Ja-ck/setfit-medical-binary-classifier

SetFit with BAAI/bge-m3

This is a SetFit model that can be used for Text Classification. This SetFit model uses BAAI/bge-m3 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: BAAI/bge-m3
• Classification head: a LogisticRegression instance
• Maximum Sequence Length: 8192 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
1	'공복 혈당 상승은 검사 이상으로 임상적으로 의미 있음(당뇨 가능성 확인 필요).' 'CPK 상승으로 근손상 가능성 및 근육 질환 의심이 제시됨.' '자궁경부세포검사에서 위축성 세포 변화가 확인되어 검사 이상 소견이 있음'
0	'소견에 구체적인 이상 소견이나 검사 결과가 명시되어 있지 않고, 결과를 별지참조라고 되어 있어 임상적 이상 여부를 판단할 수 없음.' '전립선 석회화는 과거 염증의 흔적일 수 있으며 현재 이상 소견이 없어 임상적 의미 없음' '폐경 후 상태에서 정상 소견이므로 임상적으로 이상 없음'

Evaluation

Metrics

Label	Accuracy
all	0.9976

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("Ja-ck/setfit-medical-binary-classifier")
# Run inference
preds = model("공복 혈당 상승으로 검사 이상이 확인되었습니다.")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	2	10.7004	50

Label	Training Sample Count
0	1404
1	3613

Training Hyperparameters

• batch_size: (32, 32)
• num_epochs: (1, 16)
• 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	0.1863	-
0.25	50	0.0461	-
0.5	100	0.0011	-
0.75	150	0.0008	-
1.0	200	0.001	-

Framework Versions

• Python: 3.12.3
• SetFit: 1.1.3
• Sentence Transformers: 5.1.2
• Transformers: 4.57.1
• PyTorch: 2.8.0+cu128
• Datasets: 4.4.1
• Tokenizers: 0.22.1

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}
}