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: 6 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	'i don t feel so self assured i need to compete or to justify why i m so clearly not doing as well as someone else' 'i should do but i think it means that i should always be open to opportunities of inviting and involving others in ministries and that i should be creative in finding ways for others to participate in and feel welcomed into such ministries' 'i feel like im going to be way more successful a writer because of it'
4	'i feel so weird and scattered with all wonders about a million different things' 'i mean already as a parent from the moment the iolani left my body i can tell you i feel like im constantly fearful for something horrible happening to her thats out of my control' 'i think i was feeling vulnerable due to the stress of having to buy a new sewing machine and printer'
5	'i feel like this inside theres one thing i wanna know whats so funny bout peace love and understanding' 'i feel like itd be strange at the least and possibly offensive to tell a gay friend id like to experiment or something like that' 'i am not sure why in that moment that i thought i would be able to feel it hellip but it was pretty funny'
2	'i can feel that gentle rhythm imprinted on my skin i vibrates up my arm my stomach clenches my legs squeeze i forget his own leg has somehow ended up between mine' 'i feel specially fond of' 'i just feel like i dont like supporting walmart because maceys has such good family values and is closed on sundays and isnt trying to take over mom and pop stores but i have to be a smart consumer too'
3	'i am sure the vast majority of decent working class people feel insulted about being derided as unable to be respectful towards referees and are the parents who watch their child s match shouting abuse and swearing etc' 'im feeling irritated by her friggin name' 'i feel heartless now feeling bored and not believe in love anymore'
0	'i had just begun to feel like teaching was my metier but am now resigned to the fact that i likely wont teach at university ever again' 'i think the most common one that everyone has experienced is that doom and gloom feeling where you just feel like something tragic just happened' 'i feel a bit foolish now because in the last years they havent come back to my home town and i have had to travel to england to see them'

Evaluation

Metrics

Label	Accuracy
all	0.621

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("vidhi0206/setfit-paraphrase-mpnet-emotionv")
# Run inference
preds = model("i am feeling very indecisive and spontaneous")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	5	20.4375	47

Label	Training Sample Count
0	8
1	8
2	8
3	8
4	8
5	8

Training Hyperparameters

batch_size: (8, 8)
num_epochs: (1, 1)
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
seed: 42
eval_max_steps: -1
load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0042	1	0.2804	-
0.2083	50	0.0724	-
0.4167	100	0.0512	-
0.625	150	0.0108	-
0.8333	200	0.0027	-

Framework Versions

Python: 3.8.10
SetFit: 1.0.3
Sentence Transformers: 2.3.1
Transformers: 4.37.2
PyTorch: 2.2.0+cu121
Datasets: 2.17.0
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}
}

vidhi0206
/

setfit-paraphrase-mpnet-emotionv