Matched OpenR1 SFT checkpoints and subsequent GRPO models for studying token-level behavior during the transition from SFT to RL.
🏝️ On Vacation
Zbib M
zbeeb
P(doom) Rejects the premise
AI & ML interests
KAUST - AUB
Recent Activity
updated a collection 1 day ago
OpenR1 SFT to GRPO: Token-Level Study published a dataset 1 day ago
zbeeb/openr1-token-level-training-logs-20261007 updated a dataset 1 day ago
zbeeb/openr1-token-level-training-logs-20261007Organizations
models 36
zbeeb/Qwen2.5-1.5B-Base-GRPO
Text Generation • 2B • Updated • 386
zbeeb/Qwen2.5-Math-1.5B-Base-GRPO
Text Generation • 2B • Updated • 382 • 1
zbeeb/Qwen2.5-3B-OpenR1-SFT-GRPO
Text Generation • 3B • Updated • 385
zbeeb/Qwen2.5-Math-1.5B-OpenR1-SFT-GRPO
Text Generation • 2B • Updated • 171
zbeeb/Qwen2.5-1.5B-OpenR1-SFT-GRPO
Text Generation • 2B • Updated • 169
zbeeb/Qwen2.5-3B-Instruct-OpenR1-SFT
Text Generation • 3B • Updated • 167
zbeeb/Qwen2.5-3B-OpenR1-SFT
Text Generation • 3B • Updated • 146
zbeeb/Qwen2.5-Math-1.5B-OpenR1-SFT
Text Generation • 2B • Updated • 148
zbeeb/Qwen2.5-1.5B-OpenR1-SFT
Text Generation • 2B • Updated • 149
zbeeb/Qwen2.5-Math-7B-GRPO-Staleness-4
Text Generation • 8B • Updated • 127
datasets 14
zbeeb/openr1-token-level-training-logs-20261007
Preview • Updated • 1.32k
zbeeb/OpenR1-SFT-Math-20k
Viewer • Updated • 60.8k • 386
zbeeb/Skywork-DeepScaleR-Merged-Verifiable-Dedup
Viewer • Updated • 98.9k • 40
zbeeb/DeepScaleR-Verifiable-Dedup
Viewer • Updated • 37.7k • 42
zbeeb/Skywork-OR1-Math-Verifiable-Dedup
Viewer • Updated • 97.8k • 51
zbeeb/Staleness-GRPO-DAPO-Math-17k
Viewer • Updated • 17k • 204
zbeeb/dapo-math-17k-qwen3-clean
Viewer • Updated • 17k • 33
zbeeb/beneficial-dpo-dataset
Viewer • Updated • 10k • 35
zbeeb/TAPS-Datasets
Viewer • Updated • 210k • 48 • 4
zbeeb/mixeddatasafety7030
Viewer • Updated • 10k • 7