Deduplicated Skywork and DeepScaleR math pools, plus their merged version. Only Math-Verify parser-compatible references.
Zbib M PRO
zbeeb
AI & ML interests
KAUST - AUB
Recent Activity
updated a collection 7 days ago
Exact-Age RL: Verifiable Math Data updated a dataset 7 days ago
zbeeb/Skywork-DeepScaleR-Merged-Verifiable-Dedup published a dataset 7 days ago
zbeeb/Skywork-DeepScaleR-Merged-Verifiable-DedupOrganizations
Exact-Age RL: Verifiable Math Data
Deduplicated Skywork and DeepScaleR math pools, plus their merged version. Only Math-Verify parser-compatible references.
Staleness: GRPO Models & Data
GRPO checkpoints organized by staleness cap, with the shared 17,005-row DAPO math dataset, provenance and benchmark scores.
models 27
zbeeb/Qwen2.5-Math-7B-GRPO-Staleness-4
Text Generation • 8B • Updated • 112
zbeeb/Qwen2.5-3B-GRPO-Staleness-6
Text Generation • 3B • Updated • 220
zbeeb/Qwen2.5-3B-GRPO-Staleness-8
Text Generation • 3B • Updated • 192
zbeeb/Qwen2.5-Math-1.5B-GRPO-Staleness-8
Text Generation • 2B • Updated • 292
zbeeb/Qwen2.5-3B-GRPO-Staleness-4
Text Generation • 3B • Updated • 359
zbeeb/Qwen2.5-Math-1.5B-GRPO-Staleness-6
Text Generation • 2B • Updated • 399
zbeeb/Qwen2.5-Math-1.5B-GRPO-Staleness-4
Text Generation • 2B • Updated • 184
zbeeb/Qwen2.5-Math-7B-GRPO-Staleness-2
Text Generation • 8B • Updated • 495
zbeeb/Qwen2.5-3B-GRPO-Staleness-2
Text Generation • 3B • Updated • 500
zbeeb/Qwen2.5-Math-1.5B-GRPO-Staleness-2
Text Generation • 2B • Updated • 480
datasets 12
zbeeb/Skywork-DeepScaleR-Merged-Verifiable-Dedup
Viewer • Updated • 98.9k • 39
zbeeb/DeepScaleR-Verifiable-Dedup
Viewer • Updated • 37.7k • 38
zbeeb/Skywork-OR1-Math-Verifiable-Dedup
Viewer • Updated • 97.8k • 42
zbeeb/Staleness-GRPO-DAPO-Math-17k
Viewer • Updated • 17k • 110
zbeeb/dapo-math-17k-qwen3-clean
Viewer • Updated • 17k • 28
zbeeb/beneficial-dpo-dataset
Viewer • Updated • 10k • 34
zbeeb/TAPS-Datasets
Viewer • Updated • 210k • 34 • 4
zbeeb/mixeddatasafety7030
Viewer • Updated • 10k • 2
zbeeb/llama3_MathInstruct_data
Updated • 5
zbeeb/unsafe
Viewer • Updated • 200 • 5