allenai/Olmo-3-7B-Think-SFT

Text Generation • 7B • Updated about 10 hours ago • 1.75k • 5

Note Our 7B Think model lineage.

Note Our 32B Think SFT dataset is slightly further filtered than the 7B version above.

Note The new Python code mix we created for our thinking models. We took a subset of this for our full Think SFT Mix.

allenai/Olmo-3-7B-RL-Zero-IF

Text Generation • 528k • Updated about 10 hours ago • 96 • 3

Note Our RL Zero 7B models with intermediate checkpoints per domain.

Note Our RL Zero datasets.