The official Repository of RETU: Rethinking Expert Trajectory Utilization in LLM Post-training
Ding Bowen
Daniel21Ding
AI & ML interests
DL training and inference
Recent Activity
upvoted a paper about 9 hours ago
ProAS: A Progress-Aware Advantage Shaping Framework for Multi-Turn Agentic Reinforcement Learning submitted a paper about 10 hours ago
ProAS: A Progress-Aware Advantage Shaping Framework for Multi-Turn Agentic Reinforcement Learning upvoted a paper 13 days ago
TPPO: Tracking-Value PPO For Language Model ReasoningOrganizations
None yet
models 23
Daniel21Ding/sft-qwen-2.5-7b-sp2-liger
Updated
Daniel21Ding/ft-s3-siglip-qwen2d5-7b-ins-gqa-caption-text
8B • Updated • 9
Daniel21Ding/ft-siglip-qwen2d5-7b-ins-gqa-text-mix-ocrvqa
8B • Updated • 8
Daniel21Ding/ft-siglip-qwen2d5-1d5b-ins-gqa-text-mix
2B • Updated • 9
Daniel21Ding/ft-siglip-qwen2d5-1d5b-ins-gqa
2B • Updated • 12
Daniel21Ding/pt-siglip-qwen2d5-7b-ins-llava-ckpts
8B • Updated • 14
Daniel21Ding/ft-siglip-qwen2d5-7b-ins-gqa-text-mix-general
Updated
Daniel21Ding/ft-s3-ocrvqa-siglip-qwen2d5-7b-ins-gqatextmix
Updated
Daniel21Ding/ft-s3-siglip-qwen2d5-7b-ins-gqa-text-mix
8B • Updated • 11
Daniel21Ding/ft-s2-siglip-qwen2d5-7b-ins-llava-ocrvqa
8B • Updated • 9
datasets 9
Daniel21Ding/uniform102K
Viewer • Updated • 102k • 7
Daniel21Ding/S1K_1.1
Viewer • Updated • 1k • 6
Daniel21Ding/RL62K
Viewer • Updated • 62.3k • 11
Daniel21Ding/benchmark
Viewer • Updated • 2.91k • 10
Daniel21Ding/SFT889K
Viewer • Updated • 901k • 12
Daniel21Ding/Val-199
Viewer • Updated • 199 • 11
Daniel21Ding/MIX37K
Viewer • Updated • 36.7k • 7
Daniel21Ding/Hard102K
Viewer • Updated • 102k • 7
Daniel21Ding/Easy102K
Viewer • Updated • 102k • 14