metadata

base_model: sentence-transformers/all-mpnet-base-v2
library_name: setfit
metrics:
  - f1
pipeline_tag: text-classification
tags:
  - setfit
  - absa
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
widget:
  - text: >-
      bargain:Monday nights are a bargain at the $28 prix fix - this includes a
      three course meal plus *three* glasses of wine paired with each course.
  - text: seated:We walked in on a Wednesday night and were seated promptly.
  - text: >-
      drinks:While most people can attest to spending over $50 on drinks in New
      York bars and hardly feeling a thing, the drinks here are plentiful and
      unique.
  - text: Lassi:I ordered a Lassi and asked 4 times for it but never got it.
  - text: stomach:Check it out, it won't hurt your stomach or your wallet.
inference: false
model-index:
  - name: SetFit Aspect Model with sentence-transformers/all-mpnet-base-v2
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: Unknown
          type: unknown
          split: test
        metrics:
          - type: f1
            value: 0.923076923076923
            name: F1

SetFit Aspect 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 SetFitHead instance is used for classification. In particular, this model is in charge of filtering aspect span candidates.

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 this SetFit model to filter these possible aspect span candidates.
Use a 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 SetFitHead instance
spaCy Model: en_core_web_trf
SetFitABSA Aspect Model: MattiaTintori/Final_aspect_Colab
SetFitABSA Polarity Model: setfit-absa-polarity
Maximum Sequence Length: 384 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
aspect	'price:The price is reasonable although the service is poor.' 'service:The price is reasonable although the service is poor.' 'service:The place is so cool and the service is prompt and curtious.'
no aspect	'stomach:The food was delicious but do not come here on a empty stomach.' 'place:I grew up eating Dosa and have yet to find a place in NY to satisfy my taste buds.' 'NY:I grew up eating Dosa and have yet to find a place in NY to satisfy my taste buds.'

Evaluation

Metrics

Label	F1
all	0.9231

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(
    "MattiaTintori/Final_aspect_Colab",
    "setfit-absa-polarity",
)
# 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	3	19.4137	62

Label	Training Sample Count
no aspect	430
aspect	711

Training Hyperparameters

batch_size: (64, 4)
num_epochs: (5, 32)
max_steps: -1
sampling_strategy: oversampling
num_iterations: 10
body_learning_rate: (8e-05, 8e-05)
head_learning_rate: 0.04
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
0.0028	1	0.2878	-
0.0560	20	0.2409	0.2515
0.1120	40	0.2291	0.2319
0.1681	60	0.1354	0.1835
0.2241	80	0.0654	0.1389
0.2801	100	0.0334	0.1818
0.3361	120	0.0535	0.1408
0.3922	140	0.014	0.1564
0.4482	160	0.0119	0.1453
0.5042	180	0.0158	0.1511
0.5602	200	0.0157	0.1393
0.6162	220	0.005	0.1536
0.6723	240	0.0002	0.1546
0.7283	260	0.0002	0.1673
0.7843	280	0.0004	0.1655

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.10.12
SetFit: 1.0.3
Sentence Transformers: 3.0.1
spaCy: 3.7.6
Transformers: 4.39.0
PyTorch: 2.3.1+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}
}