metadata
library_name: setfit
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
metrics:
- accuracy
widget:
- text: >-
This paper focuses on mining association rules between sets of items in
large databases, which can reveal interesting patterns and relationships
among the data.
- text: >-
In this paper, the authors explore the economic concepts of fairness and
retaliation within the context of reciprocity, demonstrating how these
principles shape market behaviors and interactions.
- text: >-
Further research is needed to explore the applicability of the proposed
model to more complex multi-echelon inventory systems with additional
features, such as lead time variability and supplier reliability.
- text: >-
The NCEP/NCAR 40-Year Reanalysis Project provides retrospective
atmospheric data sets by assimilating observational data into a model,
resulting in improved estimates of historical weather patterns for
meteorological research and applications.
- text: >-
This study aims to assess the accuracy of aerosol optical properties
retrieved from Aerosol Robotic Network (AERONET) Sun and sky radiance
measurements using ground-based reference data.
pipeline_tag: text-classification
inference: true
base_model: jinaai/jina-embeddings-v2-small-en
model-index:
- name: SetFit with jinaai/jina-embeddings-v2-small-en
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: Unknown
type: unknown
split: test
metrics:
- type: accuracy
value: 0.8492307692307692
name: Accuracy
SetFit with jinaai/jina-embeddings-v2-small-en
This is a SetFit model that can be used for Text Classification. This SetFit model uses jinaai/jina-embeddings-v2-small-en 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: jinaai/jina-embeddings-v2-small-en
- Classification head: a LogisticRegression instance
- Maximum Sequence Length: 8192 tokens
- Number of Classes: 13 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 |
|---|---|
| Aims |
|
| Background |
|
| Hypothesis |
|
| Implications |
|
| Importance |
|
| Keywords |
|
| Limitations |
|
| Method |
|
| None |
|
| Purpose |
|
| Reccomendations |
|
| Result |
|
| Uncertainty |
|
Evaluation
Metrics
| Label | Accuracy |
|---|---|
| all | 0.8492 |
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("Corran/SciGenSetfit3")
# Run inference
preds = model("This paper focuses on mining association rules between sets of items in large databases, which can reveal interesting patterns and relationships among the data.")
Training Details
Training Set Metrics
| Training set | Min | Median | Max |
|---|---|---|---|
| Word count | 11 | 28.3123 | 71 |
| Label | Training Sample Count |
|---|---|
| Aims | 200 |
| Background | 200 |
| Hypothesis | 200 |
| Implications | 200 |
| Importance | 200 |
| Keywords | 200 |
| Limitations | 200 |
| Method | 200 |
| None | 200 |
| Purpose | 200 |
| Reccomendations | 200 |
| Result | 200 |
| Uncertainty | 200 |
Training Hyperparameters
- batch_size: (256, 256)
- num_epochs: (1, 1)
- max_steps: -1
- sampling_strategy: oversampling
- num_iterations: 40
- 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: False
Training Results
| Epoch | Step | Training Loss | Validation Loss |
|---|---|---|---|
| 0.0025 | 1 | 0.2913 | - |
| 0.1229 | 50 | 0.2365 | - |
| 0.2457 | 100 | 0.185 | - |
| 0.3686 | 150 | 0.159 | - |
| 0.4914 | 200 | 0.1456 | - |
| 0.6143 | 250 | 0.1658 | - |
| 0.7371 | 300 | 0.1189 | - |
| 0.8600 | 350 | 0.1235 | - |
| 0.9828 | 400 | 0.1282 | - |
| 0.0049 | 1 | 0.1257 | - |
| 0.0615 | 50 | 0.1371 | - |
| 0.1230 | 100 | 0.1226 | - |
| 0.1845 | 150 | 0.1099 | - |
| 0.2460 | 200 | 0.0897 | - |
| 0.3075 | 250 | 0.1009 | - |
| 0.3690 | 300 | 0.0659 | - |
| 0.4305 | 350 | 0.0711 | - |
| 0.4920 | 400 | 0.0745 | - |
| 0.5535 | 450 | 0.0807 | - |
| 0.6150 | 500 | 0.0736 | - |
| 0.6765 | 550 | 0.0571 | - |
| 0.7380 | 600 | 0.0649 | - |
| 0.7995 | 650 | 0.0672 | - |
| 0.8610 | 700 | 0.0586 | - |
| 0.9225 | 750 | 0.0624 | - |
| 0.9840 | 800 | 0.0614 | - |
Framework Versions
- Python: 3.10.12
- SetFit: 1.0.3
- Sentence Transformers: 2.2.2
- Transformers: 4.36.2
- PyTorch: 2.1.0+cu121
- Datasets: 2.16.1
- Tokenizers: 0.15.0
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}
}