Meta Doubles Advertising Efficiency with GEM Training Innovations
Meta has doubled training efficiency for its GEM model in ad recommendations, overcoming engineering challenges through innovative solutions.
Meta has successfully doubled the training efficiency of its GEM model for ad recommendations on Instagram and Facebook. Achieving 20-25% Model FLOPs Utilization (MFU) while scaling training FLOPs 4x over the past year, this advancement addresses unique engineering challenges at the intersection of recommendation systems and LLMs, necessitating innovative solutions.