does anyone know where to find the recipe for training (from base model to Olmo 3 RL Zero) this model?
#2
by
dullJackplaysallday
- opened
i just cannot find it anywhere
i guess the launch script is probably this one:
https://github.com/allenai/open-instruct/blob/main/scripts/train/olmo3/7b_rlzero_math.sh
but what about dataset preparation, etc.?