gehaustein/PolyQual-2

SetFit with sentence-transformers/all-mpnet-base-v2

This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/all-mpnet-base-v2 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: sentence-transformers/all-mpnet-base-v2
• Classification head: a LogisticRegression instance
• Maximum Sequence Length: 384 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
0	'maybe coffeezilla was right that Polymarket is just a tool for insiders to make money....' 'Coping yes kids in chat ' 'The most common with ABOP'
1	"Just dispute the market even though rules always mislead sometimes others can't understand English." 'I sold some of this' "Yes you're missing something, if the GOP wins all swing states they win by between 65-104. They have to also win virginia or minnesota to get more than that"

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("gehaustein/polymarket-comments-binary")
# Run inference
preds = model("Lol prove it")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	1	16.6287	117

Label	Training Sample Count
0	307
1	307

Training Hyperparameters

• batch_size: (8, 8)
• num_epochs: (1, 1)
• max_steps: -1
• sampling_strategy: oversampling
• num_iterations: 20
• 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: True

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0003	1	0.4909	-
0.0163	50	0.3797	0.3548
0.0326	100	0.2726	0.2727
0.0489	150	0.2527	0.2652
0.0651	200	0.2342	0.2476
0.0814	250	0.1839	0.2088
0.0977	300	0.0915	0.2271
0.1140	350	0.0417	0.2716
0.1303	400	0.0197	0.3131

Framework Versions

• Python: 3.12.13
• SetFit: 1.1.3
• Sentence Transformers: 5.3.0
• Transformers: 4.49.0
• PyTorch: 2.10.0+cu128
• Datasets: 4.0.0
• Tokenizers: 0.21.4

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}
}