CoolFace
Datasetpublic

RyanYr/dapo-math-17k-drop-never-solved-qwen3-1.7b-base-k16

DAPO-Math-17k, never-solved problems dropped (Qwen3-1.7B-Base, K=16) The 7,434 problems of the DAPO-Math-17k train set that Qwen3-1.7B-Base solved at least once in K=16 samples — the other 9,964 of 17,398 problems (57.27%), solved 0/16 times, are dropped. In group-relative RL a prompt group whose K responses all score the same has a reward advantage of exactly zero and contributes no gradient. A problem the starting policy solves 0/16 times can only produce such groups, so every… See the full description on the dataset page: https://huggingface.co/datasets/RyanYr/dapo-math-17k-drop-never-solved-qwen3-1.7b-base-k16.

sourceHugging Faceapache-2.0updated 29d agoView on Hugging Face
0likes41downloads
3 commits on main
372bf6429d ago

dataset card

RyanYr
cdd7e8129d ago

7434-problem subset of DAPO-Math-17k: never-solved (0/16 under Qwen3-1.7B-Base) dropped

RyanYr
a3a675729d ago

initial commit

RyanYr