Sparse Mixture-of-Experts for Summarization
PythonPyTorchTransformersNLP
Sparse MoE and Grouped Query Attention implemented from scratch for abstractive summarization
Undergraduate at IIIT Hyderabad. Interested in cybersecurity, machine learning, and systems programming. Rated 1608 on Codeforces.
Contributed to Microsoft PowerToys , a set of utilities for power users to streamline their Windows experience.
Sparse MoE and Grouped Query Attention implemented from scratch for abstractive summarization
Distributed file system with naming server architecture for coordinated operations
Feature-rich shell in C with custom commands, I/O redirection, and piping
Full-stack MERN marketplace with real-time chat and AI-powered support
Scalable monitoring dashboard querying 50K+ logs/day with sub-second latency
Evaluating genuine reasoning vs pattern matching in Large Language Models