SetFit Polarity Model with sentence-transformers/all-mpnet-base-v2

This is a SetFit model that can be used for Aspect Based Sentiment Analysis (ABSA). This SetFit model uses sentence-transformers/all-mpnet-base-v2 as the Sentence Transformer embedding model. A LogisticRegression instance is used for classification. In particular, this model is in charge of classifying aspect polarities.

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.

This model was trained within the context of a larger system for ABSA, which looks like so:

Use a spaCy model to select possible aspect span candidates.
Use a SetFit model to filter these possible aspect span candidates.
Use this SetFit model to classify the filtered aspect span candidates.

Model Details

Model Description

Model Type: SetFit
Sentence Transformer body: sentence-transformers/all-mpnet-base-v2
Classification head: a LogisticRegression instance
spaCy Model: en_core_web_sm
SetFitABSA Aspect Model: setfit-absa-aspect
SetFitABSA Polarity Model: najwaa/absa-digital_cameras-polarity-p2
Maximum Sequence Length: 384 tokens
Number of Classes: 3 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
positive	'The autofocus is lightning fast:The autofocus is lightning fast and the image quality is absolutely stunning for portraits.' 'fast and the image quality is absolutely stunning:The autofocus is lightning fast and the image quality is absolutely stunning for portraits.' 'sharp, the auto focus is easy to:the images are sharp, the auto focus is easy to use, and the lens is decent quality.'
negative	', but the menu system is confusing for:Beautiful sharp photos with vibrant colors, but the menu system is confusing for beginners.' ', though the LCD screen is hard to:The lens produces crisp images even at maximum zoom, though the LCD screen is hard to see in bright sunlight.' 'on, but settings navigation is cumbersome.:The zoom range is impressive and focus accuracy is spot-on, but settings navigation is cumbersome.'
negative	'not worth the price value.:definitely not worth the price value.'

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 AbsaModel

# Download from the 🤗 Hub
model = AbsaModel.from_pretrained(
    "setfit-absa-aspect",
    "najwaa/absa-digital_cameras-polarity-p2",
)
# Run inference
preds = model("The food was great, but the venue is just way too busy.")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	7	20.5822	57

Label	Training Sample Count
negative	97
negative	1
positive	115

Training Hyperparameters

batch_size: (32, 32)
num_epochs: (5, 5)
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: True
warmup_proportion: 0.1
l2_weight: 0.01
seed: 42
eval_max_steps: -1
load_best_model_at_end: True

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0014	1	0.3121	-
0.0140	10	-	0.2740
0.0280	20	-	0.2615
0.0420	30	-	0.2430
0.0560	40	-	0.2219
0.0700	50	0.2693	0.1975
0.0840	60	-	0.1651
0.0980	70	-	0.1169
0.1120	80	-	0.0611
0.1261	90	-	0.0338
0.1401	100	0.126	0.0204
0.1541	110	-	0.0076
0.1681	120	-	0.0071
0.1821	130	-	0.0047
0.1961	140	-	0.0032
0.2101	150	0.0126	0.0029
0.2241	160	-	0.0027
0.2381	170	-	0.0032
0.2521	180	-	0.0035
0.2661	190	-	0.0032
0.2801	200	0.0044	0.0027
0.2941	210	-	0.0027

Framework Versions

Python: 3.11.12
SetFit: 1.1.2
Sentence Transformers: 4.1.0
spaCy: 3.7.5
Transformers: 4.51.3
PyTorch: 2.6.0+cu124
Datasets: 3.6.0
Tokenizers: 0.21.1

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}
}

najwaa
/

absa-digital_cameras-polarity-p2