README.md

metadata

tags:
  - setfit
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
widget:
  - text: >-
      Framing shop providing custom picture framing and art preservation
      services. We frame artwork, photographs, and memorabilia with quality
      materials.
  - text: >-
      Lymph node pathology reveals metastatic adenocarcinoma with extracapsular
      extension. Immunostains are consistent with breast primary tumor origin.
  - text: >-
      Fluoroscopy-guided placement of central venous catheter for chemotherapy
      administration. Port placement performed successfully with proper tip
      positioning confirmed.
  - text: >-
      Immunotherapy harnesses the body's immune system to fight cancer.
      Checkpoint inhibitors block proteins like PD-1 and CTLA-4 that normally
      prevent immune cells from attacking cancer cells, while CAR-T therapy
      genetically modifies T cells to better recognize cancer.
  - text: >-
      Endometrial biopsy shows endometrioid adenocarcinoma, grade 2, with
      squamous differentiation. Deep myometrial invasion is present involving
      the outer half.
metrics:
  - accuracy
pipeline_tag: text-classification
library_name: setfit
inference: true
base_model: TimKond/S-PubMedBert-MedQuAD
model-index:
  - name: SetFit with TimKond/S-PubMedBert-MedQuAD
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: Unknown
          type: unknown
          split: test
        metrics:
          - type: accuracy
            value: 1
            name: Accuracy

SetFit with TimKond/S-PubMedBert-MedQuAD

This is a SetFit model that can be used for Text Classification. This SetFit model uses TimKond/S-PubMedBert-MedQuAD as the Sentence Transformer embedding model. A LogisticRegression instance is used for classification.

The model has been trained using an efficient few-shot learning technique that involves:

1. Fine-tuning a Sentence Transformer with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.

Model Details

Model Description

• Model Type: SetFit
• Sentence Transformer body: TimKond/S-PubMedBert-MedQuAD
• Classification head: a LogisticRegression instance
• Maximum Sequence Length: 512 tokens
• Number of Classes: 2 classes

Model Sources

• Repository: SetFit on GitHub
• Paper: Efficient Few-Shot Learning Without Prompts
• Blogpost: SetFit: Efficient Few-Shot Learning Without Prompts

Model Labels

Label	Examples
non_oncology	'Food truck festival coordination service organizing mobile vendor events, permit management, and location logistics.' 'Traditional Japanese tea ceremony instruction teaching the ancient art of chanoyu. Cultural immersion program exploring Japanese aesthetics, philosophy, and mindfulness practices.' 'Yoga studio offering various classes including Hatha, Vinyasa, and restorative yoga. We provide a peaceful environment for mindfulness and physical wellness.'
oncology	'Combination therapy uses multiple treatment modalities to improve outcomes. For example, neoadjuvant chemotherapy can shrink tumors before surgery, while adjuvant therapy eliminates remaining cancer cells after primary treatment.' 'Hereditary cancer syndromes account for 5-10% of all cancers. BRCA1 and BRCA2 mutations significantly increase breast and ovarian cancer risk, while Lynch syndrome increases colorectal and endometrial cancer risk.' 'Endoplasmic Reticulum: The cisternae of the endoplasmic reticulum are distended by fluid in hydropic swelling. In other forms of acute, reversible cell injury, membrane-bound polysomes may undergo disaggregation and detach from the surface of the rough endoplasmic reticulum.'

Evaluation

Metrics

Label	Accuracy
all	1.0

Uses

Direct Use for Inference

First install the SetFit library:

pip install setfit

Then you can load this model and run inference.

from setfit import SetFitModel

# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("setfit_model_id")
# Run inference
preds = model("Lymph node pathology reveals metastatic adenocarcinoma with extracapsular extension. Immunostains are consistent with breast primary tumor origin.")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	12	31.1543	929

Label	Training Sample Count
non_oncology	293
oncology	180

Training Hyperparameters

• batch_size: (16, 16)
• num_epochs: (1, 1)
• max_steps: 200
• sampling_strategy: oversampling
• body_learning_rate: (2e-05, 1e-05)
• head_learning_rate: 0.01
• loss: CosineSimilarityLoss
• distance_metric: cosine_distance
• margin: 0.25
• end_to_end: False
• use_amp: False
• warmup_proportion: 0.1
• l2_weight: 0.01
• seed: 42
• eval_max_steps: -1
• load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.005	1	0.1922	-
0.25	50	0.0766	-
0.5	100	0.0081	-
0.75	150	0.0058	-
1.0	200	0.0005	0.0074

Framework Versions

• Python: 3.11.10
• SetFit: 1.1.2
• Sentence Transformers: 5.0.0
• Transformers: 4.49.0
• PyTorch: 2.7.1
• Datasets: 3.6.0
• Tokenizers: 0.21.2

Citation

BibTeX

@article{https://doi.org/10.48550/arxiv.2209.11055,
    doi = {10.48550/ARXIV.2209.11055},
    url = {https://arxiv.org/abs/2209.11055},
    author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
    keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
    title = {Efficient Few-Shot Learning Without Prompts},
    publisher = {arXiv},
    year = {2022},
    copyright = {Creative Commons Attribution 4.0 International}
}