Ultra Small SLMs designed to test how much performance can be squeezed out of low tokens and low parameters.