File size: 8,469 Bytes
944e24d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
---
base_model: FacebookAI/xlm-roberta-base
library_name: setfit
metrics:
- accuracy
pipeline_tag: text-classification
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
widget:
- text: Banco Sabadell confirma el día de pago de pensiones para jubilados en agosto.
- text: Banco Santander supera resistencias y avanza hacia máximos anuales, lo que
    tiene implicaciones para los inversores.
- text: Abre una cuenta online gratuita en BBVA, domicilia tu nómina durante 12 meses
    y recibe 250 usando el código 90030031951793.
- text: MyInvestor tiene una grave falta de oferta en acciones individuales y sus
    comisiones son peores que las de ING en ese mismo ámbito.
- text: Los recicladores están durmiendo en la vereda del BBVA y el fin de semana
    dentro del cajero, mientras la seguridad parece ausente.
inference: true
model-index:
- name: SetFit with FacebookAI/xlm-roberta-base
  results:
  - task:
      type: text-classification
      name: Text Classification
    dataset:
      name: Unknown
      type: unknown
      split: test
    metrics:
    - type: accuracy
      value: 0.7138461538461538
      name: Accuracy
---

# SetFit with FacebookAI/xlm-roberta-base

This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [FacebookAI/xlm-roberta-base](https://huggingface.co/FacebookAI/xlm-roberta-base) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) 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](https://www.sbert.net) 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:** [FacebookAI/xlm-roberta-base](https://huggingface.co/FacebookAI/xlm-roberta-base)
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
- **Maximum Sequence Length:** 512 tokens
- **Number of Classes:** 2 classes
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->

### Model Sources

- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)

### Model Labels
| Label    | Examples                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      |
|:---------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| discard  | <ul><li>'Las negociaciones para el Banco de España avanzan rápidamente y el Congreso convocará el jueves la comisión de economía para anunciar los nombres pactados, con Conthe como la candidata más firme a gobernadora.'</li><li>'Depósitos y seguros son aspectos fundamentales para atraer clientes y potenciar el negocio de Caixabank en la segunda mitad del año.'</li><li>'El Banco Santander ofrece 400€ al cambiar tu nómina a su cuenta en línea, eliminando las comisiones bancarias.'</li></ul> |
| relevant | <ul><li>'Nuevo caso de phishing relacionado con Evobanco registrado el 13 de julio de 2024.'</li><li>'El Banco Sabadell ofrece depósitos a plazo fijo con un interés del 2,5% TAE a 1 año y 3% TAE a 6 meses, lo cual es una buena opción.'</li><li>'Estoy en Abanca porque no me cobran comisiones, de lo contrario ya los habría dejado.'</li></ul>                                                                                                                                                         |

## Evaluation

### Metrics
| Label   | Accuracy |
|:--------|:---------|
| **all** | 0.7138   |

## Uses

### Direct Use for Inference

First install the SetFit library:

```bash
pip install setfit
```

Then you can load this model and run inference.

```python
from setfit import SetFitModel

# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("saraestevez/setfit-xlm-bank-tweets-processed-80")
# Run inference
preds = model("Banco Sabadell confirma el día de pago de pensiones para jubilados en agosto.")
```

<!--
### Downstream Use

*List how someone could finetune this model on their own dataset.*
-->

<!--
### Out-of-Scope Use

*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->

<!--
## Bias, Risks and Limitations

*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->

<!--
### Recommendations

*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->

## Training Details

### Training Set Metrics
| Training set | Min | Median  | Max |
|:-------------|:----|:--------|:----|
| Word count   | 1   | 21.0437 | 36  |

| Label    | Training Sample Count |
|:---------|:----------------------|
| discard  | 80                    |
| relevant | 80                    |

### Training Hyperparameters
- batch_size: (16, 16)
- 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 |
|:------:|:----:|:-------------:|:---------------:|
| 0.0025 | 1    | 0.4924        | -               |
| 0.125  | 50   | 0.2519        | -               |
| 0.25   | 100  | 0.186         | -               |
| 0.375  | 150  | 0.188         | -               |
| 0.5    | 200  | 0.0504        | -               |
| 0.625  | 250  | 0.0412        | -               |
| 0.75   | 300  | 0.0147        | -               |
| 0.875  | 350  | 0.0517        | -               |
| 1.0    | 400  | 0.0162        | -               |

### Framework Versions
- Python: 3.11.0rc1
- SetFit: 1.0.3
- Sentence Transformers: 2.7.0
- Transformers: 4.39.0
- PyTorch: 2.3.1+cu121
- Datasets: 2.19.1
- Tokenizers: 0.15.2

## Citation

### BibTeX
```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}
}
```

<!--
## Glossary

*Clearly define terms in order to be accessible across audiences.*
-->

<!--
## Model Card Authors

*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->

<!--
## Model Card Contact

*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->