metadata

library_name: setfit
metrics:
  - f1
  - accuracy
pipeline_tag: text-classification
tags:
  - setfit
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
widget:
  - text: >-
      To make introductions between Camelot's Chairman and the Cabinet
      Secretary. We discussed the operation of the UK National Lottery and how
      to maximise returns to National Lottery Good Causes as well as our plans
      to celebrate the 25th birthday of The National Lottery.
  - text: Discussion on crime
  - text: To discuss Northern Powerhouse Rail and HS2
  - text: To discuss food security
  - text: Electricity market
inference: true
model-index:
  - name: SetFit
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: Unknown
          type: unknown
          split: test
        metrics:
          - type: f1
            value: 0.92
            name: F1
          - type: accuracy
            value: 0.9658119658119658
            name: Accuracy

SetFit

This is a SetFit model that can be used for Text Classification. A SetFitHead instance is used for classification.

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

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

Model Details

Model Description

Model Type: SetFit
Classification head: a SetFitHead instance
Maximum Sequence Length: 512 tokens
Number of Classes: 4 classes

Model Sources

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

Evaluation

Metrics

Label	F1	Accuracy
all	0.92	0.9658

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("twright8/setfit-oversample-labels-lobbying")
# Run inference
preds = model("Electricity market")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	2	26.1406	153

Training Hyperparameters

batch_size: (16, 2)
num_epochs: (4, 9)
max_steps: -1
sampling_strategy: oversampling
body_learning_rate: (1.0797496673911536e-05, 3.457046714445997e-05)
head_learning_rate: 0.0004470582121407239
loss: CoSENTLoss
distance_metric: cosine_distance
margin: 0.25
end_to_end: True
use_amp: False
warmup_proportion: 0.1
seed: 42
eval_max_steps: -1
load_best_model_at_end: True

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0040	1	19.1843	-
0.2024	50	11.3434	-
0.4049	100	9.3116	-
0.6073	150	2.7233	-
0.8097	200	1.5662	-
1.0	247	-	14.3603
1.0121	250	0.0159	-
1.2146	300	0.0135	-
1.4170	350	0.0003	-
1.6194	400	0.0002	-
1.8219	450	0.0007	-
2.0	494	-	16.8205
2.0243	500	0.0023	-
2.2267	550	0.0004	-
2.4291	600	0.0001	-
2.6316	650	0.0	-
2.8340	700	0.0003	-
3.0	741	-	15.2312
3.0364	750	0.0	-
3.2389	800	3.1257	-
3.4413	850	0.0001	-
3.6437	900	0.0002	-
3.8462	950	0.0139	-
4.0	988	-	14.4995

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.10.12
SetFit: 1.0.3
Sentence Transformers: 3.0.1
Transformers: 4.39.0
PyTorch: 2.3.1+cu118
Datasets: 2.20.0
Tokenizers: 0.15.2

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}
}