AI training costs continue to plummet. @databricks slashed their training costs by 4x in just 10 months, helped by their acquisition of @NaveenGRao's @DbrxMosaicAI. Faster than our modeled 75% per year cost decline. The pace of innovation in AI right now is unprecedented databricks.com/blog/introduci…
@downingARK @databricks @NaveenGRao @DbrxMosaicAI Its crazy that you guys sold NVDA. The demand for their products goes up dramatically as the usefulness of the models increase
@downingARK @barrydauber @databricks @NaveenGRao @DbrxMosaicAI To be fair, we're comparing to our previous effort, not the best previous-gen model out there. MPT was probably 2x less efficient than LLaMA2, so we had a lot of catching up to do in addition to some major efficiency improvements beyond that.