🛍️ Product Review Sentiment Classifier
A fine-tuned DistilBERT model for classifying product reviews as Positive or Negative.
Fine-tuned on the Amazon Polarity dataset with 5,000 training samples.
📊 Model Performance
| Metric | Score |
|---|---|
| Accuracy | 92.10% |
| F1 Score | 0.9210 |
🚀 Usage
from transformers import pipeline
classifier = pipeline("text-classification", model="JuhiSaxena2002/opinion-mining")
result = classifier("This product is absolutely amazing! Best purchase ever.")
print(result) # [{'label': 'POSITIVE', 'score': 0.99}]
result = classifier("Terrible quality. Broke after one day. Very disappointed.")
print(result) # [{'label': 'NEGATIVE', 'score': 0.98}]
🏗️ Model Architecture
- Base Model:
distilbert-base-uncased - Task: Binary Sequence Classification
- Labels:
POSITIVE,NEGATIVE - Max Sequence Length: 128 tokens
- Training Epochs: 3
📦 Training Details
- Dataset: Amazon Polarity
- Train samples: 5,000
- Test samples: 1,000
- Learning Rate: 2e-5
- Batch Size: 32
- Optimizer: AdamW with warmup
🙋 Author
Built and fine-tuned by JuhiSaxena2002
- Downloads last month
- 45