vidhi0206/setfit-paraphrase-mpnet-emotion

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:

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: 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 feel so much better about that number' 'i feel like i have reached a plateau where im not buying as much as i use to and feeling more satisfied with my wardrobe and personal style' 'i feel especially thankful'
3	'i feel so violent just want to break some glass' 'i always feel rushed on the way to visit no comments' 'i think maybe about how strongly she feels about him and being there for him but brad looks really distracted'
5	'i feel like when i was a kid it was constantly impressed upon me how awesome ants are' 'i feel like it s a boy i would be pretty shocked if it was so somewhere in there my gut or my brain is saying girl' 'i feel like every day i walk around with so much stress and sadness that im literally amazed im still here that i still function that im still basically a friendly stable person'
0	'i would feel that a few words would be not only inadequate but a travesty' 'i attributed this depression to feeling inadequate against the unrealistic ideals of the lds church and while i still hold those ideals somewhat responsible i recognize this pattern of behavior' 'ive been resting and feeling generally unpleasant and queasy but in that frustrating background way where you dont feel right but cant place an exact cause'
4	'i was starting to feel scared for both of their safety and i wish those officers hadn t left no matter how much i hated them' 'i am already feeling frantic' 'i believe in you moment we all feel til then it s one more skeptical song'
2	'i do feel sympathetic to the parties involved now that their careers are down the drain' 'i like frappes and shit when im feeling naughty but i drink tea daily' 'i will pay a month for months and feel shame every time i grill a hot dog from that point on'

Evaluation

Metrics

Label	Accuracy
all	0.5225

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-emotion")
# Run inference
preds = model("i am feeling very indecisive and spontaneous")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	4	19.3333	48

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	0.3009	-
0.2083	50	0.1916	-
0.4167	100	0.0393	-
0.625	150	0.0129	-
0.8333	200	0.0034	-

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}
}