TLDR AI · QUICK LINKS
OLMO-CORE 3 FOR MOE TRAINING
Summary from the TLDR AI newsletter
Olmo-core 3 is an open training framework designed to scale mixture-of-experts models into the trillion-parameter range.
TLDR AI · QUICK LINKS
Olmo-core 3 is an open training framework designed to scale mixture-of-experts models into the trillion-parameter range.