SetFit with sentence-transformers/paraphrase-MiniLM-L6-v2

This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/paraphrase-MiniLM-L6-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:

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
Sentence Transformer body: sentence-transformers/paraphrase-MiniLM-L6-v2
Classification head: a LogisticRegression instance
Maximum Sequence Length: 128 tokens
Number of Classes: 3 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	'Draft a polite email to reschedule a meeting' "Translate the phrase 'The quick brown fox jumps over the lazy dog' into Mandarin, French, and German." 'Write a short story (under 200 words) about a robot discovering nature.'
0	'Analyze the ethical implications of CRISPR gene editing' 'Debug this Python multiprocessing code deadlock: [code snippet]' 'Implement a Python async websocket client with error handling'
2	"Python syntax to print 'Hello World'" 'What is 2+2?' 'Who is the current president of the United States?'

Evaluation

Metrics

Label	F1
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("Explain the difference between HTTP and HTTPS")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	3	8.9583	17

Label	Training Sample Count
0	8
1	8
2	8

Training Hyperparameters

batch_size: (8, 8)
num_epochs: (3, 3)
max_steps: -1
sampling_strategy: oversampling
num_iterations: 20
body_learning_rate: (1e-05, 1e-05)
head_learning_rate: 0.001
loss: CosineSimilarityLoss
distance_metric: cosine_distance
margin: 0.25
end_to_end: True
use_amp: False
warmup_proportion: 0.1
l2_weight: 0.01
max_length: 384
seed: 42
evaluation_strategy: epoch
eval_max_steps: -1
load_best_model_at_end: True

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0083	1	0.4046	-
0.4167	50	0.2913	-
0.8333	100	0.1724	-
1.0	120	-	0.1897
1.25	150	0.0825	-
1.6667	200	0.0284	-
2.0	240	-	0.1806
2.0833	250	0.0137	-
2.5	300	0.0089	-
2.9167	350	0.007	-
3.0	360	-	0.1806

Framework Versions

Python: 3.11.13
SetFit: 1.1.2
Sentence Transformers: 3.3.1
Transformers: 4.53.3
PyTorch: 2.7.1
Datasets: 3.0.0
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}
}

MykhailoS
/

llm_router