context_fine-tuned-SapBERT1 / README.md

Stevenf232

Add new SentenceTransformer model

e3b01cf verified 2 days ago

preview code

raw

history blame contribute delete

15.4 kB

metadata

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dense
  - generated_from_trainer
  - dataset_size:27120
  - loss:ContrastiveLoss
base_model: cambridgeltl/SapBERT-from-PubMedBERT-fulltext
widget:
  - source_sentence: >-
      1-bromo-1-chloro-2,2,2-trifluoroethane [SEP] ious degrees. Both compounds
      are metabolised in the same way as 1-bromo-1-chloro-2,2,2-trifluoroethane
      (halothane) to form reactive trifluoroacetyl halide intermediat
    sentences:
      - >-
        Nephrotic Syndrome [SEP] A condition characterized by severe
        PROTEINURIA, greater than 3.5 g/day in an average adult. The substantial
        loss of protein in 
      - >-
        Personality Disorders [SEP] A major deviation from normal patterns of
        behavior.
            
      - >-
        Propylene Glycol [SEP] A clear, colorless, viscous organic solvent and
        diluent used in pharmaceutical preparations.
            
  - source_sentence: >-
      bupivacaine [SEP] was to investigate the influence of calcium channel
      blockers on bupivacaine-induced acute toxicity. For each of the three
      tested calcium ch
    sentences:
      - |-
        Bupivacaine [SEP] A widely used local anesthetic agent.
            
      - |-
        Urinary Bladder Neoplasms [SEP] Tumors or cancer of the URINARY BLADDER.
            
      - >-
        Spondylarthropathies [SEP] Heterogeneous group of arthritic diseases
        sharing clinical and radiologic features. They are associated with the
        HLA-B27 ANTIGEN
  - source_sentence: >-
      proteinuria [SEP] and an increase in fractional Li excretion. Lithium also
      caused proteinuria and systolic hypertension in absence of
      glomerulosclerosis. HP 
    sentences:
      - |-
        Levofloxacin [SEP] The L-isomer of Ofloxacin.
            
      - >-
        Gastroesophageal Reflux [SEP] Retrograde flow of gastric juice (GASTRIC
        ACID) and/or duodenal contents (BILE ACIDS; PANCREATIC JUICE) into the
        distal ESOPHAGU
      - >-
        Carcinoma, Hepatocellular [SEP] A primary malignant neoplasm of
        epithelial liver cells. It ranges from a well-differentiated tumor with
        EPITHELIAL CELLS indisti
  - source_sentence: >-
      radiculopathy [SEP] OBJECTIVE: Conventional treatment methods of
      lumbusacral radiculopathy are physical therapy, epidural steroid
      injections, oral medicat
    sentences:
      - >-
        Seizures [SEP] Clinical or subclinical disturbances of cortical function
        due to a sudden, abnormal, excessive, and disorganized discharge of br
      - >-
        Desipramine [SEP] A tricyclic dibenzazepine compound that potentiates
        neurotransmission. Desipramine selectively blocks reuptake of
        norepinephrine
      - >-
        Amphetamine [SEP] A powerful central nervous system stimulant and
        sympathomimetic. Amphetamine has multiple mechanisms of action including
        blockin
  - source_sentence: Death [SEP] Death from chemotherapy in gestational trophoblastic disease.
    sentences:
      - >-
        Coma [SEP] A profound state of unconsciousness associated with depressed
        cerebral activity from which the individual cannot be aroused. Com
      - >-
        Neurotoxicity Syndromes [SEP] Neurologic disorders caused by exposure to
        toxic substances through ingestion, injection, cutaneous application, or
        other method
      - >-
        Vascular Diseases [SEP] Pathological processes involving any of the
        BLOOD VESSELS in the cardiac or peripheral circulation. They include
        diseases of ART
pipeline_tag: sentence-similarity
library_name: sentence-transformers

SentenceTransformer based on cambridgeltl/SapBERT-from-PubMedBERT-fulltext

This is a sentence-transformers model finetuned from cambridgeltl/SapBERT-from-PubMedBERT-fulltext. 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: cambridgeltl/SapBERT-from-PubMedBERT-fulltext
Maximum Sequence Length: 512 tokens
Output Dimensionality: 768 dimensions
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': 512, 'do_lower_case': False, 'architecture': 'BertModel'})
  (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("Stevenf232/context_fine-tuned-SapBERT1")
# Run inference
sentences = [
    'Death [SEP] Death from chemotherapy in gestational trophoblastic disease.',
    'Neurotoxicity Syndromes [SEP] Neurologic disorders caused by exposure to toxic substances through ingestion, injection, cutaneous application, or other method',
    'Coma [SEP] A profound state of unconsciousness associated with depressed cerebral activity from which the individual cannot be aroused. Com',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.5542, 0.6546],
#         [0.5542, 1.0000, 0.4659],
#         [0.6546, 0.4659, 1.0000]])

Training Details

Training Dataset

Unnamed Dataset

Size: 27,120 training samples
Columns: sentence_0, sentence_1, and label
Approximate statistics based on the first 1000 samples:
sentence_0 sentence_1 label
type string string float
details
min: 9 tokens
mean: 29.3 tokens
max: 82 tokens

min: 4 tokens
mean: 24.34 tokens
max: 40 tokens

min: 0.0
mean: 0.19
max: 1.0

	sentence_0	sentence_1	label
type	string	string	float
details	min: 9 tokens mean: 29.3 tokens max: 82 tokens	min: 4 tokens mean: 24.34 tokens max: 40 tokens	min: 0.0 mean: 0.19 max: 1.0

Samples:

sentence_0	sentence_1	label
`prolactinomas [SEP] l prolactin greater than 20 ng./ml. in 1.86% of 1,821 patients, prolactinomas in 7, 0.38%). Bromocriptine was definitely effective in cases w`	`Nicotine [SEP] Nicotine is highly toxic alkaloid. It is the prototypical agonist at nicotinic cholinergic receptors where it dramatically stimu`	`0.0`
`acetazolamide [SEP] reatment for periodic paralysis and myotonia. Three patients on acetazolamide (15%) developed renal calculi. Extracorporeal lithotripsy succe`	`Neutropenia [SEP] A decrease in the number of NEUTROPHILS found in the blood.`	`0.0`
`methylergonovine [SEP] Effect of direct intracoronary administration of methylergonovine in patients with and without variant angina.`	`Methylergonovine [SEP] A homolog of ERGONOVINE containing one more CH2 group. (Merck Index, 11th ed)`	`1.0`

Loss: ContrastiveLoss with these parameters:

{
    "distance_metric": "SiameseDistanceMetric.COSINE_DISTANCE",
    "margin": 0.5,
    "size_average": true
}

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 16
per_device_eval_batch_size: 16
fp16: True
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 16
per_device_eval_batch_size: 16
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
num_train_epochs: 3
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: None
warmup_ratio: None
warmup_steps: 0
log_level: passive
log_level_replica: warning
log_on_each_node: True
logging_nan_inf_filter: True
enable_jit_checkpoint: False
save_on_each_node: False
save_only_model: False
restore_callback_states_from_checkpoint: False
use_cpu: False
seed: 42
data_seed: None
bf16: False
fp16: True
bf16_full_eval: False
fp16_full_eval: False
tf32: None
local_rank: -1
ddp_backend: None
debug: []
dataloader_drop_last: False
dataloader_num_workers: 0
dataloader_prefetch_factor: None
disable_tqdm: False
remove_unused_columns: True
label_names: None
load_best_model_at_end: False
ignore_data_skip: False
fsdp: []
fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
parallelism_config: None
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch_fused
optim_args: None
group_by_length: False
length_column_name: length
project: huggingface
trackio_space_id: trackio
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
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: False
gradient_checkpointing_kwargs: None
include_for_metrics: []
eval_do_concat_batches: True
auto_find_batch_size: False
full_determinism: False
ddp_timeout: 1800
torch_compile: False
torch_compile_backend: None
torch_compile_mode: None
include_num_input_tokens_seen: no
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: True
use_cache: False
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss
0.2950	500	0.0105
0.5900	1000	0.0066
0.8850	1500	0.0054
1.1799	2000	0.0043
1.4749	2500	0.0036
1.7699	3000	0.0034
2.0649	3500	0.0032
2.3599	4000	0.0024
2.6549	4500	0.0025
2.9499	5000	0.0024

Framework Versions

Python: 3.12.12
Sentence Transformers: 5.2.3
Transformers: 5.0.0
PyTorch: 2.10.0+cu128
Accelerate: 1.12.0
Datasets: 4.0.0
Tokenizers: 0.22.2

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}
}