GradNova

Computational Techniques for Learning From Data That Are Highly Scalable With the Availability of the Compute

1 faculty research Computational Techniques for Learning From Data That Are Highly Scalable With the Availability of the Compute on GradNova.

I

Assistant Professor

Computer Science

University of Hawaiʻi at Mānoa

Computational Techniques for Learning From Data That Are Highly Scalable With the Availability of the Compute

Working on Optimizers and Model Architecture Scaling Schemes That Would Allow Predictable and Efficient Training of Transformer Models Containing Hundreds of Billions of Parameters