Will Held(@WilliamBarrHeld):To train better open models, we need predictable scaling. Delphi is Marin’s first step: we pretrained many small models with one recipe, then extrapolated 300× to predict a 25B-param / 600B-token run with just 0.2% error. Getting there took some work 🧵

Will Held

@WilliamBarrHeld

Open LLM Training @ Formerly ML PhD w/ @Diyi_Yang, 🦙 @AIatMeta, Assistant @GoogleAI, اللغة العربية @NYUAbuDhabi Burqueño

加入 October 2012

1.1K 正在關注 2.6K 粉絲

Will Held@WilliamBarrHeld

2026.05.11 19:25

To train better open models, we need predictable scaling. Delphi is Marin’s first step: we pretrained many small models with one recipe, then extrapolated 300× to predict a 25B-param / 600B-token run with just 0.2% error. Getting there took some work 🧵