38u3u/sentiment-demo-classifier

Sentiment Classification Demo

This repository packages distilbert/distilbert-base-uncased-finetuned-sst-2-english, a public DistilBERT model for binary English sentiment classification. It predicts POSITIVE or NEGATIVE.

Source and data

The model is sourced from distilbert/distilbert-base-uncased-finetuned-sst-2-english. It was originally fine-tuned on the SST-2 sentiment task; this assignment re-saves and republishes it and does not perform additional fine-tuning.

Usage

from transformers import pipeline
classifier = pipeline("text-classification", model="38u3u/sentiment-demo-classifier")
print(classifier("The product is excellent!"))

Limitations

The model is intended for English sentiment examples similar to its training data. It can misclassify sarcasm, mixed sentiment, domain-specific language, and other languages. Predictions should not be treated as objective truth.