ardi555/setfit_reuters21578_reducedto15

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 OneVsRestClassifier 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 OneVsRestClassifier instance
• Maximum Sequence Length: 512 tokens

Model Sources

• Repository: SetFit on GitHub
• Paper: Efficient Few-Shot Learning Without Prompts
• Blogpost: SetFit: Efficient Few-Shot Learning Without Prompts

Evaluation

Metrics

Label	Accuracy
all	0.7852

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("ardi555/setfit_reuters21578_reducedto15")
# Run inference
preds = model("Oper shr 69 cts vs 83 cts
    Oper net 35.9 mln vs 42.4 mln
    Revs 798.9 mln vs 659.2 mln
    Avg shrs 52.0 mln vs 50.9 mln
    Nine mths
    Oper shr 2.38 dlrs vs 2.75 dlrs
    Oper net 123.3 mln vs 135.6 mln
    Revs 2.31 billion vs 1.86 billion
    Avg shrs 51.8 mln vs 49.3 mln
    NOTE: Net excludes losses from discontinued operations of
nil vs 16.1 mln dlrs in quarter and 227.5 mln dlrs vs 42.7 mln
dlrs in nine mths.
    Quarter net includes gains from sale of aircraft of two mln
dlrs vs 6,200,000 dlrs.
 Reuter
")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	1	181.1067	788

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
• l2_weight: 0.01
• seed: 42
• eval_max_steps: -1
• load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0013	1	0.4971	-
0.0667	50	0.1826	-
0.1333	100	0.1223	-
0.2	150	0.0699	-
0.2667	200	0.0712	-
0.3333	250	0.0646	-
0.4	300	0.055	-
0.4667	350	0.0611	-
0.5333	400	0.053	-
0.6	450	0.0555	-
0.6667	500	0.0475	-
0.7333	550	0.0716	-
0.8	600	0.0587	-
0.8667	650	0.0571	-
0.9333	700	0.0436	-
1.0	750	0.0505	-

Framework Versions

• Python: 3.10.12
• SetFit: 1.1.0
• Sentence Transformers: 3.2.1
• Transformers: 4.42.2
• PyTorch: 2.5.1+cu121
• Datasets: 3.1.0
• 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}
}