subham18/setfit-paraphrase-mpnet-base-v2-twitter-sentiment

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: 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
Positive	'.@JohnKasich won this debate with a little home field advantage. #GOPDebates' 'RT @Mike_Surtel: @megynkelly your questions were more like attacks on @realDonaldTrump. Then u get upset when he got tough with u! What a jâ\x80¦' 'RT @kwrcrow: Congrats to @realDonaldTrump for your win in #GOPDebates polling last night. @Time @DRUDGE_REPORT Well done Sir! http://t.co/nâ\x80¦'
Neutral	'RT @CharleneCac: So does his position on Iran mean that Rick Perry is also pro-divestment from Israel? #GOPDebate' "We Watched The Debate With A Bunch Of Conservative Activists. Here's How They Reacted #GOPDebate http://t.co/Ug21fI5FcE via @dailycaller" "I loved the cluelessness of invoking Reagan's name on #IranDeal at #GOPDebate considering Reagan made deals w/ them."
Negative	"beeteedubs. If you have to play 'Lesser-of-17-Evils' with your party ... perhaps you need a new party. #p2 #tcot #GOPDebate" "RT @Ornyadams: Single payer... no way! I would miss paying ten different bills after my annual physical. Where's the fun in writing one cheâ\x80¦" "RT @madyclahane: srry rather not have decisions over my body being made by men that can't count to two #GOPDebate https://t.co/1Ps81yQaOl"

Evaluation

Metrics

Label	Accuracy
all	0.5307

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("subham18/setfit-paraphrase-mpnet-base-v2-twitter-sentiment")
# Run inference
preds = model("Who do you think won the #GOPDebate last night?")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	8	18.0833	25

Label	Training Sample Count
Negative	8
Positive	8
Neutral	8

Training Hyperparameters

• batch_size: (16, 16)
• num_epochs: (4, 4)
• 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
0.0417	1	0.2934	-
1.0	24	-	0.263
2.0	48	-	0.2555
2.0833	50	0.0091	-
3.0	72	-	0.2598
4.0	96	-	0.261

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.12.3
• SetFit: 1.0.3
• Sentence Transformers: 3.0.1
• Transformers: 4.39.0
• PyTorch: 2.4.0+cu121
• Datasets: 2.21.0
• Tokenizers: 0.15.2

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}
}