my-sentiment-analyzer
Fine-tuned distilbert-base-uncased for binary sentiment classification (0 = negative, 1 = positive).
Intended Use
Classify English-language text (reviews, comments, feedback) as positive or negative sentiment.
Training Data
- IMDB movie reviews (25,000 train / 2,500 validation / 25,000 test)
Training Procedure
- Base model: distilbert-base-uncased
- Epochs: 2
- Learning rate: 2e-5
- Batch size: 32
- Max sequence length: 128
Evaluation Results
| Metric | Score |
|---|
| Accuracy | 0.873 |
| Precision | 0.862 |
| Recall | 0.887 |
| F1 | 0.875 |
Limitations
- Trained on movie reviews; may not generalize well to other domains without further fine-tuning.
- Binary classification only — does not detect neutral sentiment.
- English only.
Usage
from transformers import pipeline
classifier = pipeline("sentiment-analysis", model="kabra686/my-sentiment-analyzer")
result = classifier("This was a great experience!")
print(result)
