jlee03046/hinglish-sentiment-xlmr

Hinglish Sentiment Analysis Model

First free open-source sentiment analysis model for Hinglish (Hindi-English code-mixed) text.

Model Details

• Base Model: XLM-RoBERTa Base
• Task: 3-class Sentiment Analysis (Positive / Negative / Neutral)
• Language: Hinglish (Hindi-English code-mixed)
• Dataset: 17,000+ real social media samples
• Test Accuracy: 68.5%

How to Use

```python from transformers import pipeline

classifier = pipeline("text-classification", model="jlee03046/hinglish-sentiment-xlmr")

result = classifier("yaar ye movie bahut bakwaas thi") print(result) # [{'label': 'LABEL_0', 'score': 0.91}] ```

Labels

• LABEL_0 = Negative
• LABEL_1 = Neutral
• LABEL_2 = Positive

Why This Matters

No free, open-source Hinglish sentiment model existed before this. Built for Indian startups, researchers, and developers working with Indian social media data.

Author

Built by aditya — fresher passionate about Indian language NLP.