Deep Learning/Vision-Language Models (Focusing on Indic Scripts), System Design, Advanced Data Structures.
Currently hacking on
Edge AI Optimization (Sub-100MB AI Models). Compressing state-of-the-art Transformer models(BERT) via pruning and quantization. Aim: 95%+ accuracy at 10x size reduction for offline deployment of tools
Available for
Open Source Contribution (Security/Accessibility projects), Hackathons, Technical Writing & Collaboration.
Want to connect with shambhavi525-sudo?
Create an account to connect with shambhavi525-sudo. You can also sign in below to proceed if you already have an account.