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: 5 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
product faq	'Does the Safe - Dual Doors Sneaker Crates offer protection for shoes?' 'What is the price of the Microfiber Towel (Pack of 3 & 5)?' 'How many Crease Defenders are included in the pack of 2?'
general faq	'What makes SHOEGR Safe - Dual Door Sneaker Crates different from other shoe storage solutions?' 'How should I store my shoes to keep them in good condition?' 'What can I do to keep my shoes smelling fresh?'
order tracking	"My order was supposed to arrive yesterday but it hasn't. Can you check the delivery status for me?" 'I ordered the Cupcake Cases 3 days ago with order no 34567 how long will it take to deliver?' 'What is the expected delivery time for the Baking Ingredients I ordered?'
product policy	'What happens if I breach any of the Terms of Service?' 'How long do I have to replace a product after delivery?' 'What are the different types of cookies used on your site and their purposes?'
product discoverability	'Do you carry products for cleaning mesh shoes?' 'What are the prices for products designed for all types of materials?' 'Are there any single-product options for shoe maintenance?'

Evaluation

Metrics

Label	Accuracy
all	0.8696

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("What is included in the Ultimate Shoe Cleaning Kit?")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	4	11.7841	24

Label	Training Sample Count
general faq	12
order tracking	24
product discoverability	15
product faq	21
product policy	16

Training Hyperparameters

batch_size: (16, 16)
num_epochs: (2, 2)
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: True

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0026	1	0.1187	-
0.1309	50	0.0867	-
0.2618	100	0.0003	-
0.3927	150	0.0002	-
0.5236	200	0.0001	-
0.6545	250	0.0002	-
0.7853	300	0.0001	-
0.9162	350	0.0001	-
1.0471	400	0.0001	-
1.1780	450	0.0001	-
1.3089	500	0.0001	-
1.4398	550	0.0001	-
1.5707	600	0.0001	-
1.7016	650	0.0001	-
1.8325	700	0.0001	-
1.9634	750	0.0001	-

Framework Versions

Python: 3.10.16
SetFit: 1.0.3
Sentence Transformers: 2.7.0
Transformers: 4.40.2
PyTorch: 2.2.2
Datasets: 2.19.1
Tokenizers: 0.19.1

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}
}

Shankhdhar
/

classifier_woog_shoegr