Optimizing AI systems by implementing TensorRT-based neural network graph optimization, resulting in significant performance improvements.

In this project, I start with the Llama-2 model with 7B parameters and fine-tune it for question-answering tasks on two new domains.

Implemented a solution directly from a published paper to predict 3D objects from fisheye cameras without training images, using the vanilla 3D object detector.

Optimizing neural networks via model pruning. The project highlights the efficiency of model pruning in increasing inference speed while retaining accuracy.