Hugging Face's logo

ronalhung
/
setfit-absa-restaurants-polarity

Text Classification
setfit
Safetensors
sentence-transformers
mpnet
absa
generated_from_setfit_trainer
Eval Results
Model card Files
xet
Community
setfit-absa-restaurants-polarity / README.md
ronalhung's picture
ronalhung
Add SetFit ABSA model
7fb9978 verified
preview code
|
raw
history blame
9.87 kB
metadata
tags:
  - setfit
  - absa
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
widget:
  - text: food:The food was bland oily.
  - text: >-
      soups:An oasis of refinement:  Food, though somewhat uneven, often reaches
      the pinnacles of new American fine cuisine - chef's passion (and kitchen's
      precise execution) is most evident in the fish dishes and soups.
  - text: lobster sandwich:We had the lobster sandwich and it was FANTASTIC.
  - text: >-
      sommlier:I understand the area and folks you need not come here for the
      romantic, alluring ambiance or the five star service featuring a sommlier
      and a complicated maze of captain and back waiters - you come for the
      authentic foods, the tastes, the experiance.
  - text: food:Not impressed with the food.
metrics:
  - accuracy
pipeline_tag: text-classification
library_name: setfit
inference: false
base_model: sentence-transformers/all-MiniLM-L6-v2
model-index:
  - name: SetFit Aspect Model with sentence-transformers/all-MiniLM-L6-v2
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: Unknown
          type: unknown
          split: test
        metrics:
          - type: accuracy
            value: 0.8377192982456141
            name: Accuracy

SetFit Aspect Model with sentence-transformers/all-MiniLM-L6-v2

This is a SetFit model that can be used for Aspect Based Sentiment Analysis (ABSA). This SetFit model uses sentence-transformers/all-MiniLM-L6-v2 as the Sentence Transformer embedding model. A LogisticRegression 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:

  1. Fine-tuning a Sentence Transformer with contrastive learning.
  2. 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:

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

Model Details

Model Description

Model Sources

Model Labels

Label Examples
aspect
  • 'staff:But the staff was so horrible to us.'
  • "food:To be completely fair, the only redeeming factor was the food, which was above average, but couldn't make up for all the other deficiencies of Teodora."
  • "food:The food is uniformly exceptional, with a very capable kitchen which will proudly whip up whatever you feel like eating, whether it's on the menu or not."
no aspect
  • "factor:To be completely fair, the only redeeming factor was the food, which was above average, but couldn't make up for all the other deficiencies of Teodora."
  • "deficiencies:To be completely fair, the only redeeming factor was the food, which was above average, but couldn't make up for all the other deficiencies of Teodora."
  • "Teodora:To be completely fair, the only redeeming factor was the food, which was above average, but couldn't make up for all the other deficiencies of Teodora."

Evaluation

Metrics

Label Accuracy
all 0.8377

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(
    "ronalhung/setfit-absa-restaurants-polarity",
    "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 4 18.0 37
Label Training Sample Count
no aspect 73
aspect 128

Training Hyperparameters

  • batch_size: (128, 128)
  • 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
0.0058 1 0.2702 -
0.2907 50 0.2764 0.2555
0.5814 100 0.1827 0.2240
0.8721 150 0.0239 0.2296
1.1628 200 0.0027 0.2436
1.4535 250 0.0015 0.2421
1.7442 300 0.001 0.2411
2.0349 350 0.0007 0.2431
2.3256 400 0.0005 0.2391
2.6163 450 0.0004 0.2470
2.9070 500 0.0004 0.2381
3.1977 550 0.0003 0.2465
3.4884 600 0.0003 0.2452
3.7791 650 0.0003 0.2478
4.0698 700 0.0003 0.2416
4.3605 750 0.0003 0.2453
4.6512 800 0.0002 0.2433
4.9419 850 0.0003 0.2447

Framework Versions

  • Python: 3.11.12
  • SetFit: 1.1.2
  • Sentence Transformers: 3.4.1
  • spaCy: 3.8.5
  • Transformers: 4.51.1
  • PyTorch: 2.6.0+cu124
  • Datasets: 3.5.0
  • 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}
}