Training a large language model (LLM) is among the most costly and time consuming exercises for enterprises. A new open-source model being released today by ServiceNow could make a big difference, with the promise of training 20% faster, saving enterprises time and money.