README.md

---
language:
- ne
pretty_name: Nepali Sentiment Dataset
tags:
- sentiment-analysis
- text-classification
- nepali
license: mit
task_categories:
- text-classification
task_ids:
- sentiment-analysis
dataset_info:
  features:
    - name: text
      dtype: string
    - name: label
      dtype:
        class_label:
          names:
            - Negative
            - Neutral
            - Positive
size_categories:
- 10K<n<100K
---

# Nepali Sentiment Dataset (Devanagari)

## Dataset Summary
This dataset contains Nepali sentences in Devanagari script labeled with sentiment classes: **Negative**, **Neutral**, and **Positive**.

## Supported Tasks
- Sentiment analysis / text classification

## Languages
- Nepali (`ne`) — Devanagari script

## Dataset Structure

### Data Instances
Each example contains:
- `text`: Nepali sentence (string)
- `label`: one of {negative, neutral, positive}

Example:
```json
{"text":"यो फिल्म एकदम राम्रो लाग्यो", "label":"positive"}