docs/features/nepali_text_classifier.md

# Nepali Text Classifier

## Overview

This classifier identifies whether Nepali-language text content is written by a human or AI.

## Model

- Base Model: XLM-Roberta (XLMRClassifier)
- Language: Nepali (Multilingual model)
- Fine-tuned with scraped web content (~18,000 samples)

## Dataset

- Custom scraped dataset with manual labeling
- Includes news, blogs, and synthetic content from various LLMs

## Code Location

- Controller: `features/nepali_text_classifier/controller.py`
- Inference: `features/nepali_text_classifier/inferencer.py`
- Model Loader: `features/nepali_text_classifier/model_loader.py`

## API

- Endpoint: [ENDPOINTS](../api_endpoints.md)
- Input: Raw text
- Output: JSON classification with label and confidence score

[🔙 Back to Main README](../README.md)