allenai/tulu-3-sft-mixture
Viewer
•
Updated
•
939k
•
12
All datasets released with Tulu 3 -- state of the art open post-training recipes.
Note Our main SFT mixture.
Note The full preference mixture used for DPO on our 8B SFT checkpoint.
Note The full preference mixture used for DPO on our 70B SFT checkpoint.
Note The rest from here are individual new SFT or preference datasets we created!