Yelp Review Quality Classifier

This model classifies Yelp restaurant reviews into three quality levels:

  • Low Quality: Short reviews (<30 characters)
  • Medium Quality: Average-length reviews (30-100 characters)
  • High Quality: Detailed reviews (>100 characters)

Model Details

  • Base Model: DistilBERT (distilbert-base-uncased)
  • Task: Multi-class text classification (3 classes)
  • Training Samples: 8000
  • Validation Samples: 2000
  • Training Epochs: 3

Performance

  • Accuracy: 0.8140
  • F1 Macro: 0.4449
  • F1 Weighted: 0.7853
  • Precision: 0.7914
  • Recall: 0.8140

Usage

from transformers import pipeline

classifier = pipeline("text-classification", model="RLau33/yelp-review-quality-v2")
result = classifier("The food was amazing and the service was excellent!")
print(result)

Training Details

  • Learning Rate: 2e-05
  • Batch Size: 32
  • Max Length: 256 tokens
  • Device: GPU (Tesla T4)

License

MIT License

Downloads last month
56
Safetensors
Model size
67M params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Dataset used to train RLau33/yelp-review-quality-v2