skylord/setfit-bge-small-v1.5-sst2-8-shot-talk2loop

SetFit with BAAI/bge-small-en-v1.5

This is a SetFit model that can be used for Text Classification. This SetFit model uses BAAI/bge-small-en-v1.5 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-small-en-v1.5
• 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
sensitive	'Im Amie Taylaran from Pan-ay Clarin from Solo Parent Organization grateful and excited to receive the help you are giving.' 'I want to volunteer' 'There is now a growing popular street Pennsylvania street in the annex Phase 3 of Greenland Executive Village for bikers walkers joggers every morning when the weather is fair. I presume they are groups of retirees matrons sports enthusiasts an even dance exercisers. They all wear face masks for health protection against COVID-19 infection. My concern is this: face masks are just thrown away after use when these fitness buffs are done with their morning binges. Face masks thrown on the pavement of the street the sidewalks and the grass field. Health fitness aficionados they all are but careless with the proper disposal of their face masks.'
other	'There is a man here forced us the girls in the house to have sex with him. He took videos of us and now he is asking for money. Can someone help us?' 'In this community alcohol abuse is rampant. The men go out drinking and come home and beat their wives. They are getting seriously injured.' "I find myself in a very challenging situation - I've experienced sexual abuse at work. If anyone has gone through something similar, I would appreciate your guidance and support. It's tough, but we're stronger together."

Evaluation

Metrics

Label	Accuracy
all	0.9828

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("skylord/setfit-bge-small-v1.5-sst2-8-shot-talk2loop")
# Run inference
preds = model("Drenage problem here in lilanda")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	4	38.0	171

Label	Training Sample Count
sensitive	8
other	8

Training Hyperparameters

• batch_size: (32, 32)
• num_epochs: (10, 10)
• max_steps: -1
• 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
• seed: 42
• eval_max_steps: -1
• load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.2	1	0.1988	-
10.0	50	0.019	-

Framework Versions

• Python: 3.10.11
• SetFit: 1.0.3
• Sentence Transformers: 2.3.1
• Transformers: 4.37.2
• PyTorch: 2.2.0+cu121
• Datasets: 2.16.1
• Tokenizers: 0.15.1

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}
}