metadata
library_name: setfit
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
metrics:
- accuracy
widget:
- text: >-
La falta de entrega de la denuncia en el momento de la infracción, sin
causa que lo impida, genera indefensión y la invalida.
- text: >-
No es verdad que condujera bajo los efectos del alcohol. No había ingerido
ninguna bebida alcohólica, la prueba debió estar mal hecha.
- text: >-
Se sanciona sin explicar motivadamente las razones por las que se
considera acreditada la comisión de la infracción, pese a lo alegado por
el interesado.
- text: >-
Reclamo que se practique una reconstrucción de los hechos para demostrar
que la colisión se produjo por una maniobra imprudente del otro conductor.
- text: >-
La sanción no está suficientemente motivada, pues no se justifica la
valoración de la prueba efectuada ni el rechazo de las pruebas propuestas
por el interesado.
pipeline_tag: text-classification
inference: true
base_model: desarrolloasesoreslocales/bert-leg-al-corpus
model-index:
- name: SetFit with desarrolloasesoreslocales/bert-leg-al-corpus
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: Unknown
type: unknown
split: test
metrics:
- type: accuracy
value: 0.8
name: Accuracy
SetFit with desarrolloasesoreslocales/bert-leg-al-corpus
This is a SetFit model that can be used for Text Classification. This SetFit model uses desarrolloasesoreslocales/bert-leg-al-corpus 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: desarrolloasesoreslocales/bert-leg-al-corpus
- Classification head: a LogisticRegression instance
- Maximum Sequence Length: 4096 tokens
- Number of Classes: 15 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 |
---|---|
49 |
|
269 |
|
2014 |
|
78 |
|
32 |
|
12 |
|
2017 |
|
304 |
|
2027 |
|
357 |
|
2002 |
|
994 |
|
42 |
|
1002 |
|
1001 |
|
Evaluation
Metrics
Label | Accuracy |
---|---|
all | 0.8 |
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("desarrolloasesoreslocales/bert-leg-al-setfit")
# Run inference
preds = model("La falta de entrega de la denuncia en el momento de la infracción, sin causa que lo impida, genera indefensión y la invalida.")
Training Details
Training Set Metrics
Training set | Min | Median | Max |
---|---|---|---|
Word count | 14 | 22.4 | 37 |
Label | Training Sample Count |
---|---|
12 | 16 |
32 | 18 |
42 | 16 |
49 | 16 |
78 | 18 |
269 | 17 |
304 | 16 |
357 | 13 |
994 | 18 |
1001 | 18 |
1002 | 18 |
2002 | 16 |
2014 | 19 |
2017 | 17 |
2027 | 19 |
Training Hyperparameters
- batch_size: (24, 24)
- num_epochs: (1, 1)
- max_steps: -1
- sampling_strategy: oversampling
- num_iterations: 8
- 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 |
---|---|---|---|
0.0059 | 1 | 0.2923 | - |
0.2941 | 50 | 0.143 | - |
0.5882 | 100 | 0.1346 | - |
0.8824 | 150 | 0.0339 | - |
Framework Versions
- Python: 3.10.12
- SetFit: 1.0.3
- Sentence Transformers: 2.6.1
- Transformers: 4.38.2
- PyTorch: 2.2.1+cu121
- Datasets: 2.18.0
- 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}
}