Implemented a generative pretrained transformer (GPT) model using PyTorch that is trained on the life works of Shakespeare.

A Retrieval-Augmented Generation (RAG) Chatbot. It uses a FastAPI backend, a MongoDB-integrated chat history, and a Next.js frontend styled using TailwindCSS.

Developed a deep learning model that identifies the reconstruction kernel of a CT image and transforms it into an image generated with a different kernel.

SRGAN-based super-resolution model for upscaling 64×64 images to 125×125 with perceptual quality optimization using adversarial training