muril

[SMM4H] COVID-19-related Nepali Tweets Classification in a Low Resource Setting

Extractive Nepali Question Answering System | Browser Extension & Web Application

Fine-tuned MuRIL for English/Hindi/Tamil sentiment. F1 0.898. FastAPI deployed.

A Malayalam fake news detection and fact verification system using MuRIL and llm

Multilingual phishing email detection with a language router across 5 model paths (fine-tuned XLM-RoBERTa, SecureBERT+MuRIL ensemble, TF-IDF fast path), explainable verdicts, and honest OOD evaluation no inflated benchmarks. FastAPI + React + Chrome extension + Docker. Privacy-first: no email content stored.

Hindi/Tamil text classification using MuRIL + PEFT/LoRA — MLflow experiment tracking, W&B, FastAPI serving

Hybrid retrieval and RAG research framework for low-resource Bhojpuri and Indic NLP.

This project detects abusive and non-abusive comments in Malayalm Language using the MuRIL Bert model and compares its performance with TF-IDF + SVM and XGBoost. MuRIL outperforms classical models.

AI-powered multilingual occupational coding system with semantic embeddings, cultural context, and adaptive learning.

This repository contains Python implementations for processing multilingual text data, focusing on language classification and translation tasks. The project addresses two distinct tasks: language classification and English translation, each involving different complexities in the processing of text data.

FAQ Search Application using multilingual sentence transformer model.

This project focuses on automatically identifying multiple Alankars in Hindi poetry using hyrbid models. It handles the complexity of Multi-Label Classification where a single verse may contain multiple figures of speech. It aims to support literary analysis and low-resource language processing through multilabel classification.

This project explores the application of various Language Learning Models (LLMs) for processing and understanding Sanskrit texts. We evaluate models like MuRIL, BERT, SanBERT, and ByT5 to analyze their efficiency in handling the complexities of Sanskrit through fertility score metrics.

AI that reads tender documents and bids, checks every eligibility criterion, explains every verdict, and flags doubts for human review - making CRPF procurement faster, fairer, and fully auditable.

News headline classification in English, Hindi, Gujarati & Marathi using MuRIL, DistilBERT & stacking ensemble. Published at Smart Systems 2026.