2907 update-- 2.204126 validation loss on math instruct dfa471f verified DarwinAnim8or commited on Jul 29
Update readme to reference this copy of the token distribution graph 754751b verified DarwinAnim8or commited on Jul 28
Pretraining step 2-- 100k openwebtext (val loss 4.834130) 19e72e8 verified DarwinAnim8or commited on Jul 27