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

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

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-mpnet-base-v2
Classification head: a LogisticRegression instance
Maximum Sequence Length: 512 tokens
Number of Classes: 11 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
6	'Why did my payment fail?' 'Why is my payment not processing?' 'What are common payment issues?'
0	'Your products are not good' 'Your service is bad'
8	'Good Bye' 'Take care' 'Bye'
9	'Thanks'
10	'23234dfdff' 'abc' 'djfnknf'
4	'My order ref is ' 'Order Ref is C-123-P' 'Order id is C-123-P'
5	'How can I track my order?' 'When will my order arrive?' "I didn't receive my COD points"
7	'What is your refund policy?' 'How can I get a refund?'
3	'order email' 'resend order email' 'I didnt receive my order email'
2	'How are you'
1	'Hi' 'Wassup' 'Hello'

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("Bye")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	1	3.4737	6

Label	Training Sample Count
0	2
1	3
2	1
3	3
4	5
5	7
6	3
7	2
8	3
9	1
10	8

Training Hyperparameters

batch_size: (16, 16)
num_epochs: (5, 5)
max_steps: -1
sampling_strategy: oversampling
num_iterations: 20
body_learning_rate: (2e-05, 2e-05)
head_learning_rate: 2e-05
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
0.0105	1	0.3591	-
0.5263	50	0.1235	-
1.0526	100	0.0233	-
1.5789	150	0.0057	-
2.1053	200	0.0032	-
2.6316	250	0.0025	-
3.1579	300	0.002	-
3.6842	350	0.0017	-
4.2105	400	0.0019	-
4.7368	450	0.0016	-

Framework Versions

Python: 3.10.13
SetFit: 1.1.1
Sentence Transformers: 3.4.0
Transformers: 4.48.1
PyTorch: 2.5.1
Datasets: 3.2.0
Tokenizers: 0.21.0

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}
}

sam123116
/

trained_model