Worked on fine-tuning a large language model (Falcon7B and Alpaca7B) on a custom dataset. Prepared an instruction dataset (stanford-alpaca format) for training a large language model for extracting and answering questions about resume data. Built langchain pipelines to analyze model performance on several prompts.

Worked with multiple Object Detection and Character Recognition models to design a highly efficient pipeline to be deployed on a jetson nano. Collaborated with company members to make their robot capable of multi-floor navigation by recognising elevator buttons.

Represented my institute in all Google events and workshops. Organized a guest lecture by a Google employee, led a GSOC orientation session, and conducted multiple workshops in collaboration with Google.