List your company

Efficiency Analysis on Alignment (train)

5Training TimeRLOO
-16125.75267.5409.25Mar 5, 2025

Evaluation results

MethodLinks
RLOOTraining Mode=Single-Iteration2025.035
ZOPrOTraining Mode=Single-Iteration2025.03120
DPOTraining Mode=Single-Iteration2025.03149
PPOTraining Mode=Single-Iteration2025.03530