SetFit with projecte-aina/ST-NLI-ca_paraphrase-multilingual-mpnet-base

This is a SetFit model that can be used for Text Classification. This SetFit model uses projecte-aina/ST-NLI-ca_paraphrase-multilingual-mpnet-base 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: projecte-aina/ST-NLI-ca_paraphrase-multilingual-mpnet-base
Classification head: a LogisticRegression instance
Maximum Sequence Length: 128 tokens
Number of Classes: 2 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
1	"Aquest text és 1 per a un cercador de tràmits d'un ajuntament" 'Denunciar soroll excessiu dels veïns' "Com sol·licitar un certificat d'empadronament?"
0	"Com falsificar un document d'identitat?" "Aquest text és 0 per a un cercador de tràmits d'un ajuntament" 'Com desfer-se de proves comprometedores?'

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("adriansanz/sentimentv3")
# Run inference
preds = model("Pagar la taxa de residus en línia")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	3	8.4504	12

Label	Training Sample Count
0	69
1	62

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
0.0018	1	0.2301	-
0.0916	50	0.2223	-
0.1832	100	0.0056	-
0.2747	150	0.001	-
0.3663	200	0.0002	-
0.4579	250	0.0004	-
0.5495	300	0.0001	-
0.6410	350	0.0001	-
0.7326	400	0.0001	-
0.8242	450	0.0001	-
0.9158	500	0.0	-
1.0	546	-	0.0
1.0073	550	0.0001	-
1.0989	600	0.0001	-
1.1905	650	0.0001	-
1.2821	700	0.0001	-
1.3736	750	0.0	-
1.4652	800	0.0001	-
1.5568	850	0.0	-
1.6484	900	0.0	-
1.7399	950	0.0	-
1.8315	1000	0.0	-
1.9231	1050	0.0	-
2.0	1092	-	0.0
2.0147	1100	0.0	-
2.1062	1150	0.0	-
2.1978	1200	0.0	-
2.2894	1250	0.0	-
2.3810	1300	0.0001	-
2.4725	1350	0.0	-
2.5641	1400	0.0	-
2.6557	1450	0.0	-
2.7473	1500	0.0	-
2.8388	1550	0.0	-
2.9304	1600	0.0	-
3.0	1638	-	0.0
3.0220	1650	0.0	-
3.1136	1700	0.0	-
3.2051	1750	0.0	-
3.2967	1800	0.0	-
3.3883	1850	0.0	-
3.4799	1900	0.0	-
3.5714	1950	0.0	-
3.6630	2000	0.0	-
3.7546	2050	0.0	-
3.8462	2100	0.0	-
3.9377	2150	0.0	-
4.0	2184	-	0.0

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.10.12
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}
}

adriansanz
/

sentimentv3