100	0.6025

Information Retrieval

• Dataset: dim_256
• Evaluated with InformationRetrievalEvaluator with these parameters:

  {
      "truncate_dim": 256
  }
  

Metric	Value
cosine_accuracy@10	0.9137
cosine_accuracy@15	0.9208
cosine_accuracy@20	0.9307
cosine_accuracy@25	0.9335
cosine_accuracy@30	0.942
cosine_precision@10	0.3857
cosine_precision@15	0.3173
cosine_precision@20	0.2668
cosine_precision@25	0.2298
cosine_precision@30	0.2015
cosine_recall@10	0.5816
cosine_recall@15	0.6467
cosine_recall@20	0.687
cosine_recall@25	0.7142
cosine_recall@30	0.7385
cosine_ndcg@10	0.6685
cosine_ndcg@20	0.6771
cosine_ndcg@30	0.689
cosine_mrr@10	0.791
cosine_mrr@20	0.7922
cosine_mrr@30	0.7926
cosine_map@100	0.5894

Information Retrieval

• Dataset: dim_128
• Evaluated with InformationRetrievalEvaluator with these parameters:

  {
      "truncate_dim": 128
  }
  

Metric	Value
cosine_accuracy@10	0.8925
cosine_accuracy@15	0.9109
cosine_accuracy@20	0.9137
cosine_accuracy@25	0.918
cosine_accuracy@30	0.9264
cosine_precision@10	0.3652
cosine_precision@15	0.3006
cosine_precision@20	0.2521
cosine_precision@25	0.2165
cosine_precision@30	0.1898
cosine_recall@10	0.5513
cosine_recall@15	0.6155
cosine_recall@20	0.6486
cosine_recall@25	0.6719
cosine_recall@30	0.6909
cosine_ndcg@10	0.6398
cosine_ndcg@20	0.6456
cosine_ndcg@30	0.6548
cosine_mrr@10	0.7783
cosine_mrr@20	0.78
cosine_mrr@30	0.7805
cosine_map@100	0.5534

Information Retrieval

• Dataset: dim_64
• Evaluated with InformationRetrievalEvaluator with these parameters:

  {
      "truncate_dim": 64
  }
  

Metric	Value
cosine_accuracy@10	0.8458
cosine_accuracy@15	0.8642
cosine_accuracy@20	0.8755
cosine_accuracy@25	0.8769
cosine_accuracy@30	0.8854
cosine_precision@10	0.3267
cosine_precision@15	0.2664
cosine_precision@20	0.2245
cosine_precision@25	0.1922
cosine_precision@30	0.1686
cosine_recall@10	0.4846
cosine_recall@15	0.5381
cosine_recall@20	0.5767
cosine_recall@25	0.5974
cosine_recall@30	0.6174
cosine_ndcg@10	0.5696
cosine_ndcg@20	0.5722
cosine_ndcg@30	0.5802
cosine_mrr@10	0.7236
cosine_mrr@20	0.7257
cosine_mrr@30	0.726
cosine_map@100	0.4742

Training Details

Training Dataset

json

• Dataset: json
• Size: 4,012 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 5 tokens mean: 16.13 tokens max: 49 tokens	min: 3 tokens mean: 63.38 tokens max: 485 tokens

• Samples:

  anchor	positive
  What is the implication of histone lysine methylation in medulloblastoma?	Aberrant patterns of H3K4, H3K9, and H3K27 histone lysine methylation were shown to result in histone code alterations, which induce changes in gene expression, and affect the proliferation rate of cells in medulloblastoma.
  What is the role of STAG1/STAG2 proteins in differentiation?	STAG1/STAG2 proteins are tumour suppressor proteins that suppress cell proliferation and are essential for differentiation.
  What is the association between cell phone use and glioblastoma?	The association between cell phone use and incident glioblastoma remains unclear. Some studies have reported that cell phone use was associated with incident glioblastoma, and with reduced survival of patients diagnosed with glioblastoma. However, other studies have repeatedly replicated to find an association between cell phone use and glioblastoma.

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          384,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: epoch
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 16
• gradient_accumulation_steps: 16
• learning_rate: 2e-05
• num_train_epochs: 8
• lr_scheduler_type: cosine
• warmup_ratio: 0.1
• bf16: True
• tf32: True
• load_best_model_at_end: True
• optim: adamw_torch_fused
• batch_sampler: no_duplicates

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: 16
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 16
• 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: 8
• 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: 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: True
• 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_fused
• 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
• gradient_checkpointing: False
• 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
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	dim_384_cosine_ndcg@30	dim_256_cosine_ndcg@30	dim_128_cosine_ndcg@30	dim_64_cosine_ndcg@30
1.0	8	-	0.7100	0.6989	0.6534	0.5490
1.2540	10	32.9088	-	-	-	-
2.0	16	-	0.7047	0.6961	0.6579	0.5768
2.5079	20	15.2449	-	-	-	-
3.0	24	-	0.7001	0.6899	0.6553	0.5834
3.7619	30	11.0949	-	-	-	-
4.0	32	-	0.7013	0.6882	0.6539	0.5811
5.0	40	9.344	0.7023	0.6910	0.6542	0.5809
6.0	48	-	0.7010	0.6891	0.6552	0.5805
6.2540	50	8.2338	-	-	-	-
7.0	56	-	0.7003	0.6892	0.6551	0.5812
7.5079	60	7.8342	-	-	-	-
8.0	64	-	0.7001	0.6890	0.6548	0.5802

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.12.6
• Sentence Transformers: 4.1.0
• Transformers: 4.52.4
• PyTorch: 2.6.0+cu124
• Accelerate: 1.7.0
• Datasets: 3.6.0
• Tokenizers: 0.21.1

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",
}

MatryoshkaLoss

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

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}
}