List your company

Reasoning on Mathematical Reasoning (train)

1,177.8MTL (Loss)SFT
1,158.481,288.891,419.31,549.71Sep 29, 2025

Evaluation results

MethodLinks
SFTTraining Time=9m 36s2025.091,177.875.8
R1-Qwen-7B (Base)Training Time=/2025.091,24476.9
RL (GRPO)Training Time=4h 09m 35s2025.091,481.993.2
FARLTraining Time=4h 22m 53s2025.091,660.892.4